Low Vision Principle and Applications

Low Vision
George C. Woo
Editor
Low Vision
Principles and Applications
Proceedings of the International Symposium
on Low Vision, University of Waterloo,
June 25-27, 1986
With 145 Illustrations
Springer-Verlag
New York Berlin Heidelberg
London Paris Tokyo
George C. Woo
Centre for Sight Enhancement
School of Optometry
University of Waterloo
Waterloo, Ontario
Canada
Library of Congress Cataloging in Publication Data

International Symposium on Low Vision (1986 :
University of Waterloo)
Low vision.
1. Low vision-Congresses. 2. Low vision-Patients-
Rehabilitation-Congresses. I. Woo, George C.
II. University of Waterloo. III. Title.
RE91.158 1986 617.7'5 86-31434
© 1987 by Springer-Verlag New York Inc.

Softcover reprint of the hardcover 1st edition 1987
All rights reserved. This work may not be translated or copied in whole or in part without the
written permission of the publisher (Springer-Verlag, 175 Fifth Avenue, New York, New York
10010, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use
in connection with any form of information storage and retrieval, electronic adaptation, computer
software, or by similar or dissimilar methodology now known or hereafter developed is forbidden.
The use of general descriptive names, trade names, trademarks, etc. in this publication, even
if the former are not especially identified, is not to be taken as a sign that such names, as
understood by the Trade Marks and Merchandise Marks Act, may accordingly be used freely
by anyone.
Permission to photocopy for internal or personal use, or the internal or personal use of specific
clients, is granted by Springer-Verlag New York Inc. for libraries registered with the Copyright
Clearance Center (CCC), provided that the base fee of $0.00 per copy, plus $0.20 per page
is paid directly to CCC, 21 Congress Street, Salem, MA 01970, USA Special requests should
be addressed directly to Springer-Verlag New York, 175 Fifth Avenue, New York, New York
10010, USA
96538-6/1987 $0.00 + .20
987654321
ISBN-13: 978-1-4612-9152-7 e-ISBN-13: 978-1-4612-4780-7

001: 10.1007/978-1-4612-4780-7
Foreword
The study of optics and visual science has always had as its
implicit goal the development of practical knowledge and an
advanced technology which will enhance and extend the visual
capability of mankind. In no clinical area has this benefit
been more evident than in low vision.
The low vision clinic at the School of Optometry, Univer-

sity of Waterloo, was established in 1973 as a one man opera-
tion offering service one day per week to the partially sight-
ed. In thirteen short years the clinic has grown to include a
professional staff of three optometrists, a social worker, a
rehabilitation instructor as well as an administrative, techni-
cal and secretarial staff of four. Low vision services are
offered to the public five days a week. Approximately 1,000
individuals from every region of Canada benefited from such
service during the past year. The low vision aids prescribed
include a wide variety of optical and electronic devices, rang-
ing from simple magnifiers and telescopes to closed circuit
television systems and computer operated display units.
The School of Optometry has also become an important

center for basic and applied vision research. Most of this
work relates directly or indirectly to low vision. A variety of
studies are currently underway on such topics as visual field
loss, electrodiagnostic characteristics of various visual ano-
malies, the study of accommodative anomalies, spatial fre-
quency analysis in low vision patients, the effect of aging on
the optics of the eye, early cataract development and many
others.
The School of Optometry is proud of this record of

achievement. I can not think of a more appropriate site for a
low vision meeting. I congratulate Professor George Woo, Dr.
Rodger Pace and Ms. Lynne Hanna for organizing a produc-
tive and event-filled low vision symposium. This book will be
a lasting record of the contribution made by all those who
participated.
Jacob G. Sivak
Director, School of Optometry;
Associate Dean of Science for Optometry
Preface
In July of 1983, the idea of hosting a low vision symposium

was first conceived. It was formally proposed to the National
Low Vision Committee at the biennial meeting of the Canadi-
an Association of Optometrists in Vancouver. There was
immediate enthusiasm to promote such an event. Probably
this was due largely to the exciting developments that were
taking place around the world in the field of low vision
research and low vision clinical practice in the last ten years.
Subsequently it was decided that an International Symposium
would be held and a secretariat was then formed consisting of
Dr. Rodger Pace, Ms. Lynne Hanna and myself.
Excellent research and clinical papers were presented by

a number of internationally reknowned speakers from differ-
ent countries. The contents of this volume of proceedings are
the outcome of the forty-four papers presented at the sympo-
sium.
My special thanks go to Lynne Hanna without whose

organizational skills the symposium would not have run so
smoothly. I acknowledge support of the School of Optometry,
the Faculty of Science and the University of Waterloo which
was accorded to me in arranging this symposium. Financial
support was in part provided by a number of industrial con-
cerns and professional organizations listed separately.
The objective of the symposium was to provide an oppor-

tunity for low vision clinicians, administrators and research-
ers to meet and share their common goals. Judging from the
feedback received, I believe we have met the objective.
Finally, I welcome this opportunity to sincerely thank those
who participated in the symposium. It is hoped that the 2nd
International Symposium will be held in the not too distant
future.
George C. Woo
Chairman
International Symposium on Low Vision
Sponsors
Financial support for the symposium was provided in part by

the following industrial concerns and professional organiza-
tions:
Bausch and Lomb

Canadian Optical Supply Co. Ltd.
Canadian Association of Optometrists, Section of Low
Vision
Canadian Optometric Education Trust Fund
CmA Vision Care
College of Optometrists of Ontario
CooperVision Inc.
Corning Medical Optics
E.A. Baker Foundation, Canadian National Institute for
the Blind
Essilor Canada Ltd.
Faculty of Science, University of Waterloo
Imperial Optical Canada
K&W Optical Company Ltd.
Ontario Association of Optometrists
Optocoating and Co.
Optyl Design
Plastic Contact Lens Company (Canada) Ltd.
Superlite Optical Ltd.
Dr. G.C. Woo
Carl Zeiss Canada Ltd.
Contents
PART!
1. Measurement of Vision Loss: Theory and Practice
Color Plates following page 10.
Section A
New and Improved Contrast Sensitivity Approaches

to Low Vision •••••••••••••••••••••••••••••••• 1
R.F. Hess
The Evaluation of the Reading Capability of Low

Vision Patients Using the Vision Contrast Test
System (VCTS) •••••••••••••••••••••••••••••• 17
A.P. Ginsburg, B. Rosenthal, J. Cohen
Threshold and Suprathreshold Deficits in Color

Vision in Optic Neuritis • ••••••••••••••••••••••• 2.9
K.T. Mullen, G.T. Plant
Contrast Sensitivity • •••••••••••••••••••••••••• 45

R.A. Weale
Section B
The Effect of Blur Upon Psychophysical Receptive

Field Properties . • . . • . . . . • . . . . . . . . . • . . . . . . . . . 56
M.D. Benedetto, E.M. Gaynes, A.H. Gordon,
M.J. Mintz
Quantifying the Magnitude of Visual Impairment

with Multi-flash Campimetry • ••••••••••••••••••• 66
M. Dixon, E.M. Brussell
Spatial vs. Temporal Information in Suspected and

Confirmed Chronic Open Angle Glaucoma • •••••••••• 79
J. Faubert, E.M. Brussell, O. Overbury, A.G.
Balazsi, M. Dixon
Low Vision Management in Selected Eye Diseases ••••• 96

E.E. Faye
xii
The Role of X and Simple Cells in the Contrast

Transducer Function of Low Vision and Normal
Observers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
T.B. Lawton
Differential Retinal Structural Damage Exhibited

by Image Enhancement of Fundus Photographs • •••••• 125
S. Mitra, S. Whiteside, T. Krile
The Visual Requirements of Mobility • •.••••••••••• 134

D.G. Pelli
Visual Acuity Deficits and Chromatic Aberration in

Pseudophakia ••.••.•••.••••••....•......... 147
S.J. Rog, C.W. White, P. Simonet
Measurement of Central Fields Following Macular

Degeneration .•.•..•••.•••.••............•. 156
S.G. Whittaker, R.W. Cummings
Detection of Visual Field Defect Using Topographic

Evoked Potential in Children •••••••••••••••.••• 168
P .K.H. Wong, R. Bencivenga, J .E. Jan, K. Farrell
Preliminary Study of Topographic Visual Evoked

Potential Mapping in Children with Permanent
Cortical Visual Impairment •••••••••••••••••••• 180
P.K.H. Wong, K. Farrell, J.E. Jan, S. Whiting
2. Prescribing Low Vision Aids
Prescribing Magnification: Strategies for Improving

Accuracy and Consistency ••••••••••••••••••••• 190
I.L. Bailey
The Amorphic Fresnel Prism Trioptical System •••••• 209

R.L. Brilliant, S.D. Appel, R.J. Ruggiero
Sensorimotor Adaptation to Telescopic Spectacles • ••• 216

J.L. Derner, J. Goldberg, F.I. Porter, H.A. Jenkins
Most Useful Visual Aids for the Partially Sighted • •••• 232
G. Fonda
xiii
Paradoxical Cases of Visual Improvement Offered

by Above Average Lighting Levels in Cases of
Albinism and Retinitis Pigmentosa ••••••••••••••• 243
S. Patel
Magnification Efficiency in the Low Vision Patient ••• 249

F .1. Porter, J .L. Derner
An Overview on the Use of a Low Magnification

Telescope in Low Vision • •••••••••••••••••••••• 262
G.C. Woo
3. Reading in Low Vision
Observations from the Psychology of Reading

Relevant to Low Vision Research • ••••••••••••••• 272
J. Baldasare, G.R. Watson
Contrast Polarity Effects in Low Vision Reading • •••• 288

G.E. Legge, G.S. Rubin, M.M. Schleske
Effect of Magnification and Field of View on

Reading Speed Using a CCTV • •••••••••••••••••• 308
J.E. Lovie-Kitchin, G.C. Woo
Predicting Reading Performance in Low Vision

Observers with Age-Related Maculopathy (ARM) • •••• 323
G.S. Rubin
4. Low Vision Care
Section A
Visual Impairment and Disability: Enhancement and

Substitution ............................... 334
J .A. Couturier, J. Gresset
The Clinical Profile of a Young Visually

Handicapped Population • •••••••••••••••••••••• 350
J. Gresset, P. Simonet
Pathology Characteristics and Optical Correction of

900 Low Vision Patients • •••••••••••••••••••••• 362
A.R. Hill, A. Cameron
xiv
Assessment of Vision of Deaf-Blind Persons: A

Review •••••••..••.•.•••••••••.•...•..•••• 386
Lea Hyvarinen
An Evaluation of Follow Up Systems in Two Low

Vision Clinics in the United Kingdom ••••••••••••• 396
A.J. Jackson, J.H. Silver, D.B. Archer
Vision Examinations of Handicapped Children at the

Oregon State School for the Blind • ••••••••••••••• 418
S.K. Landis, T .0. Dutson, W. Ludlum
Accommodation in the Visually Impaired Child •••••• 4ZS

E. Lindstedt
Abnormal Arm Tone, Cigarette Smoking and Use of

Blood Pressure Medication in a Sight Enhancement
Clinic Population . •••••.••••.•..•.•••....•... 436
M.E. Paetkau
The City Study - Preliminary Findings • •••••••••••• 448

J.H. Silver
Section B
The Silver Pages: Are They Easier to Read? •••••••• 463

D.A. DeSylvia, T .R. Corwin
A Hierarchy of Perceptual Training in Low Vision • ••• 471

J. Faubert, O. Overbury, G.L. Goodrich
Low Vision Performance as a Function of Task

Characteristics • •••••••••••••••••••••••••••• 490
S. Marmion
PARTll
1. Some Issues in Rehabilitation in Low Vision
Section A
National Long Term Care Facility Survey •••••••••• 502.

S.J. Aston, M. Beliveau, A. Yeadon
xv
Rights of Low Vision Children and Their Parents • •••• 519

J.L. Hill
Life Satisfaction of Low Vision Patients and Other

Disability Groups: A Preliminary Study • ••••••••••• 536
M. Santangelo, O. Overbury, R. Lang
Section B
Sight Enhancement Services - A Safety Net or a

Spider's Web? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
E.J. Herie, G. Grace
Strengthening Low Vision Rehabilitation Through

the Accreditation Process • •••••••••••••••••••• 553
A.A. Rosenbloom
A Unique Model for a Resource/Rehabilitation

Centre for Consumers with Low Vision • ••••••••••• 5.67
B. Shalinsky, P. Shaw, B. Carroll
Contributors
Appel, Sarah D. Beliveau, Monica

William Feinbloom Vision William Feinbloom Vision
Rehabilitation Center Rehabilitation Centre
Pennsylvania College of Pennsylvania College of
Optometry Optometry
Philadelphia, Pennsylvania, Philadelphia, Pennsylvania,
USA USA
Archer, Desmond B. Bencivenga, Roberto

Queen's University of Department of Statistics
Belfast University of British
Eye and Ear Clinic Columbia
Royal Victoria Hospital Vancouver, British
Belfast, Northern Ireland, Columbia, Canada
United Kingdom
Benedetto, Marcus D.
Aston, Sheree J. Center for Ophthalmic
William Feinbloom Vision Research and
Rehabilitation Center Development
Pennsylvania College of Vision Rehabilitation
Optometry Institute
Philadephia, Pennsylvania, Department of
USA Ophthalmology
Sinai Hospital of Detroit
Bailey, Ian L. Detroit, Michigan, USA
School of Optometry
University of California Brilliant, Richard
Berkeley, California, USA William Feinbloom Vision
Rehabilitation Center
Balazsi, Gordon A. Pennsylvania College of
Low Vision Center Optometry
Department of Philadelphia, Pennsylvania,
Ophthalmology USA
McGill University
Montreal, Quebec, Canada Brussell, Edward
Department of Psychology
Baldasare, John Concordia University
William Feinbloom Vision Montreal, Quebec, Canada
Rehabilitation Centre
Pennsylvania College of Cameron, Alexander
Optometry Edinburgh Royal Infirmary
Philadelphia, Pennsylvania, Edinburgh, Scotland, United
USA Kingdom
xviii
Carroll, Bill Dutson, Terrel D.

Low Vision Association of Pacific University College
Ontario of Optometry
Toronto, Ontario, Canada Forest Grove, Oregon, USA
Cohen, Jay Farrell, Kevin

State University of New Department of Paediatrics
York University of British
College of Optometry Columbia
New York, New York, USA Department of Diagnostic
Neurophysiology
Corwin, Thomas British Columbia Children's
New England College of Hospital
Optometry Vancouver, British
Boston, Massachusetts, USA Columbia, Canada
Couturier, Julie-Anne Faubert, Jocelyn

Institut Nazareth et Louis- Low Vision Center
Braille Department of
Montreal, Quebec, Canada Opthalmology
McGill University
Cummings, Roger W. Montreal, Quebec, Canada
William Feinbloom Vision
Rehabilitation Center Faye, Eleanor E.
Pennsylvania College of New York Association for
Optometry the Blind
Philadelphia, Pennsylvania, New York, New York, USA
USA
Fonda, Gerald
Demer, Joseph L. Saint Barnabas Medical
Cullen Eye Institute Center
Baylor College of Medicine Livingston, New Jersey, USA
Houston, Texas, USA
Gaynes, Ernest M.
DeSylvia, Denise Center for Ophthalmic
New England College of Research and
Optometry Development
Boston, Massachusetts, USA Vision Rehabilitation
Institute
Dixon, Mike Department of
Department of Psychology Ophthalmology
Concordia University Sinai Hospital of Detroit
Montreal, Quebec, Canada Detroit, Michigan, USA
Ginsburg, Arthur P.
VISTECH Consultants
Dayton, Ohio, USA
xix
Goldberg, Jefim Hill, Adrian

Cullen Eye Institute The Eye Institute
Clayton Neurology Oxford, United Kingdom
Laboratory Nuffield Laboratory of
Baylor College of Medicine Ophthalmology
Houston, Texas, USA University of Oxford
Oxford, United Kingdom
Goodrich, Gregory
Western Blind Rehabilitation Hill, Jennifer Leigh
Center Society for Manitobans with
Veteran's Administration Disabilities
Medical Center Winnipeg, Manitoba, Canada
Palo Alto, California, USA
Hyvarinen, Lea
Gordon, Arnold H. Department of
Center for Ophthalmic Ophthalmology
Research and University of Oulu
Development Oulu, Finland
Vision Rehabilitation Department of
Institute Ophthalmology
Department of University of Tempere
Ophthalmology Tempere, Finland
Sinai Hospital of Detroit
Detroit, Michigan, USA Jackson, Jonathan
Royal Victoria Hospital
Grace, Gerrard Belfast, Northern Ireland,
Canadian National Institute United Kingdom
f or the Blind
Toronto, Ontario, Canada Jan, James E.
Department of Paediatrics
Gresset, Jacques University of British
Institut Nazareth et Louis- Columbia
Braille Department of Diagnostic
Montreal, Quebec, Canada Neurophysiology
Ecole d'Optometrie, British Columbia Children's
Universite de Montreal Hospital
Montreal, Quebec, Canada Vancouver, British
Columbia, Canada
Herie, J .A. Euclid
Canadian National Institute Jenkins, Herman A.
for the Blind Cullen Eye Institute
Toronto, Ontario, Canada Clayton Neurology
Laboratory
Hess, Robert Baylor College of Medicine
Physiological Laboratory Houston, Texas, USA
University of Cambridge
Cambridge, United Kingdom
xx
Lovie-Kitchin, J.E. Ludlam, William

Queensland Institute of Pacific University College
Technology of Optometry
Department of Optometry Forest Grove, Oregon, USA
Brisbane, Queensland,
Australia Marmion, Shelly
Rehabilitation Research and
Krile, Thomas Training Center:
Department of Electrical Blindness and Low Vision
Engineering/Computer Mississippi State University
Science Mississippi State,
Texas Tech University Mississippi, USA
Lubbock, Texas, USA
Mintz, Morris J.
Landis, Sandra K. Center for Ophthalmic
Pacific University College Research and
of Optometry Development
Forest Grove, Oregon, USA Vision Rehabilitation
Institute
Lang, Rina Department of
Low Vision Center Ophthalmology
Depart ment of Sinai Hospital of Detroit
Ophthalmology Detroit, Michigan, USA
McGill University
Montreal, Quebec, Canada Mitra, Sunanda
Department of Electrical
Lawton, Teri B. Engineering/Computer
Jet Propulsion Laboratory Science
California Institute of Texas Tech University
Technology Lubbock, Texas, USA
Pasadena, California, USA
Mullen, Kathy T.
Legge, Gordon Physiological Laboratory
Department of Psychology University of Cambridge
University of Minnesota Cambridge, United Kingdom
Minneapolis, Minnesota, USA
Overbury, Olga
Lindstedt, Eva Low Vision Center
University Eye Clinic, Department of
Karolinska Sjukhuset Ophthalmology
Stockholm, Sweden McGill University
"TRC", Tomteboda Resource Montreal, Quebec, Canada
Center for Visually
Handicapped Children
Solna, Sweden
xxi
Paetkau, Margaret Rubin, Gary S.

Departments of Medicine Wilmer Institute, Johns
and Ophthalmology Hopkins Hospital
University of Alberta Baltimore, Maryland, USA
Edmonton, Alberta, Canada
Ruggiero, Robert J.
Patel, Sudhir William Feinbloom Vision
Department of Ophthalmic Rehabilitation Center
Optics Pennsylvania College of
Glasgow College of Optometry
Technology Philadelphia, Pennsylvania,
Glasgow, Scotland, United USA
Kingdom
Santangelo, Mary
Pelli, Denis G. Low Vision Center
Institute for Sensory Department of
Research Ophthalmology
Syracuse University McGill University
Syracuse, New York, USA Montreal, Quebec, Canada
Plant, Gordon T. Schleske, Mary M.

Department of Neurology Department of Psychology
Addenbrook's Hospital University of Minnesota
Cambridge, United Kingdom Minneapolis, Minnesota, USA
Porter, Franklin L. Shalinsky, William

Cullen Eye Institute Low Vision Association of
Baylor College of Medicine Ontario
Houston, Texas, USA Toronto, Ontario, Canada
Rog, Stanislaw J. Shaw, Peter

Department of Psychology Optometric Institute of
Concordia University Toronto
Montreal, Quebec, Canada Toronto, Ontario, Canada
Rosenbloom, Alfred A. Silver, Janet

Chicago Lighthouse for the Moorfields Eye Hospital
Blind London, United Kingdom
Chicago, illinois, USA
Simonet, Pierre
Rosenthal, Bruce P. Department of Psychology
State University of New Concordia University
York Montreal, Quebec, Canada
College of Optometry Ecole d'Optometrie,
New York, New York, USA Universite de Montreal
Montreal, Quebec, Canada
xxii
Wa tson, Gale Wong, Peter K.H.

William Feinbloom Vision Department of Paediatrics
Rehabilitation Center University of British
Pennsylvania College of Columbia
Optometry Department of Diagnostic
Philadelphia, Pennsylvania, Neurophysiology
USA British Columbia Children's
Hospital
Weale, Robert A. Vancouver, British
Department of Visual Columbia, Canada
Science
Institute of Ophthalmology Woo, George C.
University of London Centre for Sight
London, United Kingdom Enhancement
School of Optometry
White, Charles W. University of Waterloo
Department of Psychology Waterloo, Ontario, Canada
Concordia University
Montreal, Quebec, Canada Yeadon, Anne
Whiteside, Steven Rehabilitation Center
Department of Electrical Pennsylvania College of
Engineering/Computer Optometry
Science Philadelphia, Pennsylvania,
Texas Tech University USA
Lubbock, Texas, USA
Whiting, S.
Division of Neurology
Department of Paediatrics
University of Ottawa
Children's Hospital of
Eastern Ontario
Ottawa, Ontario, Canada
Whittaker, Stephen G.
Rehabilitation Center
Pennsylvania College of
Optometry
Philadelphia, Pennsylvania,
USA
New and Improved Contrast Sensitivity
Approaches to Low Vision
Robert F. Hess
1. Introduction
The orlgms of contrast sensitivity approach can be traced

back to Selwyn, who worked for Kodak in the United King-
dom, and in the 1940's was involved in reconnaisance photog-
raphy. He was keen to measure the transfer function of cam-
eras and films and recognised the advantage of using
sinew ave patterns - that is, because they produced a known
image distribution. He used human subjects but only to cali-
brate his cameras and films (SELWYN [1]). In the late 1940's
and early 1950's, Otto Schade, a television engineer, carried
this approach forward (SCHADE [2]) and concentrated mainly
on the transfer properties of human vision, using initially
square wave gratings and an optical projection (onto a televi-
sion raster display) technique. In the 1960's, Fergus Campbell
and Gerald Westheimer working together in Cambridge were
interested in measuring optical quality and accommodation
and further developed the use of sinewave stimuli. John Rob-
son, who was then a graduate student built the first electron-
ic means of spatial sinew ave generation which had good time
and space resolution using a commercially available oscillo-
scope display. Apart from a refinement of the contrast lin-
earity which was also later developed by John, this system is
essentially what is used today. The first application of this
approach which was previously only used for optical assess-
ment to neural assessment was made by CAMPBELL and
ROBSON [3] in an effort to answer some of the questions
arising from the publication of De Palma and Lowry in 1962.
Shortly after this Gerald Westheimer was beginning to have
serious doubts about the interpretation of its application to
studying neural function (later expressed in WESTHEIMER,
[4]). It was now too late as new and interesting things about
the psychophysics (CAMBPELL and ROBSON [3],
BLAKEMORE and CAMPBELL [5]) and electrophysiology
(ENROTH-CUGELL and ROBSON [6]) of vision were emerg-
ing. The impact of the approach has been substantial for it
2
allowed the psychophysical and electrophysiological investi-

gation of vision to proceed together using comparable stimuli.
The higher expectation that the visual system might Fourier
analyse the retinal image with subsequent processing occur-
ring in the frequency domain has not fared so well. It is clear
that the whole image is not Fourier analysed, but it is still
unresolved whether this is a good way of thinking about the
analysis that occurs in different patches of the visual field
(ROBSON, [7]).
The important variables in the contrast sensitivity

approach are contrast and spatial frequency. Contrast is
defined as Lmax - Lmin/Lmax + Lmin where Lmax and Lmin
are the luminance maximum and minimum respectively. In
Fig. 1 we see an aerial view of the colleges in Cambridge
where naturally occurring periodic stimuli of the type used
for contrast sensitivity testing can be observed. A low spa-
tial frequency stimulus of high contrast is easily seen in the
grounds of Kings College (center left) and as one proceeds
along Kings Parade to Trinity Street, higher spatial frequen-
cies can be observed in the forecourt of Trinity College
(towards top right). The variable contrast is well illustrated in
Trinity College "backs" (center to top left). As Selwyn so
rightly pointed out, not only should the stimuli be periodic
such as those seen in Kings and Trinity, but they should also
be sinusoidal in their luminance profile, a feat that only the
gardeners of St. John's and Caius College can manage. Unfor-
tunately, these cannot be observed from this particular angle.
It is little wonder that this approach developed in Cambridge,
a place with such a richly textured environment.
Over this same period it was inevitable that such a suc-

cessful technique would also be applied to help our under-
standing of low vision. Since the dust has now settled over
this particular application, it is a good time to look back over
the last decade or so and ask whether the application of this
approach to low vision has been as successful as it has been
for investigating normal vision. My feeling is that it has been
of only very limited success, and I don't think that this is a
reflection on the quality of the science of those who have
used it. The technique as it stands is of very limited help to
ophthalmologists, optometrists and neurologists because in
most cases it does not tell them anything new which wOlild
help them to make earlier or more accurate clinical deci-
sions. In fact, in some cases, it can actually be misleading as
we shall see later. However, all is not lost because having
3
Figure 1
recognized the faults associated with this technique as it

applies to low vision assessment, we are obviously in a better
position to do something about it. There is emerging a second
generation contrast sensitivity approach, which I believe will
be essential to the vision practitioner. Inherent in this new
approach is that there is not just one type of contrast sensi-
tivity technique suitable to all clinical occasions, but many
different types of contrast sensitivity techniques, each spe-
cifically tailored to suit particular applications. It is the
indiscriminate use of this approach in the past that has under-
mined its clinical value.
What I would like to do in this paper is to take a number

of different clinical anomalies, each affecting different sites
in the retino-geniculo-cortical pathway. Firstly, I hope to
show just how inadequate our present method of indiscrimi-
nately applying contrast sensitivity measurements to these
conditions is, and secondly, how improved methods can be
produced to answer important clinical issues.
4
2. Optical Pathology
Let us begin with the optics. When should one remove a cata-
ract and how should one assess the low vision produced by a
cataract for occupational use? There are two important but
related clinical questions. The standard contrast sensitivity
approach is a step in the right direction. Some years ago, I
had the good fortune to work with George Woo when he was
on sabbatical in Australia (HESS and WOO [8]). We investi-
gated a number of patients with cataract and found that con-
trast sensitivity was more valuable than visual acuity as a
predictor of performance. This was because for some of
them the scattering affected even very low spatial frequen-
cies. These patients found it much more difficult in everyday
situations to make visual decisions. However, the overall
impression that we get of their visual handicap is still inaccu-
rate because the testing conditions that we use in the clinic
do not match everyday conditions with which these patients
have to deal. Specifically, contrast sensitivity measurements
are usually made either without a background field (i.e., in a
darkened room), or with a background field of the same space
averaged luminance. Compare this with everyday conditions
where the local luminances surrounding objects of interest
can be very high, for example, driving at night into oncoming
headlights, or into the sunset. It is in these everyday condi-
tions that patients with media abnormalities are most disad-
vantaged because of light scatter and yet we don't assess this
situation at all with the present contrast sensitivity methods.
Consider the modified experimental arrangement shown

in Fig. 2. Here we are measuring contrast sensitivity in the
presence of an evenly lit and luminance matched surround
field through an ND filter, placed in one of two positions 1
and 2. Since NDI attenuates the light from both the stimulus
field and the background, and ND2 only attenuates the stimu-
lus luminance, the effects from scattered light will be rela-
tively greater for the ND2 filter. This allows us to assess
contrast sensitivity for backgrounds of relatively higher and
higher luminance as the density of ND2 filter is increased.
The reason for using the NDI filter is to control for any neur-
al changes that might result from reducing the stimulus lumi-
nance. Fortunately, the visual system exhibits a Weber
response for a wide range of stimulus luminance within the
photopic region, the extent of this depends only on the spatio-
temporal frequency of the stimulus (VAN NES and BOUMAN
5
[9]). It can be shown that from a knowledge of the ratio of

contrast thresholds for an ND filter placed in these two posi-
tions, the percentage of forward light scattered in the eye
can be calculated (Fig. 2). Now, with the use of one filter in
two different positions, contrast sensitivity measurements
can tell us directly about the amount of forward scattered
light and thereby we can obtain a direct prediction of the vis-
ual performance of these patients under more realistic test-
ing conditions. Some recent measurements that George Woo,
John Robson and I have made suggest that in normal eyes
about 10 - 15% of the light is scattered (Fig. 2, lower graph -
unfilled symbols), and that contrary to current opinion, in
patients without media opacities this does not vary greatly
with age. Patients with media opacities exhibit markedly
greater forward light scatter, which can be assessed directly
using this modified contrast sensitivity approach (HESS, et
al., forthcoming). Results (filled symbols) from some such
patients with incipient cataract are seen in Fig. 2. In this
figure the ratio of contrast thresholds for the two filter posi-
tions is plotted against the density of the filter. The solid
curves give the theoretical relationship for different degrees
of scattering.
6
1.
k. ND U•• + ITO) NO .Tranlml •• ion of filter

lro.Stlmulul mean luminance
ND ITO+I .. la.-Scattered light
100'110
50'110
k
.1
l~~~,,--r---ITTrrrT-r-'---'
1 .1 .01
NO TRANSM ISSION
Figure 2
3. Retinal Pathology
Let us now move on past the optics to a related problem

involving retinal pathology, and in particular senile macular
degeneration. If contrast sensitivity as measured in the tra-
ditional way is depressed in an elderly patient with media
opacities, how can one be sure the retina is normal? Since the
cloudy optics severely impair any ophthalmoscopic assess-
ment other means must be sought. Two methods have been
developed to help with this important, though difficult, clini-
cal decision. Firstly, Dan Green from Michigan suggested
that laser interferometry be used to bypass the optical ano-
7
maly and thereby measure neural contrast sensitivity directly

- this is particularly useful when the opacities are discrete
and when there are clear windows through which the laser
light can be channelled. It is of less value, except in a very
qualitative sense, when the opacities are diffuse because light
scattering in this case also affects the neural contrast meas-
urements (GREEN, [10]). The second method was devised by
Rick Williams and coworkers at Berkeley (WILLIAMS et aZ.
[11]) and uses the measurement of vernier alignment of large
spots which are less blurred by media opacities. This is also a
very successful method, relying as it does on the assumption
that retinal pathology produces anomalous vernier alignment
for such low contrast, low pass filtered stimuli. I recently
had the opportunity and good fortune to collaborate with Dan
Kersten (from Brown) in Cambridge, where we developed
another method to help with this difficult question in teasing
apart the contribution of optical and retinal pathology. In
this approach we add "visual noise" to the grating stimulus for
which contrast sensitivity measurements are made. An
example of what such a stimulus looks like for different lev-
els of visual noise (wideband noise of varying spectral density)
is shown in Fig. 3, A-C. The usefulness of this approach rests
on the following finding. Any optical anomaly will attenuate
the grating signal as much as the visual noise signal so that
the signal-to-noise ratio will remain unaltered. In other
words, if the loss of contrast sensitivity is due solely to opti-
cal factors, the addition of visual noise will not further
depress contrast sensitivity. This is not the case if the loss of
contrast sensitivity is due to retinal pathology such as senile
macular degeneration or optic neuropathies. In these cases
we found that the noise was not attenuated to the same
extent as the signal and so the signal-to-noise ratio changes.
This means that by measuring contrast sensitivity with and
without visual noise we can address the more important clini-
cal question concerning whether the retina, which cannot be
adequately observed because of the cloudy media, is normal.
The results of adding visual noise are shown in Fig. 4. In each
part of Fig. 4 we are comparing contrast sensitivity functions
with (filled symbols), and without (open symbols) visual noise
of high spectral density. In the top figure contrast sensitivity
has been measured through a diffuser to simulate an even
optical opacity. The function is severely depressed (maxi-
mum sensitivity around 30 and cut-off acuity around 1 c/deg).
Notice that when visual noise is added, contrast sensitivity is
not affected. Compare this with the results in the lower fig-
ure which are similarly depressed due to retinal pathology.
8
Now the addition of visual noise of the same spectral density

further depresses contrast sensitivity. Here we have a meth-
od which will allow the separation of neural and optical ano-
malies. Unlike the other methods so far proposed, this one
allows not only an estimate of how good the neural apparatus
is, but it also allows one to estimate relative neural and opti-
cal contributions to depressed contrast sensitivity (KERSTEN
et aZ., forthcoming).
Figure 3
9
DIFFUSER
100 o Affected eye
without noise
>- • Affecled eye
t:
> with noise
;::
"'-Zl"l
in
z
UJ
(f)
I-
(f)
10 - I1-l"
«
a:
I-
z
0
(J
\
0.06 0.12 0.25 0.5 1.0

SPATIAL FREQUENCY (c/d)
MACULAR DEGENERATION
100
>-
I-
:;
;::
-I---n""
in
z
UJ
(f) 10
I-
I'\\ \
(f)
«
a:
I-
z
0
-T--T
(J
~T
0.06 0.12 0.25 0.51 1.0
SPATIAL FREQUENCY (c/d)
Figure 4
4. Optic Nerve Pathology
Let us now progress to retrobulbar pathology. How successful

has the traditional measurement of contrast sensitivity been
at assessing the nature of the pathology underlying optic
nerve demyelination? The traditional approach of using large
field grating stimuli has shown us that there are four differ-
ent types of spatial contrast sensitivity anomaly in optic neu-
ritis, which are depicted in the left half of Fig. 5 (REGAN et
ale [12]; for review see HESS and PLANT [13]). Some recent
work that I have been fortunate to do with Gordon Plant
(PLANT and HESS [14]), suggests that all of these deficits can
be successfully modelled by simulating different types of
10
regional loss of sensitivity (scotoma) in normal observers.

The results of these simulations are seen in the right half of
Fig. 5 with the type of regional loss simulated shown above
each contrast sensitivity curve. Compare the scotoma simu-
lation on normal subjects (right half of Fig. 5) with the actual
results on representatives of the four different types of visual
loss found in optic neuritis (left half of Fig. 5). This close
correspondence suggests one explanation for the above results
in optic nerve pathology, namely that there may be discrete
regional loss of sensitivity in these patients. If this is so, and
there is other evidence for it from how the plaques of demye-
lination are distributed in the optic nerve itself and from per-
imetry, then the use of a large field of stripes to test such a
patchily distributed anomaly is most inappropriate. Clearly,
our stimulus should be small enough and sufficiently well
positioned to probe the nature of the local regions of dysfunc-
tion. Indeed, if contrast measurements are done with a small
patch of grating, the patchy nature of the visual loss in this
condition reveals itself. This is seen in Fig. 6. Here contrast
sensitivity has been measured with a well-localized (at least
equally localized in space and spatial frequency)
2-dimensional Gaussian patch of grating for different retinal
regions at the same eccentricity (3.5 and 7.5 degrees). The
true nature of the pathology is seen in the lower figures
which are contrast sensitivity functions measured at these
locations. Contrast sensitivity is affected differently in the
same parts of the visual field at the same eccentricity -
hence the anomaly has a patchy distribution. In future, the
contrast sensitivity measurements in cases of optic nerve
demyelination will have to be done with sufficiently small
patches of grating so that the regional variations in the
pathology can be adequately examined.
Plate I: Topographic map of control group, showing the
appearance, growth and decay of the occipital
DEF component; 90 ms post stroboscopic stimu-
lus. The color scale is at the right proportional
to the amplitude. The map is orientated as a
vertex view, patient's right is to the right, ante-
rior is up.
Plate II: Topographic map of control group showing the
DEF component; 100 ms post stroboscopic stim-
ulus.
Plate III: Topographic map of control group, showing the
ulus.
Plate IV: Topographic map of control group, showing the
ulus.
Plate V: Integrated map of the control group. Note the
color scale now has no polarity, as it reflects
area or absolute amplitude.
Plate VI: Integrated map of patient with type I complete
abnormality (see text).
Plate VII: Integrated map of patient with type 2 abnor-
mality (see text).
Plate II
Plate ill
"S
198. ee
FILE:A1868618 RECORD:I
EPOC~
512
SCRLE BRNK
2 1
~O\ out uV
1,
•
A LEF~ •• _-,-~_.~ '11'1 ~

;
.
~
\
\
,•
:..... , ',' .,'
,
Plate IV (Top) Plate V (Bottom)

Plate VI (Top) Plate VII (Bottom)
11
§
A
...,- 0'
B
..-..--..... - 1,·0
_e",-'--'-- ., 0-°
/ 0\
1000
..."" \
q A B
/.°b 100
!.-.-.,\ li~\
I \\\q
~ • 0
100
,re,\'q
~ 1/ e'l 100
I \\ /
\\,
,0 / .0
.
:J:
IIe~
I
\\ ;"
10
~
e 0
'0
,
«
~\ e, .0
~ '0
,
60 01 10 60
I I
I
I
I , ''
"
I
'"
SPATIAL FREOUE N(V (c I dpg ) 01 10 01 10 60
dB e
0
............ ...
-
'111,1,0:,0
!.!)~~ ~'v'- 0' eo.. C
,.
g~~ 0'
•• _..._e
1000
°
0
/~
~w
~
100
/o-q 100 ,r'" 100
J '-'\g
~ /;~/.\ 0, 0 /0 '\
i
~
/.-.....°\
\
• • 0, 6
I°
'/
:J: .. °0
1/\0 6
Ii
I
10 10 /e 10
~ ~. \~ fie ~
§ I
' ~'" I
~, I
I
01
SPATIAL
60 01
FREQUENCY (c/de>9)
10 60 01 10
'" O'
SPATIAL . FREQUENCY (c Ideg)
10 60
Figure 5
12
OPTIC NEURITIS
100 c.s. c.s.
30
10
0.3 10 30 0.3 10 30
SPATIAL FREQUENCY (Cycl •• /deg)
Figure 6
5. Cortical Pathology
Let us finally go to the cortex. The use of contrast sensitivi-

ty techniques in assessing the visual loss in amblyopia, partic-
ularly strabismic amblyopia, has been particularly unsuccess-
ful. While it is true that contrast sensitivity is depressed in
most cases, the level of disability produced by this is very
small for two reasons:
1. The contrast sensitivity effects are small.
2. Although more contrast is needed to detect these

stimuli, that does not mean that they are necessarily
seen at a reduced contrast above threshold. In fact,
when one assesses this in strabismic amblyopia one
13
finds that above the raised threshold contrast is seen

normally (HESS and BRADLEY [15]).
The problem concerning strabismic amblyopia cannot be

uncovered by using the contrast sensitivity approach, for
these amblyopes see things as distorted, and the contrast sen-
sitivity approach does not measure distortion. The best
example of this is in the next figure (Fig. 7). Here we have a
strabismic amblyope with reduced Snellen single letter acuity
but normal contrast sensitivity. We also see how the grating
stimuli are perceived (these are drawings made by the
amblyope of how the stimuli appeared through the amblyopic
eye), they are distorted and the distortions vary with spatial
frequency. The problem here is that there is scrambling in
the cortical map for which traditional contrast sensitivity
measurements have no answer.
300 1 C/d 4 C/d 10 c/d 15 c/d
>- ~~30Cd/m2
I. \\
I-
:> 100 IG.
i=
o. VERTICAL
C.
V)
z
w
I/)
HORIZONTAL
l-
I/)
«
a::
·10
I-
z
0
() o
\
\
\
0·1 1 10 100
SPATIAL FREQUENCY (c/cleg)
Figure 7
14
In conclusion, there are a number of inadequacies with

the present contrast sensitivity technique as applied to
assessing low vision resulting from visual pathology. They
stem from the fact that to be useful the contrast sensitivity
approach needs to be adapted to particular clinical questions
which are worth answering. If it turns out to become just
another way of saying vision is abnormal it will be of little
real clinical use.
Some of its problems stem from the fact that
1. Its testing condition can be unrepresentative of real,

everyday conditions - for example, in cases of scat-
tering,
Z. Reduced contrast sensitivity does not have any dif-

ferential diagnostic value in itself, e.g., assessing the
retina behind cloudy optics - "is the loss optical,
neural or both?"
3. The use of large field gratings for contrast sensitivity

measurements is misleading when the pathology is
not evenly distributed over the tested area, e.g.,
optic nerve pathology,
4. When the abnormality does not involve a primary loss

of contrast sensitivity, contrast sensitivity measure-
ment tells us nothing about the pathology, e.g., stra-
bismic and anisometropic amblyopia are due to spa-
tial scrambling not a primary loss of contrast
sensitivity.
By realizing these inadequacies a series of new and

improved contrast sensitivity approaches can be developed
that have great promise for clinical practice.
15
6. References
1. E.W.H. Selwyn: The photographic and visual resolving

power of lenses. Photogr. J. 88B, 6 (1948)
2. O.H. Schade: Electro-optical characteristics of televi-

sion systems. RCA Rev. 9, 5 (1948)
3. F.W. Campbell, J.G. Robson: Application of Fourier

analysis to the visibility of gratings. J. Physiol.
(Lond.) 197, 551 (1968)
4. G. Westheimer: Fourier analysis of vision. Invest.

Ophthalmol. Vis. Sci. 12, 86 (1973)
5. C. Blakemore, F.W. Campbell: On the existence of

neurones in the human visual system selectively to the
orientation and size of retinal images. J. Physiol.
(Lond.) 203, 237 (1969)
6. C. Enroth-Cugell, J .G. Robson: The contrast sensitivi-

ty of retinal ganglion cells of the cat. J. Physiol.
(Land.) 187, 517 (1966)
7. J.G. Robson: Neural images: The physiological basis of

spatial vision. In Visual Coding and Adaptability, ed.
by C.S. Harris (Laurence Erlbaum, Hillside 1980) pp.
177-214.
8. R. Hess, G. Woo: Vision through cataracts. Invest.

9. F.L. Van Nes, M.A. Bouman: Spatial modulation trans-

fer in the human eye. J. Opt. Soc. Am. 57, 401 (1967)
10. D.G. Green: Testing the vision of cataract patients by

means of laser generated interference fringes. Sci-
ence 168, 1240 (1970)
11. R.A. Williams, J .M. Enoch, E.A. Essock: The resis-

tance of selected hyperacuity configurations to retinal
image degradation. Invest. Ophthalmol. Vis. Sci. 25,
389 (1984)
16
12. D. Regan, R. Silver, T.J. Murray: Visual acuity and

contrast sensitivity in multiple sclerosis: Hidden visual
loss. Brain 100, 563 (1977)
13. R.F. Hess, G. T. Plant: The psychophysical loss in optic

neuritis: Spatial and temporal aspects. In Optic Neu-
ritis, ed. by R.F. Hess, G.T. Plant (Cambridge Univer-
sity Press, 1986) p. 106.
14. G.T. Plant, R.F. Hess: Regional threshold contrast sen-

sitivity within the central visual field in optic neuritis.
Brain (in press)
15. R.F. Hess, A. Bradley: Contrast perception above

threshold is only minimally impaired in human amblyo-
pia. Nature 287, 463 (1980)
The Evaluation of the Reading Capability
of Low Vision Patients Using the Vision
Contrast Test System (VCTS)
Arthur P. Ginsburg, Bruce Rosenthal, Jay Cohen
1. Introduction
Visual acuity, the main measure of visual capability for the

last 124 years, has not related well to functional vision. In
particular, low vision practitioners have noted the discrepan-
cy between the "quantity" of vision and patient performance
as measured by acuity [1]. Contrast sensitivity is emerging as
a more complete performance-related measure of the func-
tional "quality" of vision. Differences in individual contrast
sensitivity, but not visual acuity, of normal observers have
been shown to be related to differences in complex visual
tasks [2]. These results from normal observers suggest that
contrast sensitivity may also help in the evaluation of func-
tional vision of low vision patients.
A major drawback to incorporating contrast sensitivity

into low vision clinical practice is that most research has
been done using computer-video systems. These systems are
limited in practicality for routine clinical use because of
expense, the careful calibration and monitoring of contrast
required, test time, and variability [3,4]. An early photo-
graphic plate system, the Arden gratings, that solved some of
those problems nevertheless has limitations of test reliability,
frequency range, and scoring technique [3,4]. A new scientif-
ically based vision contrast test system (VCTS) overcomes the
limitations of the previous test systems to provide a quick,
simple, repeatable, standardized chart system for measuring
contrast sensitivity that is comparable to computer-video
systems [3,4].
In this study we report the use of the VCTS to measure

the contrast sensitivity of 55 low vision patients having a
wide variety of eye diseases. This initial study was designed
to determine if contrast sensitivity as measured by the VCTS
18
could provide more information than visual acuity in predict-

ing patient performance using aids for a continuous reading
task. A secondary goal was to determine if contrast sensitivi-
ty could predict the preferred eye of the low vision patient
more accurately than acuity, an important part of helping to
treat the low vision patient. Comparison of the contrast sen-
sitivity of the low vision population to that of a normal popu-
lation was an additional goal.
2. Methods
The 55 low vision patients in this study were consecutively

chosen patients of the Lighthouse of the Blind of New York
and the Low Vision Clinic at the State University of New
York. Average patient age was 63.4 +/- 20.3 years. The
major eye diseases of these patients are shown in Table 1,
with 40% having atrophic macular degeneration.
Visual acuity was measured at 3 m (10 feet) under stan-

dard luminance with Lighthouse distance acuity charts. If the
patient had difficulty, visual acuity was measured at 1.5 m (5
feet), or as close as necessary.
Contrast sensitivity was measured using the Vision Con-

trast Test System (VCTS 6500, Vistech Consultants, Inc.)
shown in Fig. 1. The VCTS is a chart system that displays a
series of sine-wave gratings of different spatial frequencies,
contrasts and tilts. At the standard test distance of three
meters, five spatial frequencies are tested: 1.5, 3, 6, 12, and
18 cycles per degree (c/deg). The contrast of the gratings
decreases from left to right in approximately 0.2 log unit
steps from a high of 33% to zero contrast. The gratings are
tilted +/- 15 degrees or are vertical to create a forced-choice
response. The patient is asked to report the orientation of
each grating patch viewed from left to right. The last correct
response for each row indicates the contrast thresholds for
that spatial frequency. That point is marked on the data
form. Connecting all five points creates a contrast sensitivi-
ty curve. Illuminance of the VCTS is standardized by using a
calibrated light meter included with the system, a feature
allowing a comparison of these new data with previously col-
lected normative data.
19
Table 1
Low Vision Patient Data
·· ........
n" "" ,,,
Dh·~•• ... 1 t,.
AUdl,
WI'.
.. ltv
.... r.,.......,.. ..
(19 hAM'! .,t
,.,.,'"''" """"U
,.&~
!XI ".,'1 1
w
• Ot"
Irk" H
6D
• • k c;:;a
AN"
tI "'y •
i ... 1
.... 1 'n • c:Mii
Si IIHL .... 1 H
•...,i
' "
in
ill _ _ II ;.; H H ..
It}; i..-I H.
til H • a
" mil •• 'tHon k GOiII
w. ·
H [..,.1 '""
• in • I."
B .. tn.... "'"
'01" • in
Ie"
II OW Ii in R JO llil II
n ItIP • ft. .. III~ • is '"
I ll..... ;::
.... i i " '......
" 1:l1L
lOW!
H
_iH __
tn Ii ....
nl0R.t.... , 1m........
lit£;; .....
ram
" DiM
· · ....
Ii K ....
Ii 1:l1to h
H
'"
eft M ....
"
H \Ill;
(... 1
• • Tn • w
• fn
'''II s ... • tiMii
'"
H ....
..........
Ii U "iN.
• In
• 1" " ...."" :~: II Ii in 6\ CMiiI
Il'a
· ·· .....
H W
" mat • • •
::
D H M in h ....
I"""'.,..". . III.
i ll_ ...
u R II .. II
IIUIIF ....
Ii • in
II ill,. 61 lit....
• 1n
,"
ItlI.
" lIS!'.
.
• · ...
II 1118 it R • .. , .... " it
",yrm •
"'f...ir..
.1
• In • I.e;
fi tt. H W
it QUI H
I""
Wlt ,... '
Il
ilion"," H i"
'n n ....
u
"Is J'Nihi. mil I ..,
• in
Iii. tiM
• 1n • t410I
h ltitte h H it. • Gil
II H i t! ,
· ....
R '"
lit'.. ,.. ::: ."'1 H 1n
----ww"lt~ar. •.__.•._--'rnn--,gH--~~._ i.-I [hllotf '"'"
tJ :::~.. 1D • i"
II
From previous studies [5-7], it was expected that low

vision patients would have limited contrast sensitivity at the
standard three meter distance so the patients were also test-
ed at a distance of one meter. This closer viewing distance
changes the spatial frequencies to 0.5, 1, 2, 4 and 6 c/deg and
increases the visual field tested by the test patches from 1.5
to 4.5 degrees. Since the previous normal population data
from the VeTS was made at a 3 m (10 foot) viewing distance,
appropriate norms for this closer test distance were estab-
lished at one meter. A population of 47 normally-sighted sub-
jects was used, each individual with a visual acuity of at least
6/6 (20/20) in both eyes. All subjects were free of significant
media opacities and eye diseases. The normal observers were
age and sex matched to the low vision patients.
Two measures of the contrast sensitivity function appro-

priate to visual performance have been the magnitude of con-
20
._..,..1 __ _
W ..__
Figure 1. The Vistech Vision Contrast Test System

(VCTS), a photographic plate system used to
evaluate contrast sensitivity. Patches of sine-
wave gratings with different contrasts and spa-
tial frequencies are the visual targets used for
testing. The patient is instructed to identify
the orientation of the lines within the patches.
trast sensitivity and spatial frequency bandwidth relevant to

the particular task [8]. For example, the identification of
Snellen or Sloan letters on visual acuity charts required a cer-
tain magnitude of contrast sensitivity over a one-to-two
octave bandwidth [8]. The particular levels of contrast sensi-
tivity and range of spatial frequencies depend upon each let-
ter and letter size. The contrast sensitivity data of the
patients who performed well with the reading task will pro-
vide the relevant unaided contrast sensitivity magnitude and
spatial frequency bandwidth required for reading the Sloan I
m Continuous Text Reading Cards.
The preferrea eye was determined to be the eye the

patient used for visual tasks and while using the visual aid for
the reading task in this study. If either eye was used or no
SUbjective preference was shown, the patient was rated as
either.
21
The appropriate low vision aid used was determined by a

standard low vision examination. Initial addition was deter-
mined from Kestenbaum's rule using a near acuity card.
Reading ability of the low vision patient was determined
using the 1 m Sloan Continuous Text Reading Cards while
using the low vision aid judged to be most appropriate. A
four level rating scale was employed to rate the patient's
reading ability fluency: good, fair, poor and very poor. A
good score was given if the patient read the cards fluently
while a fair score was given if the patient could read the
cards but not fluently. A patient not able to read the cards
but able to read some isolated words or a continuous line of
letter acuity without skipping letters was rated poor; a
patient unable to read even isolated words or letter acuity
without skipping letters was rated very poor.
3. Results and Discussion
The results of this study are shown in Table 1. Tabulated

data are the patient diagnosis, patient age, visual acuity, the
eye having best visual acuity, the eye having best contrast
sensitivity, whether or not the eye with the best contrast sen-
sitivity had a two-octave range of spatial frequencies, the
preferred eye, and continuous reading rating with aids.
Visual acuities of these patients range from 3/3 (10/10) to

hand movement with an average of 3/115 (10/317) to 3/222
(10/740). Of these patients, 46 of 55 (83%) had one eye with
higher acuity than the other.
From contrast sensitivity measurements of the low vision

patients, 48 of 55 (85%) had higher contrast sensitivity over
more spatial frequencies in one eye when compared to the
fellow eye. The means and 90% of the population (based on
the 5th and 95th percentile) of the contrast sensitivity of the
normal population and low vision patients for the more sensi-
tive eye are shown in Fig. 2. At the one meter viewing dis-
tance, the average contrast sensitivity of the low vision
patient is at least a factor of four less than that of the nor-
mal population at all spatial frequencies. This finding is in
agreement with other contrast sensitivity data obtained from
low vision patients [5-7].
22
CONTRAST SENSITIVITY FUNCTIONS

OF NORMAL & LOW VISION POPULATIONS
AT ONE METER VIEWING DISTANCE
1..0_ VISION ~PuLATION
NO"MAL PO,"ULATION LOW
__ VIIIO'"
MIAN I'O~LATION
IIIlf;PONOIJrfQ TO AIDS
1111 . . . MUN
. . . . . . .UN
IIIIII1 =~~'T",.
o
...;;:>
..J
o%
(/)
E UJ
a:
......
(/) 100 01 %
Z
UJ
...
(/)
«
(/)
(/)
« ...a:
...a: ~
~ JO 03 U
U
SPATIAL FREQUENCY
ICYCLES PER DEGREe)
Figure 2. Shown here are the means and 90% of the pop-
ulation (based on the 5th and 95th percentile) of
the contrast sensitivity functions of the normal
population, low vision patients for the most sen-
sitive eye, and those low vision patients with
fair to good fluency reading Sloan 1 m Continu-
ous Reading Text Cards using visual aids. At
this one meter viewing distance, the average
contrast sensitivity of the low vision patient is
at least a factor of four less than that of the
normal population at all spatial frequencies.
The mean and 90% of the population of the low VISIon

patients having aids who could read the Sloan 1 m Continuous
Text Reading Cards with fair to good fluency are shown in
Fig. 2. The average contrast sensitivity and spatial frequency
bandwidth of these patients are less than that of the normal
population but greater than that of the general low vision
population. These data may be useful as a screening tool to
help guide the low vision practitioner in predicting patient
reading capability with aids.
23
Most low vision patients performed the reading task bet-

ter with one eye than the fellow eye. Of 55 patients, 47
(86%) showed either equal or more consistent fluency in the
reading task in the eye having better sensitivity and/or visual
acuity. For only one patient did the preferred eye have poor-
er visual acuity and contrast sensitivity. Of the eight patients
with better visual acuity in one eye and better contrast sensi-
tivity in the fellow eye, six patients (75%) preferred to use
the eye with better contrast sensitivity. Thus contrast sensi-
tivity magnitude seems to be more important than visual acu-
ity in determining eye preference [5-7].
In responding to aids, the reading of 1 m Sloan Continuous

Text Reading Cards by 36 of 55 (66%) of the patients was rat-
ed good, 6 of 55 (11%) patients were rated fair, 10 of 55
(19%) were rated poor, and 3 of 55 (6%) were rated very poor.
The number of low vision patients having some contrast sensi-
tivity over at least a two-octave bandwidth of spatial fre-
quencies (0.5, 1 and 2 c/deg) was 42 of 55 (77%). Of those
with two or more octaves, 39 of 42 (93%) were noted as hav-
ing fair to good reading ability, capable of moderate to highly
fluent reading of the 1 m Sloan Continuous Text Cards using
aids. Ten of thirteen patients with a spatial frequency band-
width of less than two octaves, showed poor reading ability
and three had good reading ability. Thus the criteria of
patients having a two-octave spatial frequency bandwidth
identified 93% (39 of 42) of patients who were able to read
well with visual aids and 77% (10 of 13) of those who did not
perform well with aids. On the other hand, visual acuity pre-
dicted 96% (53 of 55) of patients' additions within the range
of optical magnification that should have allowed the patient
to read, but only 76% (42 of 55) were actually able to read
continuous text. Only 15% (2 of 13) of the patients were
identified who could not read continuous text. Visual acuity
predicted 11 false positives whereas contrast sensitivity pre-
dicted only 3 false positives. Therefore, contrast sensitivity
gave a more accurate measure of reading performance capa-
bility than did visual acuity.
It is interesting to note that two of three patients who

showed good reading ability but had less than a two-octave
bandwidth required two times higher magnification for their
visual aids than was expected using Kestenbaum's rule (the
predicted additions were 4X and 5X magnification; the
required additions were 8X and lOX, respectively). Of the
other patients having good reading ability, none required
more than IX (4 D) above their predicted additions.
24
The previous results centered on averaged data. An

examination of individual contrast sensitivity functions pro-
vides important information about the low vision patient, too.
For example, although each patient shown in Fig. 3 has 3/42
(10/140) acuity, considerable differences exist in their con-
trast sensitivity functions, especially between patients MB
and RK. Patient MB, having higher contrast sensitivity at all
but 4 c/deg, did well on the reading task whereas patient RK
did poorly. Clearly, patient RK will not see large, low-
contrast objects to the same degree of patient MB.
. Another measure that could be used to test reading per-

formance of low vision patients is the peak of the contrast
sensitivity function. One problem with obtaining the peak
value is evident from the data. For example, patients RK and
MB have different peaks at different spatial frequencies (Fig.
3). One way to circumvent that problem is to use large tar-
gets having sharp edges such as low frequency square-wave
gratings or bisected disks. These targets would sum the visu-
al receptive field (channel) responses over the whole contrast
sensitivity function to create one number. However there,
too, very important patient information can be lost. Consid-
er, for example, the very different contrast sensitivity func-
tions of patients JT, HM and SM of Fig. 4. A summed grating
score (SGS) using a simple channel summation model and a 0.5
c/deg square-wave (multiply contrast sensitivity values by
square-wave harmonics and addition) yields 16.5, 14.13, and
9.27 for patients SM, HM and JT, respectively. Note the lack
of relationship between acuity and the summed grating score.
Patient JT had the highest acuity 3/37.5 (10/125) but also had
the lowest summed grating score, 9.27. Although SM has the
highest summed grating score (16.5), his contrast sensitivity
is very poor at spatial frequencies greater than 1 c/deg. The
lowest summed grating score of JT reflects the lack of meas-
urable contrast sensitivity at 0.5 c/deg. Clearly, these two
patients have quite different visual capabilities. In this case,
the summed grating score does predict that patient SM would
perform the reading task better than patient JT. However,
the summed grating score does not predict that patient HM
would do worse than patient JT. The higher contrast sensitiv-
ity of patient HM than that of patient JT at 0.5 c/deg does
support their differences in reading performance. Further,
note that the poor reading capability of RK (Fig. 3) compared
to the fair reading capability of JT is also not predicted by
their summed grating scores of 15.8 vs. 9.27 respectively.
However, their difference in reading performance is predict-
ed from their contrast sensitivity scores at 0.5 c/deg. It
25
appears from initial analysis that the patients' contrast sensi-

tivities at 0.5 and 1.0 c/deg and not at the peak or from a
summed grating score is the best predictor of the continuous
reading task.
EXAMPLES OF CONTRAST SENSITIVITY OF

MACULAR DEGENERATION PATIENTS
VA SGS RTA
MB 10/140 55.8 GOOD
. . . . . . . . . . IU . . HS 10/140 28.13 GOOD
RK 10/140 15.8 POOR
3 0 0 t - - - - - - - - - - - - - - - - f 003
01
i
II:
j!:
.03 UI
10
..
SPATIAL FREQUENCY
(CVCLES PER DEGREE)
Figure 3. Individual contrast sensitivity functions of

three low vision patients. Note that while each
patient shown in Fig. 3 has 3/42 (10/140) acuity,
there are significant differences in their con-
trast sensitivity functions, especially between
patients MB and RK. Also note that while
patient MB, with an overall higher contrast sen-
sitivity function, did well on the reading task,
patient RK did poorly. Having higher contrast
sensitivity, patient HS also performed better on
the reading task than patient RK.
26
EXAMPLES OF CONTRAST SENSITIVITY OF

MACULAR DEGENERATION PATIENTS
VA SGS RTA
............. HM 10/200 14.13 POOR
JT 10/125 9.27 'FAIR
SM 101180 18.50 GOOD
3 0 0 t - - - - - - - - - - - - - -__~ 003
...
Q
~
cr
~
I
0'
03 (,)
'0 .. ,
.3
.. e
SPATIAL FREQUENCY
(C'lCLES PER DEGREE)
Figure 4. Shown here are the diverse contrast sensitivity

functions of patients JT, HM and SM. The
summed grating scores (SGS) are 16.5, 14.13 and
9.27 for patients SM, HM and JT respectively.
Note the lack of relationship between acuity
and the summed grating score. Also note that
the higher contrast sensitivity at 0.5 c/deg of
patient HM than patient JT supports their read-
ing performance differences, a result not pre-
dicted from the SGS value.
27
4. Conclusions
Contrast sensitivity measurements from the VCTS were a

better predictor of the patients' preferred eye and ability to
read continuous text than were acuity measurements. Con-
trast sensitivity functions also predicted reading performance
better than did a contrast sensitivity score obtained from a
summed grating measure for seven patients having macular
degeneration.
Contrast sensitivity bandwidth criteria identified 39 of 42

who were capable of reading and 10 of 13 who could not read.
Visual acuity predicted 53 patients who had additions within
the range of optical magnification, but only 42 were able to
read continuous text, while 2 of 13 patients were identified
who could not read. Therefore, contrast sensitivity as meas-
ured by the VCTS is a better predictor of a continuous read-
ing task than visual acuity. More research is needed to deter-
mine if these results generalize to other visual tasks
important to low vision patients such as recognizing faces,
street curbs and other objects.
These results help show that abbreviated measures of vis-

ual capability such as acuity and/or peak contrast sensitivity
values can ignore important individual differences in visual
performance. Although almost any lower contrast target can
potentially provide more information about visual capability
than will high contrast acuity targets with certain tasks, it
seems premature at this stage of research to use contrast
sensitivity to throwaway important information about indi-
vidual visual capability by not using sine-wave gratings. Rap-
id, clinically reliable tests such as the VCTS make obtaining
complete contrast sensitivity data clinically practical.
Finally, the mean and 90% of the population of the con-

trast sensitivity functions of the low vision patients able to
read continuous text with aids may be useful as a patient
screening test in determining which patients may be helped to
read with aids. It is suggested that continued determination
of the relationship between contrast sensitivity functions and
relevant low vision tasks may provide a clinically useful data
base to guide the analysis of visual capabilities of low vision
pa tients. Indeed, the VCTS is part of the Low Vision Func-
tional Battery of vision tests used for low vision evaluation at
the New York Lighthouse.
28
5. Acknowledgements
Acknowledgement is given to Ann R. Kurzer, Ken Blauvelt,

and Michelle Whitley Turner for their assistance with this
paper.
6. References
1. E.E. Faye: Clinical Low Vision (Little, Brown, Boston

1976)
2. Committee on Vision, National Research Council:

Emergent Techniques for the Assessment of Visual
Performance (National Academy Press, Washington
D.C. 1985)
3. A.P. Ginsburg: A new contrast sensitivity vision test

chart. Am. J. Optom. Physiol. Opt. 61,403 (1984)
4. G.C. Woo, H. Bohnsack: Comparison of the distance

and near Vistech Vision Contrast Test Systems (VCTS).
Can. J. Optom. 85, 12 (1986)
5. D.S. Loshin, J. White: Contrast sensitivity: The visual

rehabilitation of the patient with macular degenera-
tion. Arch. Ophthalmol. 102, 1303 (1984)
6. G.C. Woo, J.A. Wessel: Use of contrast sensitivity

function in prescribing low vision aids. Am. J. Optom.
Physiol. Opt. 59, 924 (1982)
7. L. Hyvarinen, P. Laurinen, J. Rovamo: Contrast sensi-

tivity in evaluation of visual impairment due to macu-
lar degeneration and optic nerve lesions. Acta
Ophthalmol. (Copenh.) 61, 161 (1983)
8. A.P. Ginsburg: Spatial filtering and vision: Implica-

tions for normal and abnormal vision. In Clinical
Applications of Visual Psychophysics, Symposium on
Clinical Applications of Visual Psychophysics in Con-
junction with the 2nd Study Group on Human Vision,
San Francisco, 1978, ed. by L.M. Proenza, J.M. Enoch,
A. Jampolsky (Cambridge University Press, New York
1981) pp. 70-106.
Threshold and Suprathreshold Deficits in
Color Vision in Optic Neuritis
Kathy T. Mullen, Gordon T. Plant
1. Introduction
Disturbances of color VISIon are frequently reported by

patients following an attack of optic neuritis (see
MACKA~ELL [1] for an illustrated account of the symptoms
of optic neuritis, and FOSTER [2] for a recent review of color
vision anomalies in optic neuritis). For example, typical com-
ments of our subjects have been that colors look duller, paler
or less glossy with their affected eye: reds have been report-
ed as pink, and greens as grey. Occasionally the subject will
report an apparent hue change, such as calling red orange, or
green yellow. Color anomalies such as these were first
described in the literature just over one hundred years ago by
NETTLESHIP [3]. Subsequently, in 1897, GUNN and
BUZZARD [4] reported that frequently colors cannot be rec-
ognized around absolute scotomas in optic neuritis. They
considered that deficits to red and green were chiefly
involved, and this type of observation is embodied in KOLL-
NER's rule [5] which holds that deficits to red and green but
not blue are characteristic of optic nerve disorders. The
introduction of pseudoisochromatic plates into neuro-
ophthalmic testing by SLOAN [6] emphasized the usefulness
of color vision testing in addition to visual acuity and visual
field examination in optic nerve disease. Tests such as the
Ishihara test and the Farnsworth-Munsell 100 hue test are
particularly sensitive indicators of a previous attack of optic
neuritis, even when visual acuity has returned to normal [7,8].
We consider two main questions in this paper, which are

addressed using two different experimental approaches.
Firstly, we investigate whether there is any evidence for a
selective loss of color vision at threshold in optic neuritis,
which might indicate, as the clinical literature implies, that
chromatic mechanisms are more severely affected than lumi-
nance mechanisms in the disorder. To answer this question,
we have used a method of testing color vision independently
of luminance vision. Secondly, we have investigated the
30
nature of the color deficits at suprathreshold levels, using

surface colors. We asked whether the suprathreshold deficit
can be accounted for by changes in perceived saturation or
whether specific hue shifts occur, and also whether the defi-
cit is selective for specific hues, as the clinical descriptions
have suggested.
2. Methods
2.1 Stimuli, Apparatus and Procedure: Threshold

Experiment
A red/green or a blue/yellow sinusoidal chromatic grating,

with a spatial frequency of 1 cycle per degree (c/deg) and
sinusoidally reversed at 0.5 Hz is produced by displaying two
gra tings varying in luminance, each on oscilloscope screens
with white (P4) phosphors. A diagram of the apparatus is
shown on the left side of Fig. 1. The two screens, placed at
right angles, are each viewed through a narrow band interfer-
ence filter. Interference filters with peak wavelength trans-
missions at 526 nm and 602 nm are used in the red/green
grating, and 470 nm and 577 nm for the blue/yellow grating.
The two monochromatic gratings are combined optically 180
degrees out of phase to form the composite red/green or
blue/yellow chromatic grating. The grating patch viewed is
circular and subtended 6.5 degrees in diameter. Viewing is
monocular and with a natural pupil. A fuller description of
the stimuli and methods is given elsewhere [9].
The contrast of either component grating is defined by

the usual formula:
Lmax Lmin
c
Lmax + Lmin
where Lmax and Lmin are the peak and trough luminance val-
ues respectively. The contrasts of the two component
monochromatic gratings are yoked together electronically so
that they are always equal, even though their respective
mean luminances may differ. Henceforth, these contrasts are
31
used to describe both the monochromatic and the chromatic

gratings. The mean luminances of the composite red/green or
blue/yellow gratings are both fixed, although the ratio of the
luminances of the two component colors in each is variabl~
The mean luminance of the red/green stimul~ is 15 cd/m
and that of the blue/yellow stimulus is l.l cd/m (see Fig. 1).
Output contrasts were calibrated using a UDT (United Detec-
tor Technology) light meter. All mean luminances were
measured with a calibrated SEI spot photometer.
s as IF L os GREEN RED
R
fl:G
'I,
0J\-f' +
[S] DO
~
0
<Q
= '?'s,-A-P + yA-A-, 50
f\p + dVv 100

(Contrast =I)
Figure 1. Left: apparatus used to produce the stimuli.

DS = display screen; L = lens for correction of
chromatic aberrations; IF = interference filter;
BS = beam splitter; S = subject. Right: diagram
of the luminance profiles across space of the
red and green component gratings which, when
added together 180 degrees out of phase,
produce a sinusoidal red/green chromatic stimu-
lus. The ratio of the red (R) to green (G) mean
luminances in the chromatic grating is variable,
and is expressed as the percentage of red light
in the mixture. Three ratios in the range are
shown. The mean luminance of each stimulus
(R + G) is constant. The contrasts of the red
and green component gratings are always equal
to each other and have a value of 1 in this
example. The same method is used to produce a
blue/yellow grating.
32
A method of adjustment was used to measure threshold.

The subject was able to increase or decrease the contrast of
the stimulus by pressing an appropriate button. A third but-
ton could be pressed to indicate the chosen threshold value.
A jump in contrast either above or below threshold occurred
after each threshold setting. Threshold was described to sub-
jects as being when a vertical bar pattern could just be seen,
and all subjects were given several practice runs before
beginning the experiment. Each data point represents the
mean of at least two thresholds set nonsequentially. General-
ly, however, the mean of 3 to S settings is taken for each
plotted point. Error bars show typical standard deviations
unless described otherwise.
2.2 Suprathreshold Experiment
A between eye matching paradigm was used in this study,

with non spectral stimuli viewed at suprathreshold levels.
Both test and matching stimuli were colored chips taken from
the Munsell Book of Color. The book is a color atlas of 40
pages. Each page consists of arrays of colored chips and all
the chips on a page have the same hue. Along each row the
chips increase in chroma {saturation} but have equal value
(lightness), and up each column the chips increase in value but
have equal chroma. Page by page the hue of the chips alters
and the complete atlas forms a color circle. Within each
range of hue, chroma and value, the increments between
chips are intended to be perceptually equal. Unique hues on
the Munsell system are SR, SY, SG and SPB.
A central screen was arranged to obtain dichoptic view-

ing conditions. The chips subtended 3 x 2.4 degrees at a
viewing distance of 40 cm and were viewed on a plain white
background. The subject viewed a test chip with their worse
eye, and used their better eye to select a matching chip.
Potentially matching chips were drawn from the book by the
experimenter, as requested by the subject, and were placed
on the same white background so that both test and matching
chips were view~d under equivalent conditions. Room lumi-
nance {l80 cd/m } with a color temperature of 7,000 K was
used.
33
2.3 Subjects
Nine subjects were studied. All had had an episode of optic

neuritis in the recent past which had left them with a stable
residual visual deficit. Most were cases of unilateral disease
and all showed a marked asymmetry between the two eyes.
This has permitted a comparison of luminance and chromatic
sensitivities and color appearance between the two eyes of
each subject.
3. Results
3.1 Threshold study
The principle of the experiment is illustrated on the right side

of Fig. 1. The ratio of the mean luminances of the two com-
ponent gratings in the stimulus is varied, covering the com-
plete range. The ratio is expressed as the percentage of red
in the red/green mixture or as the percentage of yellow in the
yellow/blue mixture. In the figure the range begins and ends
with a green or red monochromatic grating at 0 and 100% red
respectively. Such stimuli vary sinusoidally in luminance, but
are of uniform color and so thresholds will be based on the
detection of the luminance contrast in the stimulus. At a
point in the midrange, the stimulus will contain color differ-
ences but be of uniform luminance, thus detection thresholds
will be based on color contrast. Contrast sensitivity was
measured at nine points in both the red/green and blue/yellow
ranges for each eye of each subject.
Results for the two subjects on both red/green and bluet

yellow stitnuli are shown in Fig. 2. Each subject had suffered
a unilateral attack of optic neuritis and each of the affected
eyes has the lower overall contrast sensitivities. The better
eyes of the subjects illustrate the type of function typically
found for normal SUbjects at this spatial frequency [9]. Con-
trast sensitivity is greatest to the monochromatic gratings at
either end of each range (0% and 100% red; 0% and 100% yel-
low). As luminance contrast is removed from the stimulus,
contrast sensitivity declines and a minimum is reached. Nor-
mal subjects show such a minimum for both red/green and
blue/yellow stimuli at this spatial and temporal frequency.
The minimum is taken to indicate the contrast sensitivity of
34
the sUbject when thresholds are based solely on the detection

of the color contrast in the stimulus. Thus, we determined
the appropriate match for the intensities of the two colors in
the stimulus for each eye of each subject under the spatial
and temporal conditions which are used in the experiment.
The depth of each minimum (indicated by the vertical arrows
in Fig. 2), corresponds to the difference between the logs of
the contrast sensitivities to luminance (L) and chromatic (C)
gratings, or the log of the ratio of luminance to color con-
trast sensitivity (L/C). We found some variability in the posi-
tion of the minimum between subjects, and in some cases
between the better and worse eyes, but such differences were
small.
A consideration of the red/green results for subject PS

(bottom left panel) shows that the luminance contrast sensi-
tivities, given at each end of the range, are lower in the right
(affected) eye compared to the left (unaffected) eye. Color
contrast sensitivity, given at the minimum, is also reduced in
the right eye and by a greater extent than luminance sensitiv-
ity. The depths of the minima for the red/green gratings are
greater in the right eye than in the left (or the value log L/C
is higher), indicating that this subject has a greater deficit in
color sensitivity than luminance sensitivity. Similarly, a
selective color deficit is also found for the blue/yellow grat-
ings (bottom right panel), although it is slightly less marked
for these stimuli. A similar pattern of results is obtained for
the other subject IN (upper panels). The minima are deeper in
the left (affected) eye than the right (unaffected) eye, indi-
cating a selective loss of color sensitivity which is slightly
more marked for the red/green stimuli. One question arising
from these data is whether the deeper minima occur in the
affected eyes simply because their overall contrast sensitivi-
ty levels are lower. An analysis of our data shows that there
is no correlation between depth of minima and either lumi-
nance or color contrast sensitivity levels [10]. This confirms
that the greater log L/C ratios in the affected eyes are due
to a genuine, selective loss of color sensitivity, and not to an
anomaly of scaling.
In order to analyse our data we made comparisons

between the two eyes of each of our SUbjects. This serves to
eliminate from our results the high degree of variability
which is found in the color and luminance contrast sensitivi-
ties of the normal population. Eyes of each subject were cat-
egorized as better or worse according to their absolute color
35
, IN
1
IN
O~'----.r-~----.,~O
.0
RJR+O .. Y'Y+B ..
100 100
~tLE ~:~!LE
\/1-
. , P8
100
·~l·
1 , , - -_ _---,.-,_ _ _..:..:;P8
o 60 100
R/R+G" YIv+. "
Figure 2. Contrast sensitivity plotted as a function of

the red/green luminance ratio in the stimulus,
expressed as the percentage of red in the mix-
ture (left panels), or the yellow/blue luminance
ratio, expressed as the percentage of yellow in
the mixture (right panels). Data for the left
eye and right eye of two subjects (IN, PS; top
and bottom panels, respectively) are shown.
Contrast sensitivities at the minimum of each
function indicate the contrast sensitivity to
chromatic gratings, and contrast sensitivity at
either end of each function indicate contrast
sensitivity to the monochromatic luminance
gratings. The vertical arrows show the differ-
ence between the color and luminance contrast
sensitivities. The gratings have a spatial fre-
quency of 1 c/deg and are phase reversed sinu-
soidally at 0.5 Hz +/- I SD as shown for each
function.
contrast sensitivity levels. In Fig. 3, the log of the ratio L/C

36
is plotted for the better eye as a function of the ratio for the
worse eye of each sUbject. Data for both red/green and bluet
yellow gratings are shown. If the L/C ratios are the same in
the better and worse eyes (i.e. the depths of the minima are
equal), the data would lie along a slope of unity given by the
dashed line. Clearly, the results are not described by this
function. Instead, the results indicate that in most cases the
ratio L/C is greater in the worse than the better eyes, or the
color deficit is greater than luminance deficit. The data also
indicate that the color deficit is not proportionally greater
than the luminance deficit in which case the data would again
fall along a slope of unity with the intercept indicating the
constant of proportionality. The form of the data indicate
that the larger L/C ratios in the better eyes are associated
with disproportionately larger ratios in the worse eyes; this
means that the relationship in linear terms is a power func-
tion. In summary, overall in our population of subjects the
deficit for color contrast is greater than the deficit for lumi-
nance contrast. An analysis of the results given elsewhere
[10] indicates that overall the size of the color deficit is
about equal for the red/green and blue/yellow gratings.
3.2 Suprathreshold Study
In this part of the study we investigated the color deficits of

our subjects at suprathreshold levels. We asked two main
questions: firstly, whether the color deficit can be accounted
by changes in perceived saturation or whether specific altera-
tions in hue appearance occur, and secondly whether the defi-
cit is selective for specific hues. We used a between-eye
matching paradigm with stimuli taken from the Munsell Book
of Color (see Methods for details). We used four of our previ-
ous nine subjects, and one additional subject, all of whom had
one eye considerably more affected than the other, particu-
larly on the Farnsworth-Munsell 100 hue test.
Results for two subjects (PS and IN) are shown in Figs. 4
and 5 (their threshold data are given in Fig. 3). The results
are represented on a circular color wheel: the center of the
wheel represents white and chroma (saturation) increases in
perceptually equal steps radially along the spokes of the
wheel. Hue varies around the circumference of the wheel.
Series of matches were made using chips differing in hue and
chroma, but of substantially the same value (lightness), gen-
erally at 5 or 6. At least half the matches were repeated
37
/
/
/
/
o Red - green grot in g
/
oBlue-yellow grating
/
/
/
/
0/
/
/
/ o 0
0
/
o 0
/ 0
/
0·5
LOG LIe WORSE EYE
Figure 3. The ratio of the contrast sensitivity for lumi-

nance gratings (L) to the contrast sensitivity for
chromatic gratings (e), in log units, is shown for
the better eye as a function of that for the
worse eye of each subject. The calculation of
this ratio for two subjects (IN, PS) is shown in
Fig. 2. The eyes of each subject were categor-
ised as "better" or "worse" depending on which
had the greater or lesser contrast sensitivity to
chromatic gratings, respectively. The dashed
line has a slope of unity. Data for red/green
gratings (squares) and blue/yellow gratings (cir-
cles) are shown. Typical standard errors are 0.1
log units. (After Mullen and Plant, 1986)
once or twice and the plotted data points show the averaged
results. The origin of each arrow on the figures, marked by a
filled spot, indicates the hue and chroma of each test chip,
and the arrow head shows the matching chip.
For subject IN (Fig. 4), the results show large changes in

both hue and saturation in the affected eye. From 2.5 BG
through the green (G) and yellow (Y) chips to 2.5 Y the arrows
38
all point straight towards white. They indicate that very

large losses in saturation have occurred without any measura-
ble changes in hue. From 7.5 BG through blue (B), purple-blue
(PB) to 2.5 RP there are little or no changes in saturation but
there are large changes in hue. The hue shifts in the BG and
PB regions show that the test chips tend to be matched to
hues of 2.5 PB or 5 PB; 5 PB is a unique hue. Another less
dramatic hue shift occurs around SR - 7.S R; 5 R is another of
the four unique hues. The changes in value (not shown) were
evenly distributed across hue. On average the matched chips
were 2 Munsell units darker than the test chips, corresponding
to approximately 0.2 neutral density (ND) units. In this sub-
ject the changes in perceived hue have occurred independent-
ly of the large alterations in perceived chroma, and neither
appear to be correlated with alterations in perceived value.
Figure S shows the results for another subject (PS). This

subject has quite large losses in perceived chroma at all hues,
although these are most marked in the GY (green/yellow) and
Y (yellow) regions. Hue shifts can be identified in two
regions. Hues in the GY and BG regions tend to be matched
with SG, and those in the RP region are matched with 2.S or
SR; SG and SR are unique hues. Changes in perceived value
were uniformly distributed across hue and overall the average
change was zero. This pattern of results was repeated in the
other subjects we tested. We found changes in perceived
chroma occurring at most or all hues. When perceived hue
altered, the match tended to be made with one of the four
unique hues. Changes in value occurred at all hues, and we
could find no correlation between the sizes of the changes in
hue and those for the changes in chroma or value.
39
>-
<Q '"-0
>-
a
a-u
In
'"<3:. if
<Q
IN 5SG
Figure 4. Results from the between-eye matching exper-

iment described in the text. The hue and chro-
ma of the test and matching stimuli are plotted
in Munsell co-ordinates on the color wheel. Hue
is given around the circumference of the wheel
and the center of the wheel represents white
with chroma increasing along the radii. The
origin of the arrows (filled circles) indicates the
hue and chroma of each test chip and the arrow
head indicates that for the matching chip. Sub-
ject: IN.
40
..,>- ll'
-<l
>- C;
a -u
ro
U'
Gl ..,~
PS 5BG
Figure 5. Results for subject PS. See legend of Fig. 4 for

details.
4. Discussion
The first part of the study has shown that overall in our group
of subjects there is a deficit in color contrast sensitivity
which is greater than the deficit in luminance contrast sensi-
tivity. This deficit was about equal for red/green and blue/
yellow gratings. Other studies, using different stimulus
arrangements, have not revealed a selective color deficit at
threshold. ALVAREZ et al.[11] and ALVAREZ and
KING-SMITH [12] compared the incremental sensitivity of
opponent color mechanisms (isolated using a 1 degree test
spot on a white background) with that of luminance mecha-
nisms (isolated by flickering the test spot). They report that
the luminance sensitivity is reduced more than the color sen-
sitivity under these conditions. FOSTER [13,2] used a method
41
of isolating and comparing color and luminance detection

thresholds under the same temporal conditions, using very
small (0.25 degree) test spots. Results indicate that the color
and luminance deficits are about equal.
An approach more closely resembling our own has been

used by FALLOWFIELD and KRAUSKOPF [14]. A uniform
field (2 degrees) was modulated in saturation while the lumi-
nance of the field was held constant. By comparison with
normal subjects thresholds for detection changes in saturation
were elevated compared to those for detection of a change in
brightness. Thus the results of this study support those
reported here. However, since F ALLOWFIELD and
KRAUSKOPF [14] made comparisons between color and lumi-
nance thresholds using different scales for the measurement
of each, direct comparisons of their magnitudes are difficult
to interpret. While at this stage it is not possible to explain
why all these studies have produced different results, one
possible cause lies in the very different spatial and temporal
configurations of stimuli which have been used.
One potential explanation for the selective color loss

which our results show is that optic neuritis selectively
affects mechanisms sensitive to color contrast. Since in the
primate it is the ganglion cells with small diameter axons
{projecting to the parvocellular layers of the lateral genicu-
late nucleus} which respond to color differences (GOURAS
[15], DeMONASTERIO and GOURAS [16]) it is possible that
these axons are selectively affected by the disease. How-
ever, various other possibilities should also be considered. As
mentioned above, it is not yet known whether a selective col-
or deficit occurs for stimuli of other spatial or temporal fre-
quencies. For example, the differences which exist between
the effect of frequency on spatial sensitivity of color and
luminance vision {VAN DER HORST and BOUMAN [17];
MULLEN [9]} might mean that the greatest deficits in color
and luminance vision occur at different spatial frequencies,
or possibly in different regions of the visual field. These
aspects remain to be investigated.
In the second part of the study, subjects who had selec-

tive losses of color contrast at threshold were investigated at
suprathreshold levels. These results suggest that the predom-
inant deficit is a loss of perceived chroma {saturation} which
may occur at any hue. This is supported by the subjective
reports of our subjects that colors look paler, dimmer or less
42
glossy through their affected eye. The results are also sup-
ported by other clinical reports (GLASER [18]) that deficits in
saturation are a sensitive indication of optic nerve disease.
However, it has also been suggested (see Introduction) that
deficits of red and green are especially characteristic of
optic nerve dysfunction. Our data do not support this conclu-
sion since across our group of subjects deficits in perceived
chroma (saturation) were also found. These changes were
such that hues tended to be matched to one of the unique
hues. Thus, these results also contradict any suggestion that
only reds and greens are affected. It is also worth emphasiz-
ing the great range of results which we obtained among our
subjects, both in the extent by which different hues were
affected by losses in perceived chroma and in the range and
type of hues which altered their color appearance.
One question which arises is whether the changes in per-

ceived hue occur secondarily to any alterations in the per-
ceived chroma or value of the test chips. For examplet in
normal subjects, reducing the mean luminance of a test light
causes its color appearance to alter towards one of the three
invariant hues (which are similar to three of the unique hues),
and these changes produce the Bezold-Brucke phenomenon
(see HURVICH [19]). Thus a perceived darkening of the test
chip in the affected eye, or a perceived loss in luminance
contrast might account for the alterations in color appear-
ance. Similarly, the changes in hue might arise from the loss
of perceived chroma. There is evidence against this explana-
tion since the greatest changes in hue are not associated with
the greatest changes in value or chroma; this is especially
marked in the case of IN (Fig. 4). A pilot study on a normal
subject also confirms that the value changes manifest in our
subjects (usually of less than 2 Munsell units) cannot account
for their alterations of hue perception. In summary, our
results suggest that optic nerve disorders are associated with
loss of perceived saturation, which can occur for any hue, and
that in addition, genuine alterations in perceived hue can
occur. These effects are likely to be coupled with selective
deficits for the detection of color contrast at threshold.
There is no evidence that red or green hues are more likely to
be affected.
43
5. Acknowledgements
We are grateful to our subjects for their kind co-operation

and we thank Dr. I.M.S. Wilkinson for permitting us to study
patients under his care.
6. References
1. P. MacKarell: Interior journey and beyond: An artist's

view of optic neuritis. In Optic Neuritis, ed. by R.F.
Hess, G.T. Plant (Cambridge University Press, Cam-
bridge 1986) pp. 283-293.
2. D.H. Foster: Psychophysical loss in optic neuritis. In

Optic Neuritis, ed. by R.F. Hess, G.T. Plant (Cam-
bridge University Press, Cambridge 1986) pp. 152-191.
3. E. Nettleship: On cases of retro-ocular neuritis.

Trans. Ophthalmol. Soc. U.K. 4, 186 (1884)
4. M. Gunn, T. Buzzard: Discussion on retro-ocular neu-

ritis. Trans. Ophthalmol. Soc. U.K. 17, 107 (1897)
5. H. Kollner: Die Storungen des Farbensinnes: Ihre

Klinische Bedeutung und Ihre Diagnose. (Karger, Ber-
lin 1912)
6. L.L. Sloan: The use of pseudo-isochromatic charts in

detecting central scotomas due to lesions in the con-
ducting pathways. Am. J. Ophthalmol. 25, 1352 (1942)
7. B.H. Lynn: Retrobulbar neuritis: A survey of the

present condition of cases occurring over the last
fifty-six years. Trans. Ophthalmol. Soc. U.K. 79, 701
(1959)
8. J.F. Griffin, S.H. Wray: Acquired color vision defects

in retrobulbar neuritis. Am. J. Ophthalmol. 86, 193
(1978)
9. K.T. Mullen: The contrast sensitivity of human color

vision to red-green and blue-yellow chromatic grat-
ings. J. Physiol. (Lond.) 359, 381 (1985)
44
10. K.T. Mullen, G.T. Plant: Colour and luminance vision

in human optic neuritis. Brain 109, 1 (1986)
11. S. Alvarez, P .E. King-Smith, S.K. Bhargava: Lumi-

nance and colour dysfunction in l'etrobulbar neuritis.
In Colour Vision Deficiencies VI, Proceedings of the
International Symposium, Berlin, 1981, ed. by G. Verri-
est, Doc. Ophthalmol. Proc. Sere Vol 33 (Dr. W. Junk,
The Hague 1982) pp. 441-443.
12. S.L. Alvarez, P.E. King-Smith: Dichotomy of psycho-

physical responses in retrobulbar neuritis. Ophthalmic
Physiol. Opt. 4, 101 (1984)
13. D.H. Foster, R.S. Sne1gar, J.R. Heron: Nonselective

losses in foveal chromatic and luminance sensitivity in
multiple sclerosis. Invest. Ophthalmol. Vis. Sci. 26,
1431 (1985)
14. Lo Fallowfield, J. Krauskopf: Selective loss of chro-

matic sensitivity in demyelinating disease. Invest.
15. P. Gouras: Identification of cone mechanisms in mon-

key ganglion cells. J. Physiol. (Lond.) 199, 533 (1968)
16. F .M. DeMonasterio, P. Gouras: Functional properties

of ganglion cells of rhesus monkey retina. J. Physiol.
(Lond.) 251, 167 (1975)
17. G.J .C. Van der Horst, M.A. Bouman: Spatiotemporal

chromaticity discrimination. J. Opt. Soc. Am. 59,
1482 (1969)
18. J.S. Glaser: Clinical evaluation of optic nerve func-

tion. Trans. Ophthalmol. Soc. U.K. 96, 359 (1976)
19. LoM. Hurvich: Color Vision (Sinauer Associates Inc.,

Sunderland 1981) pp. 72-75.
Contrast Sensitivity
R.A. Weale
1. Introduction
Contrast sensitivity has been measured during the last quar-

ter of a century or so as a function of a large number of
parameters including that of age. A handful of detailed stud-
ies [1-6] have shown that, in general, low spatial frequencies
are unaffected by senescence, whereas there is a reduction in
contrast sensitivity at those on the high frequency side of the
customary peak.
It is held by some to be unlikely that this can be

explained in terms of a deterioration of the ocular media.
The quality of the lens as an image-forming device remains
unimpaired (Fig. 1) for long spectral wavelengths [7] and
MORRISON and McGRATH [8] have been able to rule out
optical factors by comparing data obtained with external
gratings and interference fringes imaged on the retina in
turn. The tentative conclusion of these considerations is that
the high-frequency deficit is due to neuro-senescence.
The object of this brief note is to examine the plausibility

of this hypothesis. The suggestion advanced here is that
there is more than one mechanism available to the eye-brain
for the detection of contrast. This is not new [9]. The
present slant rests on the view that low frequency mecha-
nisms are determined genetically and function soon after
birth, whereas high frequency ones are developed by visual
experience and are comparatively frail.
46
0·09
10 0·1
-
m
...
Q)
a.
~ 5 0·2
.; 0 0·25
u
0·3
0·4
0·5
0·7
1·0
o o 30 50
Age (years)
Figure 1. Cut-off period of human crystalline lenses as a
function of age. The scale on the RHS gives
the efficiency in terms of the diffraction limit.
Females:<p-, males: ~. The half-black symbols
represent cataractous lenses. The dashed line is
the regression through F, the continuous one
through M (WEALE [7]).
2. Focus and Luminance
An early indication of the fundamentally different functions

at low and high frequencies derived from the observation [10]
that the contrast sensitivity for a grating is independent of
focus at low, but not at high spatial frequencies (Fig. 2). This
has some interesting corollaries, e.g. some optical illusions
such as that due to Zollner vanish when their retinal image is
slightly defocused [11]. It would appear to follow that such
illusions should be invisible to those suffering from high-
frequency loss.
Another example of a marked dlfference between low and

high frequencies relates to the ratio of 6. log contrast-
sensitivity/6. log luminance, which equals -F, where F is the
47
1000
i'
..=~
Z"
0
100
(,)
~
'0
i
2:
..c:
..
~
>.
os:
~
0;;
..=~
III 10
..=
0
0
1 L.._.....L_--'_ _...l....-_--L._ _'--_-L_ _ __

-2 -1 o +1 +2 +3 +4
Lens power, dioptres
Figure 2. Contrast sensitivity as a function of ocular

refraction•• : 1.5 c/deg; 0: 9 c/deg; 0: 22
c/deg; .: 30 c/deg (Courtesy CAMPBELL and
GREEN [10]).
Fechner "constant". I measured this with an orange filter

(Ilford 607) interposed between a sinusoidal grating display
and the right eyes of the two observers whose results are
shown in Fig. 3. The insets indicate the approximate size of
field used, together with the upper limit of the retinal illumi-
nance. The data (0 and <» are compared with those due to
VAN MEETEREN and VOS [12] who used white light, and also
those obtained by FIORENTINI and MAFFEI [13]. There is a
systematic difference correlated with luminance level; how-
ever, all sets of data show that the ratio of tJ. log contrast-
sensitivity/ tJ. log luminance is low or zero at low spatial fre-
quencies, and then rises with the frequency. There is
48
evidently a drop at the highest frequencies: if these cannot

be resolved then a reduction in intensity cannot get round this
failure. Note that, just like the variation with focus, with
retinal illuminance the drop is small at low frequencies.
1·5
'; 1·0
o
<l
'"
(/)
o o I
I
u &.
o
/
OJ
o ~//
<l 0·5 o /
o M'
/
00 /0
/
o ____- L______ ____ ______ ____ ____
-1·0 o 1·0 2·0

log c/o
Figure 3. Gradients of log contrast-sensitivity/log lumi-
nance as a function of spatial frequency •• :
FIORENTINI and MAFFEI [13] IA: VAN
MEETEREN and VOS [12]; 0 and o present data.
49
3. Amblyopia
The detailed studies carried out on amblyopes [14,15] offer

numerous examples which show that, in different types of
amblyopia, deficits occur which are marked at high, but
weaker or altogether absent at low spatial frequencies (Fig.
4). Insofar as amblyopia may be due to retarded or otherwise
impaired postnatal visual development, these data buttress
our working hypothesis.
SPATIAL SENSITIVITY
1000
o.s.
Tr *1 Hz
X: 10'
100
...>
...
~ 10
in
z
w
'"... 1 I
'"'a:" 0·1 10 100 0'1 10 100
;Z1000 1000
0
OL TS
u Tr : 1 Hz Tr : I Hz
X= 20' x= 20'
100 ....9"
9- ft \ 100
o~(J()- °'-0
/0/'
/9 ft,,
,,
(/0---
0.... ./!....
\
°,
10 Q /1....,-. , 10 !---!/ "\ ' ,,
,' ....
I~ ____L -____L-~~
!
10 _01 0'1 1 10 0'01 0'1 10
SPAliAl FREQUENCY (c Id.g)
Figure 4. Contrast sensitivity as a function of spatial

frequency as measured in different types of
ambylopia (Courtesy Hess et aZ. [14]).
50
4. A Computer Experiment
Low spatial frequencies can, however, create a problem of

their own. When the spatial wavelength is large systematic
sampling errors can occur. For example, at a frequency of
0.2 c/deg, only three cycles will appear in a 15 degree field.
One asks whether, at threshold, this number is sufficient to
give a reliable measure. More generally, the problem can be
put as follows. If the standard deviation of the noise in the
sensor compares with the contrast amplitude to be detected,
how many cycles are needed on the detecting field for the
response to be well above chance level? The problem can be
tackled in more than one way, but sampling theory yields the
data shown in Table 1.
Table 1
Number N of cycles per sampling area as function of P when

the signal and average noise amplitudes are equal.
p N
0.05 2.5
0.01 4
0.001 6.5
Note that the extent of the fovea is less than one degree
of arc: on the basis of this argument contrast sensitivity
should approach the values to be expected from the ocular
modulation transfer function only for (angular) frequencies
higher than those shown in Table 1. This is observed. There-
fore the relatively high sensitivity/luminance gradient seen in
FIORENTINI and MAFFEI's data [13] may be due to the very
large field which they used. This is evidently a point to
check.
51
5. Senescence
OWSLEY et al.[5] addressed the question of the variation of

retinal illuminance as a function of age. It is evident that
clinical tests of visual acuity ignore the fact that less light
may reach the older patient's retina than is true of younger
eyes [16]. This may be due both to senile miosis and lenticu-
lar yellowing. The above workers showed that, when full
allowance is made for this factor, the decrement due to
senescence persists even though it is significantly reduced
(Fig. 5). Note, however, that, as in all the other examples
quoted, sensitivity at low frequencies are preserved.
500 500
200 200 ... - " ,,

.....
....... "'\ ,
100 100 \
, \
\
\
50 50 '\ \
~ A \ \
;;
;; \
\ 20', -no filter
in
c 20 20 \ 20's -filter
Q)
en \
10 10 \ 60's
5 \80' s -.- 5
2 N·91 2
0.5 I 2 4 8 16
Spatia I frequency, c / deg
Spatial frequency, c/deg

frequency. Left: different age groups. Right:
two age groups, with the younger of the two
repeating the measurements when the retinal
luminance of the target has been made equal to
that of the older group (Courtesy OWSLEY et
al. [5]).
52
6. Infancy
The last decade or so has brought to light a great deal of

information on the development of infant vision. Data on con-
trast sensitivity form no exception to this. ATKINSON et al.
[17] were the first to demonstrate that a low-frequency sys-
tem is available very early in life (Fig. 6), with visual experi-
ence leading to the development of one capable of tackling
frequencies resolved by the adult eye. It is idle to speculate
at this stage whether the genetically primary low-frequency
system is governed by rod mechanisms or whether the latter
respond only to low frequencies because they evolved before
the photopic system.
Suffice it to emphasise that the low-frequency system is

the sturdier of the two, which is probably why it forms the
basis for low vision.
53
, (1,5
- ......
c~
"-
\
\
\ ,
\
\
\
\
\
\
......-0
----, 0
o
.
/0// • '\
~o
/ \
~. \
\
50 \
\
IOOL-.._ _...J..._ _'--_-L.._ _ _'--~....J...._
\
O,c U,S • .5. 10 0
Spatial frequency (cycles per deg)

frequency Top: adult. Bottom: infant. White:
flashing gratings. Black: drifting gratings
(Courtesy ATKINSON et aZ. [17]).
7. Acknowledgements
I thank Miss G.M. Villermet for acting as an observer and for

technical assistance, and Mr. Peter Clarke for data-
processing in connection with Table 1.
54
8. References
1. K. Arundale: An investigation into the variation of

human contrast sensitivity with age and ocular pathol-
ogy. Br. J. Ophthalmol. 62, 213 (1978)
2. L.D. Beazley, D.J. lllingworth, A. Jahn, D.V. Greer:

Contrast sensitivity in children and adults. Br. J.
Ophthalmol. 64, 863 (1980)
3. G. Derefeldt, G. Lennerstrand, B. Lundh: Age varia-

tions in normal human contrast sensitivity. Acta
Ophthalmol. (Copenh.) 57, 679 (1979)
4. C. McGrath, J.D. Morrison: The effects of age on spa-

tial frequency perception in human subjects. Q. J.
Exp. Physiol. 66, 253 (1981)
5. C. Owsley, R. Sekuler, D. Siemsen: Contrast sensitivi-

ty throughout adulthood. Vision Res. 23, 689 (1983)
6. R.A. Weale: Age and human contrast sensitivity. J.

Physiol. (Lond.) 360, 25P (1985)
7. R.A. Weale: Transparency and power of post-mortem

human lenses: Variation with age and sex. Exp. Eye
Res. 36, 731 (1983)
8. J.D. Morrison, C. McGrath: Assessment of the optical

contributions to the age-related deterioration in
vision. Q. J. Exp. Physiol. 70, 249 (1985)
9. C. Blakemore, F.W. Campbell: On the existence of

neurones in the human visual system selectively sensi-
tive to the orientation and size of retinal images. J.
Physiol. (Lond.) 203, 237 (1969)
10. F.W. Campbell, D.G. Green: Optical and retinal fac-

tors affecting visual resolution. J. Physiol. (Lond.)
181,576 (1965)
11. R.A. Weale: Experiments on the Zollner and related

optical illusions. Vision Res. 18, 203 (1978)
55
12. A. van Meeteren, J.J. Vos: Resolution and contrast

sensitivity at low luminances. Vision Res. 12, 825
(1972)
13. A. Fiorentini, L. Maffei: Contrast in night vision.

Vision Res. 13, 73 (1973)
14. R.F. Hess, T.D. France, U. Tulunay-Keesey: Residual

vision in humans who have been monocularly deprived
of pattern stimulation in early life. Exp. Brain Res.
44, 295 (1981)
15. R.F. Hess, A. Bradley, L. Piotrowski: Contrast-coding

in amblyopia. I. Differences in the neural basis of
human amblyopia. Proc. R. Soc. Lond. (Biol.) 217, 309
(1983)
16. R.A. We ale: A Biography of the Eye: Development,

Growth, Aging (H.K. Lewis, London 1982)
17. J. Atkinson, O. Braddick, F. Braddick: Acuity and con-

trast sensitivity of infant vision. Nature 247, 403
(1974)
The Effect of Blur Upon Psychophysical
Receptive Field Properties
Marcus D. Benedetto, Ernest M. Gaynes, Arnold H. Gordon,

Morris J. Mintz
1. Introduction
The receptive field is one of the basic functional elements in

vision. It is defined as that "area of the visual field within
which a light stimulus can cause a change in the average fir-
ing rate of a single retinal ganglion cell" [1]. Experimentally,
receptive fields are usually studied in animals, due to the
necessity for electrode tissue invasion. In humans, this can
rarely be done. As a result, human psychophysics must be
confined to the examination of receptive field analogs (or
psychophysical receptive fields) which are determined from
the analysis of tests designed to manifest localized response
properties. There are different types of receptive fields just
as there are different receptive field properties; the focus of
this particular research endeavor is the relationship of the
center-surround characteristics of psychophysical receptive
fields to induced optical blur, retinal eccentricity, and sur-
round parameters.
BARLOW et al.[2,3] noted an antagonistic center-

surround relationship under single unit recording conditions
with varying distributions of light in an annular zone centered
on the receptive field. Under dark-adapted conditions, pure
"on" and "off" units were found. Under light-adapted condi-
tions these units became, respectively, "on-center with off-
surround" and "off-center with on-surround". Within the cen-
ter, increasing stimulus area produced a summation effect;
but when light extended into the surround, inhibition was
seen. WERBLIN and COPENHAGEN [4] introduced surround
motion as a factor in the organization of physiological
center-surround receptive field properties. WERBLIN and
COPENHAGEN [4] described two separate kinds of surround
inhibition and related these to different layers and cells: I}
static surrounds produced sustained inhibition present at the
bipolar cells and organized at the outer plexiform layer,
where lateral antagonistic interactions are mediated by the
57
horizontal cell system, and then passed through the inner

plexiform layer; 2) moving or dynamic surrounds produced
selectively active lateral antagonistic interactions mediated
by the amacrine system in the inner plexiform layer. Hori-
zontal cells responded tonically to the presence of illumina-
tion within their receptive fields, but amacrine cells respond-
ed only transiently to changes in illumination within their
receptive fields. WERBLIN and COPENHAGEN [4] used the
terms "sustained" and "transient", without their current con-
notations, to describe the differential response characteris-
tics of the retinal ganglion cells to these two conditions.
Each system of lateral interneurons were activated by specif-
ic stimuli and each contributed to the control of sensitivity
within the central portion of the receptive field. The hori-
zontal cells responded to total luminous flux and formed a
concentric antagonistic surround for the ganglion cell system
via the bipolar cells. The amacrine cells responded to pattern
movement and added another dimension to the antagonistic
surround.
WESTHEIMER [5-7] adapted the physiological work of

BARLOW, FITZHUGH and KUFFLER [2,3] for psychophysical
testing with analagous results. He found that the psychophy-
sical response to a small central test flash exhibited summa-
tion and inhibition effects with varying sizes of static sur-
rounds. ENOCH [8] using a methodology similar to
Westheimer, examined the psychophysical analog of relation-
ships, with analogous results. This test has proven valuable in
clinical applications.
The examination of psychophysical receptive field prop-

erties has equally important applications in the investigation
of some of the underlying mechanisms of vision. Given that
the receptive field is a basic unit of visual processing and
that blur and retinal eccentricity are fundamental variables
in vision, the investigation of the effects of blur and retinal
eccentricity on psychophysical center-surround receptive
field relationships was a natural progression.
The results of this experiment should help to define modi-

fications of psychophysical receptive field center-surround
relationships in increasingly peripheral vision (retinal eccen-
tricity) and compromised optical imagery (blur). The motiva-
tion for this research was, first, additional understanding of
basic visual mechanisms and, second, to provide a tool for the
investigation of the function of the compromised visual sys-
tem. The availability of an effective technical approach may
58
encourage further investigation of the underlying mechanisms

of residual visual function, with refractive error as a first
approximation for a quantifiable simulation, or model, of low
vision.
2. Methods
2.1 Subjects
Three normal observers, ranging in age between 20 and 40

years, were carefully selected for their consistency and reli-
ability on psychophysical threshold measurements. Normal,
as pertains to this particular set of experiments, was defined
as free of visual defects, ocular pathologies, and correctable
to a visual acuity of 6/6 (20/20). This was confirmed by full
ophthalmological examinations.
2.2 Apparatus
Figures 1 and 2 show the apparatus and schematic, respec-

tively. Light was projected on the tangent screen of a com-
mercial perimeter (Autoplot, Bausch & Lomb) from four dif-
ferent light sources: the perimeter projection system, two
high intensity slide projectors, and a 40 watt circline fluores-
cent tube. The perimeter projection system was used to pro-
vide a fixation target and to place the test stimuli in the
appropriate position in the visual field. One high intensity
slide projector was used to provide the test spot which was
centered in the surround provided by the second high intensity
slide projector. The intensities of the center test spot and the
surround stimuli were modulated by independent sets of
counterbalanced neutral density wedges. Light from the first
high intensity slide projector passed through the center test
spot slide, a focussing lens, a shutter, neutral density wedge
system, onto a front surface mirror, another front surface
mirror and onto the screen. Light from the second high
intensity slide projector passed through the surround stimulus
slide mounted in a mechanical rotator, a focussing lens, neu-
tral density wedge systems, onto a front surface mirror,
another front surface mirror and onto the screen. The output
of the circline bulb was diffused and used to provide a homo-
geneous background adaptation field. The system was cali-
59
bra ted with a Prichard photometer {Photo Research}. This

system, not unlike that described by ENOCH [8], permits the
examination of visual psychophysical thresholds for the influ-
ence of static and dynamic surrounds on a central spot as a
function of visual field position and visual blur.
Figure 1. Test apparatus showing master control panel,

optical housing, and stimulus on screen.
2.3 Stimuli
The stimuli used in this experiment were those used by

ENOCH for the examination of the "receptive field-like prop-
erties" with the Westheimer/Werblin functions [8]. These
consisted of circular sectored discs (windmills) of various
sizes and light/dark ratios. The size of each set of windmill
stimuli varied with respect to visual field test location. A
preliminary set of experiments were conducted to verify that
the diameters chosen by Enoch corresponded to the appropri-
ate values on the inhibition arm of the Westheimer function
and permitted the reproduction of this portion of the func-
60
SCHEMATIC OF OPTICAL SYSTEM

Projection System P2 Projection System P 1
S2 - Llgllt Source S 1 - Llgllt Source
T2 - Telt FI.III Stimulul T 1 - Windmill Stimulul Slide
L2- Lenl R 1 - Slide Rot.tor
H2-SIIuUer L 1-Lenl
W2 - Neutr.1 Denllt, Wedge W1b-Neutr.1 Denllt, Wedge
F2 - Filter Wlleel W1e-Neutr.1 Denllt, Wedge
M2. - Front Surf.ce Mirror F 1 - Filter Wlleel
M2b-Front Surf.ce Mirror M1I-Front Surf.ce Mirror
M 1b - Front Surf.ce Mirror
F2 W2 H2 L2 T2 u
P - Pro jectlon Screen A - Auto - Plot

I- Background illuminator B-SubJect With Head
-Beam Path on Chinrest
-------- - Line Of Sight
Figure 2. Optical schematic of test apparatus.
tion. Slide mounted pinholes were used for the center test
spot.
2.4 Stimulus Parameters
The parameters varied in this study were visual field position,

amount of blur, motion and light/dark ratio of the windmill
arms. The locations in the visual field which were examined
were 0, 2, 4, 10 and 16 degrees. At each visual field position,
threshold sensitivity was measured with induced blur of 0.0,
1.0, 3.0, 5.0, 7.5 and 10.0 diopters of plus sphere. The stimuli
either moved at 2 rotations per second or were stationary.
Light/dark ratios of 25/75, 50/50, and 75/25 were used. The
61
center test spot was flashed for 150 msec per second. The
surrounding test stimuli consisted of the appropriate size
Enoch/Werblin stimuli for each eccentricity. The threshold
sensitivity for each of these conditions was examined under
two situations: stationary (static surround) with center test
flash, and rotating (dynamic surround) with center test flash.
2.5 Procedures
The subject was seated at the apparatus with chin in chinrest,

head in headrest, fellow eye patched, and tested eye cyclo-
pleged to arrest accom~odation. The background adaptation
field was set at 10 cd/m to avoid Troxler's effect. The sub-
ject's sensitivity to the flashed central test spot was deter-
mined. The central test spot was then set at 0.7 log units
above threshold. The intensity of the surround was then
increased until the flashing central test spot could not be
detected. The surround was then increased 0.2-0.5 log units
greater and then decreased until the flashing central test spot
was detected. The surround was then decreased 0.2-0.5 log
units. Four of these appears/disappears determinations were
made for every stimulus configuration (i.e., static, dynamic,
light/dark ratio, amount of blur, and eccentricity).
3. Results
Figures 3 &: 4 show the effect of blur across eccentricity with

different surround conditions. The results presented are
characteristic for all three subjects tested under all condi-
tions.
Figure 3 is a plot of threshold (in log candelas per meter

squared) versus eccentricity for the static 25/75, 50/50,
75/25% light/dark ratio windmill surrounds for each of 0, 1, 3,
5, 7.5 and 10 diopters of spherical blur (D.S.). Figure 3 shows
that the thresholds for the three light/dark ratios are central-
ly affected by blur but peripherally remain relatively intact.
Figure 4 is a plot of the magnitude of the response (dif-

ference between static and dynamic conditions) afforded by
the dynamic characteristic of the rotating windmill surround
conditions for the 25/75, 50/50, 75/25% light/dark ratio wind-
62
ST ATIC SURROUND
2.0
COl
:I
~ 1.5
u
O. LIGHT I DAM RAno

21171 71/21 21171 71121 21171 71/2. 21171 7112. 21171 7112.
L .0110 I L.!,01l0 I L.!,01l0 I L!01l0 I L .0/!!.....J
0' 2 ' 4' 10 ' Ie '
ECCENTRICITY
Figure 3. Representative threshold data in candelas per

meter squared for the static surround condition
at 0, 2, 4, 10 and 16 degrees with 25/75, 50/50
and 75/25% light/dark ratio stimulus configura-
tions for 0, 1, 3, 5, 7.5 and 10 diopters sphere of
optically induced blur.
mill surrounds for each of 0, 1, 3, 5, 7.5, and 10 diopters of

spherical blur. Figure 4 shows that the magnitude of the
response increases with eccentricity. The closeness of the
values at 16 degrees as compared to 0 degrees demonstrates
that the effect of blur decreases as eccentricity increases.
63
DYNAMIC SURROUND
1.5 0.0
B - 1.0
C - 3.0
COl
D - 5.0
:a::I 1.0
U
:~j.~r:~': ;~";
III
!5a:
...is~ 0.5
D~:g~'.'.
0.0
.
:~':;"'--
LIGHT/DARK RATIO
21175 75/25 25/75 75/25 25/75 75/25 25/75 75/25 25/75 75/25
SO/50 LSD/50 L...!.0/!!!......J L...!.0/!!!......J ~/50
O' 2 • 10 • 15 •
ECCENTRICITY
Figure 4. Representa tive da ta in candelas per meter

squared for the dynamic surround condition at
0, 2, 4, 10 and 16 degrees with 25/75, 50/50 and
75/25% and 10 diopters sphere of optically
induced blur. The axis which shows the magni-
tude or the response is the difference between
the static and the dynamic threshold conditions.
4. Discussion
The blurred psychophysical static/dynamic function appears

shifted in comparison to a normal, unblurred psychophysical
static/dynamic function (Figs. 3 & 4). This appears to be the
case centrally and peripherally. Centrally, the psychophysi-
cal dynamic function appears to be shifted less than the stat-
ic psychophysical function. Whereas peripherally, the psycho-
physical dynamic function appears to be even less affected by
blur than the central or peripheral psychophysical static func-
tion or the central psychophysical dynamic function.
64
It appears that, relative to center-surround relationships

{with a static or dynamic windmill as the surround for a cen-
ter test flash} a dynamic surround yields a lower threshold for
the center test flash than does a static surround. Blur
increases the threshold to the center test flash for a static
surround more than with a dynamic surround. With increasing
eccentricity the threshold for the center test flash was pro-
gressively reduced more for the dynamic surround than for
the static surround. At progressively greater eccentricities
blur increased the center test flash threshold with a static
surround and/or decreased the center test flash threshold
with a dynamic surround.
Generally, under surround conditions, blur appears to

have a greater effect upon threshold centrally than peripher-
ally for static and dynamic conditions. In fact, in some cases,
central thresholds were raised as much as a log unit differ-
ence from the peripheral thresholds. For the surround
condition, dynamic thresholds were lower than static thresh-
olds, and with greater eccentricity came a greater separation
between the static and dynamic thresholds {with the static
thresholds generally becoming more elevated}. The light/dark
ratio of the windmill surround had the least effect on thresh-
old under all stimulus conditions. Eccentricity had the great-
est effect. The order of magnitude of effect is as follows
{from greatest to least}: Centrally - blur, motion, light/dark
ratio; peripherally - motion, light/dark ratio, blur.
These results suggest that blur and eccentricity have a

substantial systematic effect upon stimulus perception and
upon center-surround relationships. Generally, as blur
increases, threshold also increases and as eccentricity
increases the effect of blur decreases. These data also seem
to indicate that under certain conditions motion and light/
dark ratio have less effect upon threshold than blur or eccen-
tricity.
The usual course of events is that psychophysical

research is developed based upon neurophysiological findings.
Psychophysical research has become prominent for one rea-
son, because it is a non-invasive means of gathering data.
These psychophysical analogs appear to follow the same rules
as the neurophysiological correlates. These data should
inspire further parametric investigation.
65
5. References
1. D. Cline, H.W. Hofstetter, J.R. Griffin: Dictionary of

Visual Science, 3rd ed. (Chilton, Radnor 1980)
2. H.B. Barlow, R. Fitzhugh, S.W. Kuffler: Dark adapta-

tion, absolute threshold and Purkinje shift in single
units of the eat's retina. J. Physiol. (Lond.) 137, 327
(1957)
3. H.B. Barlow, R. Fitzhugh, S.W. Kuffler: Change of

organization of receptive fields of the eat's retina dur-
ing dark adaptation. J. Physiol. (Lond.) 137, 338 (1957)
4. F .S. Werblin, D.R. Copenhagen: Control of retinal

sensitivity. m. Lateral interactions at the inner plex-
iform layer. J. Gen. Physiol. 63, 88 (1974)
5. G. Westheimer: Spatial interaction in the human reti-

na during scotopic vision. J. Physiol. (Lond.) 181, 881
(1965)
6. G. Westheimer: Lateral inhibition in the human rod

retina. In Performance of the Eye at Low Luminances.
Proceedings of the Colloquium Delft, 1965, ed. by
M.A. Bouman, J.J. Vos (Exerpta Medica, Amsterdam
1966) pp. 53-56.
7. G. Westheimer: Spatial interaction in human cone

vision. J. Physiol. (Lond.) 190, 139 (1967)
8. J.M. Enoch: Quantitative layer-by-Iayer perimetry.

Invest. Ophthalmol. Vis. Sci. 17, 208 (1978)
Quantifying the Magnitude of Visual Impairment
with Multi-Flash Campimetry
Mike Dixon, Edward M. Brussell
1. Introduction
Multi-flash campimetry is a computer implemented clinical

psychophysical technique that uses the ability to detect flick-
er to distinguish between healthy observers and ophthalmolo-
gical patients [1-3]. On each trial, a computer randomly
selects a point from a 36 point display and begins to flicker it
at 5 Hz. The duty cycle of this flickering point (the propor-
tion of the flicker period that is lit), is decreased from 100%
in 1.4% steps each cycle until the observer makes a manual
response indicating that flicker has been detected.
Based on the duty cycles that occurred at the time of the

response, two and three dimensional maps can be created
depicting the temporal resolving power of the visual system
across 40 degrees of the visual field as shown in Fig. 4. For
both types of visual field maps, an interpolation algorithm is
used to estimate the temporal resolving power of areas of the
retina that lie between tested points.
The main benefit of two and three dimensional visual

field maps is the ease with which they can be interpreted. In
the two dimensional display, a seven category grey scale is
used with darker shadings indicating areas of poorer temporal
resolution. Thus patients who have either a general decrease
in temporal resolving power or patients who have highly
localized areas of poor flicker detection capability, will both
be readily distinguished from the field maps of healthy indi-
viduals. Similarly, by presenting such areas of poor temporal
resolution as elevations corresponding to the severity of the
resolution loss, the clinician can immediately visually distin-
guish these "mountains" from "valleys" of normal resolution.
While the ease with which both these types of presenta-

tions can be interpreted is clearly a major benefit, both two
and three dimensional data representations have inherent
67
drawbacks. Three dimensional representations, because of

the problem of occlusion, may require a number of different
viewing orientations to illustrate any collection of data in its
entirety, while two dimensional grey scales rely on the ordi-
nal categorization of data, preventing the portrayal of small
but potentially meaningful differences in score values within
a given category.
If numbers could be provided in conjunction with the

maps, the clinician would be able to both visually discrimi-
nate between the visual fields of patients and healthy observ-
ers, as well as use quantitative methods to increase the power
of this discrimination process.
The base unit which we have adopted to quantify the

degree of visual impairment in patients tested using multi-
flash campimetry was derived from a study investigating two
different luminance presentations and their effect on tempo-
ral resolution. Until very recently the form of flicker
employed in multi-flash campimetry was of a constant pulse
variety in which a light pulse was turned on to a predeter-
mined luminance level and then turned completely off.
Unfortunately, reducing the duty cycle in this type of flicker
also reduces the time-average luminance and subsequently
the Talbot brightness of the point. Thus it could be argued
that patients might base their responses on this reduction of
the apparent brightness of the point rather than on the detec-
tion of flicker.
Such a confound can be easily avoided by holding the

time-average luminance of the point constant. This is
accomplished merely by increasing the intensity of the on-
period in proportion to any reductions in its duration. Fur-
thermore, evidence from early critical fusion frequency
(CFF) literature concerning the effect of duty cycle reduc-
tion on these two types of luminance displays suggests
another advantage of maintaining a constant time-average
luminance. While constant pulse luminance displays reveal
inverted U-shaped temporal resolution functions with the
highest resolvable frequencies at a 50% duty cycle [4], time-
average luminance displays afford monotonically increasing
temporal resolution curves as the duty cycle of the stimulus
is reduced [5]. If these relationships were maintained in
multi-flash campimetry, then holding the pulse luminance
constant would result in flicker becoming easier to see from
duty cycles between 100% and 50%, whereupon further reduc-
68
tions in this parameter would cause flicker detection to

become more difficult. A preferable situation would involve
the implementation of a time-average luminance display in
multi-flash campimetry, whereby the systematic reduction of
duty cycle would elicit a continuous increase in flicker sensi-
tivity. The question of empirical interest, therefore, was
whether the relationships found in CFF studies, where vari-
able frequencies are used to assess flicker detection capabili-
ty, could be extrapolated to multi-flash campimetry, in which
duty cycle is reduced in order to assess temporal resolution at
a fixed frequency of 5 Hz. In order to address this question a
depth of modulation experiment was conducted to test sub-
jects' sensitivity to both time-average and constant pulse
luminance stimuli.
2. Method
The sensitivity of subjects to 7 different duty cycles (20, 30,

40, 50, 60, 70 and 80%) and two different luminance presenta-
tions (time-average and constant pulse luminance) was
assessed using a depth of modulation technique. The stimulus
display consisted of a line of six points spaced at retinal
eccentricities of 0.625, 1.25, 2.5, 5, 10 and 20 degree of visu-
al angle. This line of points appeared on one of 8 meridians,
either the temporal horizontal or rotated by 45, 90, 135, 180,
225, 270 or 315 degrees away from this meridian.
The type of flicker employed was either of a mean con-

stant pulse luminance (MCPL) variety, where the average of
the ~aximum and minimum luminance levels remained at 3.1
cd/m regardless of duty cycle, or alternatively, these lumi-
nance levels were manipulated as a function of duty cycle
such that the time-average lumin~ce (TAL) was maintained
at a steady state level of 3.1 cd/m •
The experiment utilized the following split plot design:

eight meridii X six eccentricites X seven duty cycles X two
luminance types. Subjects were blocked only by meridian;
each subject was exposed to all other treatment combina-
tions. On any given trial, subjects were told to focus on a
central fixation cross and to indicate with a manual paddle
press whether flicker was detected in any of the six presented
points. Points were flickered using a randomly chosen combi-
69
nation of duty cycle, luminance type, and eccentricity. The

minimum depth of modulation required to see flicker for this
stimulus combination was determined using a randomized
staircase procedure with a stopping criterion of four rever-
sals.
Thirty two subjects with corrected or uncorrected acui-

ties between 6/4.5 (20/15) and 6/7.5 (20/25) in their best eye
were used in the experiment. The ages of these subjects
ranged from 20 to 39. All viewing was monocular with sub-
jects wearing corrective lenses for far vision if so required.
Panels A and B of Fig. 1 illustrate the results of this sensitiv-

ity experiment. In order to detect flicker: a) subjects
required greater depths of modulation for peripheral com-
pared to foveal points, and b) within an eccentricity, inverted
U-shaped functions were obtained for both time-average and
constant-pulse displays. Panels A and B of Fig. 2 present the
amplitudes of the fundamental Fourier frequency component
for each of the threshold stimuli that comprise the data in
Fig. 1. A further analysis of this data reveals a significant
effect of eccentricity on the amplitude of the fundamental
F(5,120) = 61.21 p<O.OI (omega squared = 22% of the variance
accounted for). Also, a significant effect of duty cycle was
found F(6,144) = 56.69 p<O.Ol (omega squared = 4% of the
variance accounted for).
The initially surprising discrepancy between the monoton-

ically increasing TAL function obtained when CFF is plotted
over duty cycle and the inverted U-shaped function acquired
for the time-average luminance display in the sensitivity
study may be explained in terms of the amplitude of the fun-
damental Fourier frequency component. Essentially, the idea
is that manipulations in the characteristics of the flickering
stimuli that lead to increases in this amplitude will in tUrn
elicit increases in temporal resolving power. Evidence for
this postulate comes from studies in which decreasing the
duty cycle of a time-average luminance stimulus elicited
monotonic increases in the amplitude of the fundamental and
consequently, monotonic increases in temporal resolution as
measured by CFF [5]. Further evidence comes from similar
70
A B
~
I- M ,62S· 20· TAL , .. '
> 7
.. MCPL-
~
III
6
5
~
." ..
"...
"" "
"I.
:z 4 '"
w
III
.A~.
Z
C
101
~ ~
~
"..
0
203040606070 80 20304060607080
DUTY CYCLE
Figure 1. Log mean sensitivity (l/Threshold Contrast)

over duty cycle for retinal eccentricities of
0.625 and 20 degrees.
studies using constant pulse luminance displays where reduc-

tions in duty cycle caused increases in amplitude, and thus
CFF, between 100% and 50% duty cycles, with further reduc-
tions in duty cycle evoking decrements in both the amplitude
of the fundamental and CFF [4]. H one assumes that the
amplitude of the fundamental would also underlie the detec-
tion of flicker in a sensitivity study, then in such a study,
there should be a constant threshold amplitude above which
subjects would be able to detect flicker.
To test this hypothesis, the amplitudes of the fundamen-

tal were calculated using the sensitivity data portrayed in
Fig. 1. As can be ascertained by looking at the relatively flat
functions depicting amplitude over duty cycle in Fig. 2, this
prediction appears to be confined, for within a given retinal
eccentricity, regardless of the duty cycle of the presented
stimulus, the same amplitude of the fundamental seemed to
be required in order for flicker to be detected. Statistically,
however, there was a main effect of duty cycle on the ampli-
tude of the fundamental indicating that these functions did
not have a slope of zero. Despite such statistical significance
71
ECCI!: NT"ICITY
TAL----- .625"
.4 MCPL-
\OJ
...
x
.3 A
-' .2 ~ ...... [. +····.i...... :....:i
... i!
z .1
0 \OJ
:;:
C
\OJ Q
20·
Q Z
.4
t····· -1-····· i
::l ::l
!:
......
"-
.3 ;-- ··:f···· ...I ...... t~ B
:;: .2 '. I
c(
.1
20 30 40 50 60 70 80
DUTY CYCLE
Figure 2. Amplitude of the fundamental required to

detect flicker plotted over duty cycle for reti-
nal eccentricities of 0.625 and 20 degrees.
the finding that this effect accounts for only 4% of the vari-
ance causes us to conclude that these amplitudes are virtually
the same for all duty cycles, and as such the amplitude of the
fundamental is the principal determinant of flicker detection
in this sensitivity study. Because these amplitudes seem to
underlie the detection of flicker in both CFF studies, as well
as the sensitivity study, it seemed reasonable to postulate
that it would also account for duty cycle detection thresholds
in the multi-flash paradigm. If this postulate is correct then
we should be able to use the amplitude of the fundamental
required to detect flicker in the sensitivity experiment to
predict the performance of subjects in the multi-flash para-
digm. Before we could test this hypothesis, however, it was
necessary to equate the multi-flash procedure with the sensi-
tivity task in terms of the effect of reaction time on the
threshold amplitude of the fundamental.
In the multi-flash procedure, since the duty cycle of the

flickering stimulus is decremented every 200 ms, the obtained
amplitudes that occurred at the time of a manual response
were comprised of two components, the amplitude at thresh-
72
old, and increases in this amplitude due to reaction time. In

order to equate the sensitivity and multi-flash procedures
therefore, it was necessary to evaluate and subtract out
increases in the amplitude of the fundamental due to reaction
time.
To this end, eight subjects who participated in the sensi-

tivity experiment, were administered both the time-average
luminance version of the multi-flash procedure and a single
quadrant of the multi-flash display which was used to meas-
ure reaction time. In the latter display, rather than system-
atically reducing the duty cycle of flicker only threshold duty
cycles were presented, and based on the average of six repli-
cations for each of 30 points, reaction times were determined
for each of the six retinal eccentricities common to both the
sensitivity study and the multi-flash procedure. Increases in
the amplitudes of the fundamental due to reaction time were
then calculated using these average reaction times, and these
reaction time induced increases were then subtracted from
the obtained multi-flash amplitudes. Figure 4 compares these
corrected multi-flash amplitudes to the amplitudes required
by the same subjects in the sensitivity experiment. Overlap
among the standard error bars indicate that differences
between these amplitudes within a given eccentricity are due
to chance. Because one can use the amplitude of the funda-
mental to predict the performance of subjects on the multi-
flash paradigm based on their performance in the sensitivity
experiment, it seems that this amplitude underlies the detec-
tion of flicker in both procedures.
73
.40
.38
~ .36
W· 34
~ .32
!i::> .30
I.L .28
- OBTAINED SENSITIVITY VALUES

- - - CORRECTED MULTI-FLASH THRESHOLDS
2.5 10 20
RETINAL ECCENTRICITY
Figure 3. Amplitudes of the fundamental required to

detect flicker in sensitivity experiment and in
multi-flash campimetry for six retinal eccen-
tricities. Multi-flash amplitudes are corrected
for spurious increases due to reaction time.
4. Discussion
The reliance of flicker detection on the amplitude of the fun-

damental in the CFF paradigms, as well as the sensitivity and
multi-flash paradigms suggests that regardless of the task
used to assess temporal resolution, this single measure may
be used to quantify the magnitude of visual impairment
reflected by any observed losses in temporal resolving power.
The equivalence of the amplitudes required by normals in
multi...flash and in the sensitivity study suggests that this
amplitude could be used to compare healthy individuals' per-
formance to that of patients, both between different para-
digms, or within a particular design. For our present purpos-
es, we have chosen to numerically represent the severity of
any visual deficits that are illustrated using the multi-flash
procedure by comparing the amplitude of the fundamental
required by a patient in multi-flash to that required by a nor-
mal using this same temporal resolution technique.
74
In order to ascertain the best estimate of the normal

amplitudes of the fundamental that are required to detect
flicker the amplitudes required by the eight subjects tested
using the time-average luminance version of multi-flash were
averaged to form a "control map". This control map was then
used as a reference for evaluating patient performance in this
task.
Figure 4 depicts the visual field maps of a healthy

observer, a patient with anisometropic amblyopia and a
patient with strabismic amblyopia. Presented along with
these two and three dimensional representations are five sta-
tistics which serve to numerically summarize these data.
B c
CONTROL AMBLYOPE
STRABISMUS
A.0.=-.016 A.0.=-.02 A.0.=.160 A.0.=.155
L.0.=.182 L.0.=.188 A.0.=.264 A.D.=.295
L.0.=.089 L.0.=.107 L.0.=.333 L.D.=.342
151.= 9.0 151.= 7.0
151.= 7.0 151.= 9.0
Sev.= 49% Sev.= 557.
Sev.= 67% Sev.= 691
Ar. = .5% Ar. =.657.
Ar. =3.8% Ar. =3.9%
Figure 4. Average Deficit (A.D.), Local Deficit (L.D.),

Islands (IsI.), Average Severity (Sev.), and Aver-
age Area (Ar.) statistics, along with two and
three dimensional field maps for a control sub-
ject, an anisometropic amblyope, and a stra-
bismic amblyope.
The first two statistics, the Average Deficit and the

Local Deficit, draw on the work of FLAMMER et al.[6]. The
Average Deficit, as the name suggests, reflects any overall
75
increase in the amplitude of the fundamental required to

detect flicker in the Multi-flash paradigm. It is calculated by
taking the amplitude required for each point on the patients
map and subtracting the corresponding amplitude from the
control map. These differences are summed and divided by
the number of points (120) tested. As can be seen in Fig. 4
the value of the average deficit for a healthy observer is near
zero, but for the two patients is elevated by amounts corre-
sponding to the severity of their condition.
The second statistic, the Local Deficit, is a measure of

dispersion around this Average Deficit. Patients with local-
ized areas of dysfunction in an otherwise normal field would
show high variability around the Average Deficit, and there-
fore display high Local Deficit values. A patient with a uni-
form elevation in required amplitudes spread across the visual
field, on the other hand, would show little variability around
their obtained Average Deficit, thus affording low Local Def-
icit values. The strong localized nature of deficit in the two
patients in Fig. 4 is illustrated by the elevation of this statis-
tic. Interestingly, even normals show some elevation in Local
Deficit because of the irregular placement of patches of
reduced sensitivity beyond 1.25 degree of visual angle.
Despite the irregularity of the location of such areas between
subjects, the test-retest reliability of the technique was 0.B7
when the subject depicted in panel A was given two adminis-
trations of this test.
The second strategy for quantifying visual impairment

makes use of statistics that can be said to more accurately
reflect what is visually depicted by the two-dimensional visu-
al field maps. Unlike the Average and Local Deficit statis-
tics, this latter category of measures makes use of values
arrived at through the previously mentioned interpolation
algorithm, and as such are subject to the same constraints as
the maps themselves.
The "Islands" statistic reflects the number of areas in the

visual field in which higher amplitudes of the fundamental
than normal are required to detect flicker. The criterion
used to determine whether an amplitude is to be considered
abnormal involves the grey scale categories used in the two
dimensional maps. Such grey scales are composed of ranges
of amplitudes. Looking at the healthy observer in Panel A of
Fig. 4, patches of reduced temporal resolving power are noted
throughout areas of the visual field beyond 1.25 degrees.
76
Because all healthy observers tested using multi-flash campi-

metry displayed such patches, they are considered to be part
of a normal visual field. Because these patches are found only
beyond 1.25 degrees (the second circle on the map), areas
associated with this shade of grey that occur within 1.25
degrees of the fovea would be considered an Island of dys-
function. Beyond 1.25 degrees only areas reflecting ampli-
tudes of the fundamental that are more severe than those
associated with this shade of grey qualify as an Island. An
example of a foveal Island can be found in the left eye of the
strabismic amblyope, while parafoveal Islands are dispersed
throughout the four fields of both patients.
The Average Severity statistic indicates how severe the

deficiency is within a typical Island for a given map. Since
the different shades of grey represent seven different ranges
of amplitudes, an Average Severity measure can be obtained
by taking all the points that are above normal amplitude val-
ues, assigning these points values equal to the midpoint of
their respective range, summing these midpoint values and
dividing by the total number of points. For ease of interpre-
tation, the resulting value is then expressed as a percentage
of the maximum possible severity. The correspondence
between what is visually depicted by the maps and this aver-
age severity statistic is attested to by the fact that the stra-
bismic patient has higher Average Severity values than the
anisometropic amblyope.
Finally, the Average Area statistic reflects the average

size of these Islands of deficit. This statistic is calculated by
summing the number of sampled and interpolated points that
have abnormal amplitudes and dividing by the number of
Islands. Once again for clarity of interpretation this statistic
is expressed as a percentage of total map area in order to
provide an upper limit as a point of reference. Panels Band
C of Fig. 4 indicate that the Islands of dysfunction in the
strabismic map are larger than those found on the anisome-
tropic amblyope.
To summarize, the Average Deficit and Local Deficit can

be used to determine whether a patient has visual field defi-
ciencies and whether such deficits involve local areas of
reduced temporal resolving power, or are uniformly spread
across the visual field. The Island, Average Severity and
Average Area statistics serve to both corroborate these sta-
tistics and in so doing, numerically depict what is portrayed
77
visually by the two and three dimensional maps. By using the

amplitude of the fundamental as a measurement of visual
deficit, we feel that we have chosen a measure that will
enable researchers to predict performance on any temporal
resolution task. As such, it reflects the functional capability
of an integral part of the visual system, and will therefore
enable the clinician to accurately distinguish between a
healthy individual and a patient afflicted with one of the
many ocular pathologies that affect the human visual system.
78
5. References
1. E.M. Brussell, C.W. White, M. Bross, P. Mustillo, M.

Borenstein: Multi-flash campimetry in multiple scle-
rosis. Curro Eye Res. 1,671 (l98l/2)
2. O. Overbury, E.M. Brussell, C.W. White, W. Jackson,

D. Anderson: Evaluating visual loss with multi-flash
campimetry. Can. J. Ophthalmol. 19, 255 {l984}
3. E.M. Brussell, C.W. White, J. Faubert, M. Dixon, G.A.

Balazsi, O. Overbury: Multi-flash campimetry as an
indicator of visual field loss in glaucoma. Am. J.
Optom. Physiol. Opt. 63, 32 {l986}
4. R.T. Ross: The fusion frequency in different areas of

the visual field: IV. Fusion frequency as a function of
the light-dark ratio. J. Gen. Psychol. 29, 129 (l943)
5. H.E. Ives: Critical frequency relations in scotopic

vision. J. Opt. Soc. Am. 6, 254 (l922)
6. J. Flammer, S.M. Drance, L. Augustiny, A. Funkhous-

er: Quantification of glaucomatous visual field defects
with automated perimetry. Invest. OphthalmoZ. Vis.
Sci. 26,176 {l985}
Spatial vs. Temporal Information in Suspected
and Confirmed Chronic Open Angle Glaucoma
Jocelyn Faubert, Edward M. Brussell, Olga Overbury, A. Gor-

don Balazsi, Mike Dixon
1. Introduction
Vision care professionals are exhibiting a growing interest in

psychophysical testing. In the case of glaucoma, this may be
partially due to the fact that the diagnosis and treatment can
depend upon the results of these tests. For this reason,
researchers have been exploring a variety of psychophysical
techniques in order to determine visual loss at the earliest
stages. The use of automated perimeters, for instance, has
demonstrated that glaucomatous eye damage may be present
before previously realized [1,2]. Other than light sensitivity
thresholds, spatial and temporal aspects of vision have been
found to be abnormal in glaucoma patients and suspects as
well. A number of investigators have reported that contrast
sensitivity, especially at low and medium frequencies, can be
abnormal in glaucoma patients and suspects [3-S]. In the
1940s and 1950s [9-11], and more recently [12-14], loss of
temporal resolving power and sensitivity has also been report-
ed. Given that both spatial and temporal losses are observed
in glaucoma, the question remains as to whether the same
neural mechanisms underlie each.
Neurophysiological findings [15-1S] have suggested that

the mammalian visual system consists of at least two neuron-
al mechanisms served by X (sustained) and Y (transient) type
ganglion cells. The X cells are more sensitive to higher spa-
tial frequencies, and the Y cells to time related information.
These latter cells have larger axons and receptive field areas.
QUIGLEY and his colleagues [19] have recently reported a
grea ter loss of large and medium sized fibers in the optic
nerve of monkeys in whom glaucoma was induced experimen-
tally. Since this type of tissue loss is consistent with one
interpretation of the psychophysical findings, we felt that it
was important to collect spatial and temporal data from the
same glaucoma patients and suspects so that they could be
compared directly.
80
Because there is evidence that the processing of spatial

and temporal information is interrelated [20,21], and there
can be response criterion problems when assessing both
dimensions with grating stimuli [22], it was decided to use
distinct psychophysical techniques. Temporal resolving power
was measured with multi-flash campimetry, a rapid, comput-
er implemented technique, that produces data indicating the
duty cycle (the proportion of a flicker cycle that is lit) neces-
sary to detect 5 Hz flicker. Multi-flash campimetry [23,24]
has been shown to be sensitive to visual loss due to multiple
sclerosis, optic neuritis not due to demyelination, macular
degeneration, amblyopia, as well as glaucoma [23-25,12,13].
It verifies 120 points across a 40 degree visual field, allowing
a fine probing of the central visual field. Once the data is
collected, two and three dimensional maps are printed for
easy and rapid evaluation. Spatial contrast sensitivity was
assessed with a methodology known as the "Anticipated
Threshold Technique" [26]. It is an algorithm that allows the
testing of six spatial frequencies in less than ten minutes.
2. Methods
The technical and procedural details of multi-flash campime-

try [23,24,12] and the "Anticipated Threshold Technique" [26]
have been described elsewhere. However, the essence of
each will be outlined below.
Multi-flash campimetry entails the presentation of six

concentric circles, each composed of 20 points. The radius of
the innermost circle subtends 0.625 degrees of visual angle,
with the length of each succeeding radius doubling such that
the largest radius subtends 20 degrees. A computer randomly
selects one quadrant of the display and tests each point with-
in a quadrant in a random order. A point is flickered at 5 Hz
with its duty cycle decreased from 100% in steps of no more
than 2%, until the detection of flicker is indicated with a
manual paddle press response. After 120 points are tested in
each eye, points that exhibited statistically deviant data are
retested to ensure that any abnormality is due to a loss in
temporal resolving power.
The "Anticipated Threshold Technique" [26] is a computer

implemented algorithm in which the contrast of a grating is
81
raised from zero in two stages. During the first half of a tri-
al, whose duration can be approximately predetermined, the
contrast rises rapidly to a point just below an anticipated
threshold, and then rises slowly as the threshold is crossed.
On succeeding trials the anticipated threshold is modified
based upon previous responses as the algorithm attempts to
optimize the efficiency with which the threshold is reached.
If the parameters are appropriately set, it takes about one
minute to obtain seven estimates of the threshold for a given
spatial frequency.
Both the assessment of contrast sensitivity and temporal

resolving power were implemented with the use of a
PDPll/IO computer interfaced with a large screen CRT
(Hewlett-Packard 1310A equipped with a PIS phosphor) either
through digital to analog converters, in the case of multi-
flash campimetry, or through function generators in the case
of contrast sensitivit!. The lum~nce level for the multi-
flash test was 3 cd/m ,and 5 cd/m for the contrast sensitiv-
ity test. Sine wave gratings were presented within an 8 X 8
degree window. Both tests were presented in the same ses-
sion and in a random order across observers.
Twenty nine eyes of 16 patients with either glaucoma or

suspected of having glaucoma, were used in the study. Nine-
teen of the eyes were "suspect" eyes, and 10 were glaucoma-
tous. For our younger age group (20 to 35 with a mean of
28.6), the data were compared with that collected from 40
healthy eyes in that age category. For the older age group
(over 50 with a mean age of 64 years), we also tested 12
healthy eyes with a mean age of 57.
Glaucoma patients were distinguished from suspects by

having either I} early Goldmann visual field defects as meas-
ured with the Armaly-Drance technique [27], widely used in
clinical settings for rapid screening of glaucoma, or 2} disc
cup abnormalities (e.g. disc cup asymmetry). All patients had
intraocular pressures of 21 mmHg or greater during several
visits and corrected visual acuity of 6/7.5 (20/25) or better.
All viewing was monocular with an observer asked to wear
his/her best correction for far vision.
82
3. Results
Figure 1 presents the contrast sensitivity data for the individ-

uals between the ages of 20 and 35. The hatched area repre-
sents a normal range that is two standard errors wide. The
most notable aspect of these data is the low and middle spa-
tial frequency loss exhibited by the patients and suspects;
contrast sensitivity for the high frequencies fall within or
near the normal range. Figure 2 presents the multi-flash data
for the same patients. A multi-flash threshold is defined as
the duration of the off-period within a flicker cycle that
existed at the time of a response. Again, the hatched area
represents a two standard error normal range. As can be
seen, data from both the patients and suspects fell outside
the normal range, with patient data exhibiting the higher
thresholds.
Contrast sensitivity data from the older age group is pre-

sented in Fig. 3. For the sample of eyes that we tested, con-
trast sensitivity fell within the normal range. Although we do
not consider this typical for glaucoma, it is an interesting
finding. The reason is that as illustrated in Fig. 4, these indi-
viduals with normal contrast sensitivity, show abnormal tem-
poral resolving power.
Table 1 presents, separately for the patients and

suspects, Pearson correlation coefficients for data from both
tests. Consider the correlations from the patients. The
strongest relationship, negative because of the inverse rela-
tionship between sensitivity and threshold, is between the
data from the lowest three spatial frequencies and the multi-
flash thresholds from the central 10 degrees of the visual
field. With the exception of the data from the lowest three
spatial frequencies and all of the multi-flash thresholds, none
of the other correlations are significant. The same general
pattern, although weaker, emerges from the correlations for
the suspects.
83
GROUP AVERAGES (20 TO 35)
I
I
§
1.00 1.82 3 31 6.03 10.99

SPATIAl. F&EIf:Y W
... '!IISPECT' X EMU U .
t tOIIAI. +1SE D tOIIAI. -lSE

Figure 1. Average log contrast sensitivity for the glauco-
ma and glaucoma suspect eyes as a function of
spatial frequency for the group between 20 and
35 years of age. The crosshatched section rep-
resents a 2 standard error normal range.
84
GROUP AVERAGES (20 TO 35)

200
I i
i
III
!Et-
128
itO
~ 92
't
H 74
a 58
38 :t'tt,~"lI1
20
0.6 1.3 2.5 5.0 10.0 20.0
RETINAL ECCENTRICIT1
+ 'SUSPECTS' X 1iL.4UCOMA
• NORMAL +iSE o NORMAL -1E
Figure 2. Average multi-flash thresholds for the glauco-

retinal eccentricity for the group between 20
and 35 years of age. The crosshatched section
represents a 2 standard error normal range.
85
GROUP AVERAGES (OVER 50)
I.
I 1.
I o.
o.
o.
o.
1.00 1.82 3.31 6.03 10.99

SPATIAL FIDEICY tAlI
... 'IISPECT' X EMU I..AImi
f IIIIIAI. t~ D IIIIIAI.-~
Figure 3. Average log contrast sensitivity for the glauco-

spatial frequency for the group over 50 years of
age. The crosshatched section represents a 2
standard error normal range.
86
GROUP AVERAGES (OVER 50)

200
182
lB4
i
~
H
I-
74
i
56
3B
20
O.B 1.3 2.5 5.0 10.0 20.0
RETINAL ECCENTRICITY
+ ·SUSPECTS· X EAR..Y a AIICOMA
• NOAMAl. +ISE o NORMAl. -ISE
Figure 4. Average multi-flash thresholds for the glauco-

retinal eccentricity for the group over 50 years
of age. The crosshatched section represents a 2
standard error normal range.
87
Table 1
Correlations between multi-flash campimetry and contrast

sensitivity data.
GlAUC(}!A
MULTI-FLASH THRESHOLDS
SPATIAL FREQ. I CENTRAL 10 DEGREES BEYOND 10 FULL 40 DEGREES
L(lJER 3 FREQ. I -.71 P' .05 -.34 -.67 P' .05

I
HIGHEST 2 FREQ.I -.19 .24 -.07
I
ALL FREQ. -.51 -.05 -.41
GLAUCCl1A SUSPECTS
MULTI-FLASH THRESHOLDS
SPATIAL FREQ. I CENTRAL 10 DEGREES BEYOND 10 FULL 40 DEGREES
L(lJER 3 FREQ. I -.40 p' .09 -.37 -.45 p' .06

I
HIGHEST 2 FREQ. I -.28 -.19 -.29
I
ALL FREQ. I -.36 -.30 -.39
4. Discussion
The finding that the multi-flash data correlates with contrast

sensitivity for low and medium spatial frequencies suggests
that the two tests may be tapping into the same neural
mechanism. In conjunction with Quigley's neuroanatomical
finding that large optic nerve fibers are affected earlier by
increased intraocular pressures than small fibers [19], it
might be concluded that the Y system represents the common
neural substrate [28]. However, the data from the older
group in which normal contrast sensitivity was accompanied
by abnormal temporal resolving power, suggests that each
test may be sensitive to nonoverlapping neural popUlations as
well. The data from the older group also implies that multi-
flash campimetry may be the more sensitive of the two tech-
niques.
Perhaps this latter conclusion is not surprising in light of

the fact that multi-flash campimetry allows a fine grained
probing of the visual field, whereas contrast sensitivity,
88
depending upon the psychophysical detection model assumed,

may reflect general functioning over the entire area being
tested. If further research demonstrates that multi-flash
campimetry is the more sensitive of the two tests, the ques-
tion remains as to how it compares to more conventional
types of perimetry.
Figure 5 presents two dimensional (2D) and three dimen-

sional (3D) multi-flash field maps for both eyes of two normal
observers (a 27 year old female, and a 58 year old female).
For both types of maps, retinal eccentricity is plotted on an
octave scale, such that equal spacing represents a doubling of
retinal distance. This tends to exaggerate the size of central
visual field, compared to the peripheral visual field, which
corresponds to its functional importance and representation
in the cortex. In the 2D and 3D fields, darker shadings and
higher mountains, respectively, are indicative of worse tem-
poral resolution. What characterizes these 2D and 3D fields
as normal is their relative lightness and flatness, respective-
ly.
Figure 6 presents multi-flash and Octopus fields from a

55 year old glaucomatous left eye and suspect right eye. (All
Octopus fields were generated from program =11'32 on model
2000R, '*38 on model SOOE.) Although the Octopus fields con-
firm the diagnosis, the multi-flash fields suggest a severe loss
of temporal resolving power in the right eye as well. Figures
7 and 8 present multi-flash and Octopus fields from 33 year
old and 63 year old glaucoma suspects, respectively.
Although the Octopus fields are quite normal in both cases,
the multi-flash fields are not.
These sets of data illustrate that while abnormal Octopus

fields are associated with abnormal multi-flash fields, there
are numerous cases of normal Octopus fields being associated
with abnormal multi-flash fields. In some sense this is to be
expected because temporal resolving power correlates with
light sensitivity, implying that abnormal multi-flash fields
should be observed when abnormal Octopus fields are
observed. However, multi-flash campimetry should also
reveal losses in temporal resolving power that are not accom-
panied by an attenuation of relative light sensitivity, and this
seems to be the case in many of the glaucoma suspects we
have tested. The overall extent to which multi-flash campi-
metry will be predictive of glaucomatous optic atrophy will
be determined through our ongoing longitudinal study of glau-
coma patients and suspects.
89
"'-....
. ~.
' .
.-.
III .... . -. . ~ ...
Figure 5. Two and 3D multi-flash maps for a 27-year-old

normal observer and a 58-year-old normal
observer. Temporal resolution loss is represent-
ed by darker shading and higher mountains
respectively.
90
'" ....
.... ..
........
I, •.'" t ••••••
..
..
..... . ::
f', •
:~:
" :: : . ... ...
: :.:.::.' ...
............. .......
.... , .....
:.; dink!!!!!!!)!>: .... ....

........
" "" ,
......
... .... ... .....
...
........ ... ..
...........
18 I" L': , "I ,~ I
11 I 0:; • I
~ 21 '04 I 1 ., fl ') I 113 • 2 --- .., 1 ,'1. li I

• I 2 1 1 .1 I ~r;,. L I' l .zl t .. ":I: " ' .., 1 I'"J
:1 ... a " ." '·1

J • .... .. ..~
z.. . . ~ i'~':t ... I • .?~ ~ I ~"i
I.:n A .. t
I ,
Figure 6. Two and 3D multi-flash maps and Octopus

fields represented by a grey scale and numeric
values for a 55-year-old glaucoma patient in the
left eye and glaucoma suspect in the right eye.
91
.... - ... .. .... ... .

-: ~.;;; :.:: : ...... ::.
..•.. .... . .
.:: !!:: :-............. . .. .. --..
;~ ~ iii ... !
ii i!....iii...iii..iii...iii
...........
i ".I@ . . . ........ . .
':G 21 5~ 2:t
-'5~dT~~'"
~ .. Z9 2" 10 10 0
Z'" 1 " 30. 2: 5 .28 2.
'I:~ 1'JO 1.1 .z~ ,p JO
'. . . , -, ~O "1? 'l l :!? "0 .l ~ ';:IIi.
w'" ,:, :lo ::1 3l ; 1 !O ],0 ':9 :1
-Q _. l oi) n 1,2 12 lol JI ~9 ~,
w"" ;0 'OJ. ; 1 1'3 ;: ')I r ,:, ... ':Q
:'l ':" .) :n 12 1 .. 3 1 33 30 :~ n
n H 1I
~: 30 ~~ JJ 11 l~ ~ oZ l ~ J.
J~ J2 10 l~ H 29 , . :!:' 10 )0 10 :U 2; 311 21 Z9 18
2:, ?3 1 1 1 1 30 12 l8 2'
1" 19 18 13 2& ~9 ~i 27
29 3J 29 2' 38 lB
.:~ Jl ~t 10
Figure 7. Two and 3D multi-flash maps and Octopus

fields represented by a grey scale and numeric
values for a 33-year-old glaucoma suspect in
both eyes.
92
...
...
if:
,.,
...............
...
, IQ
" H
21 t o;"
23
~I ~J
~.) -v
~J ~4 .?2 ~~ 14 ~ I 2 4 2 1 21 ~5 25 ~.1
jJ
\'
_(J =1
23 103 14 Z3 2 -
.1 4 2~ 1·' .::'!; 2 1
1} ~5 ::;2 ~3 .22
~
~S
1 24 :!) 15 _4 24
2'5 ~ 13 .:~ 11
-- ,-
,~
:! ~ 15 _$
,- -,
-~
1~ ~4 .?1 .!~ ':0 2'!i .<5 I '~ .?o ~~ ~~ .::.;. .?3 '::6 ::-.; 13 2' .l~
,-
:: t : 1 ~ ~1 2' :tl), 15 : 4 ~3 :1
- "> =?~ ~~
,-
~ .;. 14 2';' 2i>
..
0 1~
:l
.,:' ~ 1 ':: 4 1 1 Z1 ~-;
'" ..!:~ 23 2e- l':1 24 24 1'5 >-
2;;
~
.?i. ~5 1.; ::5
"" -"
;?j 21 ':4 2. 25 2'30 2~ 24 ~J 1'5 .?6 25 25 , ~
20
.:..a ::1 2'5 24 '?'5 --', ~~ l~ 2'5 2b 2 - .2$
~-
.0 :"; 1· ' ::'3 ,? 4 .::~ 2:!
Figure 8. Two-dimensional multi-flash maps and Octopus

fields of a 63-year-old glaucoma suspect in both
eyes.
93
5. References
1. J.L. Anctil, D.R. Anderson: Early foveal involvement

and generalized depression of the visual field in glau-
coma. Arch. Ophthalmol. 102, 363 (1984)
2. J. Flammer, S.M. Drance, M. Zulauf: Differential light

threshold: Short- and long-term fluctuation in patients
with glaucoma, normal controls, and patients with
suspected glaucoma. Arch. Ophthalmol. 102, 704
(1984)
3. G.B. Arden, J.J. Jacobson: A simple grating test for

contrast sensitivity: Preliminary results indicate value
in screening for glaucoma. Invest. Ophthalmol. Vis.
Sci. 17, 23 (1978)
4. A. Atkin, I. Badis-Wollner, M.D. Wolkenstein, A. Moss,

S.M. Pados: Abnormalities of central contrast sensi-
tivity in glaucoma. Am. J. Ophthalmol. 88, 205 (1979)
5. A. Atkin, M. Wolkstein, I. Bodis-Wollner, M. Anders, B.

Kels, S.M. Podos: Intraocular comparison of contrast
sensitivities in glaucoma patients and suspects. Br. J.
Ophthalmol. 64, 858 (1980)
6. J. Faubert, O. Overbury, G.A. Balazsi, E.M. Brussell:

Contrast sensitivity and Goldmann visual fields in
glaucoma. Invest. Ophthalmol. Vis. Sci. (Suppl.) 26,
217 (1985)
7. R.A. Hitchings, D.J. Powell, G.B. Arden, R. M. Carter:

Contrast sensitivity gratings in glaucoma family
screening. Br. J. Ophthalmol. 65, 515 (1981)
8. J.E. Ross, A.J. Bron, B.C. Reeves, P.G. Emmerson:

Detection of optic nerve damage in ocular hyperten-
sion. Br. J. Ophthalmol. 69, 897 (1985)
9. R. Weekers: L'exploration des fonctions visuelles en

clinique par la mesure de la frequence critique de
fusion. Bull. Soc. Fr. D'Ophtalmol. 69, 331 (1947)
10. P.W. Miles: Flicker fusion fields. Arch. Ophthalmol.

43, 661 (1950)
94
l1. C.J. Campbell, M.C. Rittler: The diagnostic value of

flicker perimetry in chronic simple glaucoma. Trans.
Am. Acad. Ophthalmol. Otolaryngol. 63, 89 (1959)
12. E.M. Brussell, C.W. White, J. Faubert, M. Dixon, G.A.

Balazsi, O. Overbury: Multi-flash campimetry as an
indicator of visual field loss in glaucoma. Am. J.
Optom. Physiol. Opt. 63, 32 (1986)
13. J. Faubert, O. Overbury, G.A. Balazsi, E.M. Brussell:

Multi-flash campimetry and Octopus perimetry in
chronic open angle glaucoma. Invest. Ophthalmol. Vis.
Sci. (Suppl.) 27, 109 (1986)
14. C.W. Tyler: Specific deficits of flicker sensitivity in

glaucoma and ocular hypertension. Invest. Ophthal-
mol. Vis. Sci. 20, 204 (1981)
15. C. Enroth-Cugell, J.G. Robson: The contrast sensitivi-

ty of retinal ganglion cells of the cat. J. Physiol.
(Lond.) 187, 517 (1966)
16. B.G. Cleland, M.W. Dubin, W.R. Levick: Sustained and

transient neurones in the eat's retina and lateral geni-
culate nucleus. J. Physiol. (Lond.) 217, 473 (1971)
17. H. Ikeda, M.J. Wright: Receptive field organization of

"sustained" and "transient" retinal ganglion cells which
subserve different functional roles. J. Physiol. (Lond.)
227, 769 (1972)
18. R. Blake, A. DiGianfilippo: Spatial vision in cats with

selective neuronal deficits. J. Neurophysiol. 43, l197
(1980)
19. H.A. Quigley, G.R.Dunkelberger, R.M. Sanchez:

Chronic experimental glaucoma causes selectively
greater loss of larger optic nerve fibers. Invest.
Ophthalmol. Vis. Sci. (Suppl.) 27, 42 (1986)
20. D.H. Kelly: Motion and vision. II. Stabilized spatio-

temporal threshold surface. J. Opt. Soc. Am. 69, 1340
(1979)
21. E.M. Brussell, C.W. White, P. Mustillo, O. Overbury:

Inferences about mechanisms that mediate pattern and
95
flicker sensitivities. Percept. Psychophys. 35, 301

(1984)
22. C.A. Burbeck: Criterion-free pattern and flicker

threshold. J. Opt. Soc. Am. 71, 1343 (1981)
23. E.M. Brussell, C.W. White, M. Bross, P. Mustillo, M.

Borenstein: Multi-flash campimetry in mUltiple scle-
rosis. Curro Eye Res. 1, 671 (1981/2)
24. C.W. White, E.M. Brussell, O. Overbury, P. Mustillo:

Assessment of temporal resolution in multiple sclerosis
by multi flash campimetry. In Advances in Diagnostic
Visual Optics, 2nd International Symposium on Advanc-
es in Diagnostic Visual Optics at the International
Congress of Ophthalmology, Tuscon, 1982, ed. by G.M.
Breinin, I.M. Siegel, Springer Series in Optical Scienc-
es, Vol. 41 (Springer, Berlin, Heidelberg 1983) pp.
239-46.
25. O. Overbury, E.M. Brussell, C.W. White, W.B. Jackson,

P. Anderson: Evaluating visual loss with multi-flash
campimetry. Can. J. Ophthalmol. 19, 255 (l984)
26. E.M. Brussell, P. Cavanagh: An anticipated threshold

technique for measuring contrast sensitivity. Am. J.
Optom. Physiol. Opt. 61, 125 (l984)
27. S.M. Drance, P. Brais, M. Fairclough, J. Bryett: A

screening method for temporal visual defects in chron-
ic simple glaucoma. Can. J. OphthaZmol. 7, 428 (l972)
28. D. Regan, D. Neima: Balance between pattern and

flicker sensitivities in the visual fields of ophthalmolo-
gical patients. Br. J. Ophthalmol. 68, 310 (l984)
Low Vision Management in Selected Eye Diseases
Eleanor E. Faye
1. Introduction
One of the greatest challenges in doing low vision work is the

opportunity to combine an interest in medicine with rehabili-
tation, psychology, optics and psychophysics. In this paper,
some of these interests are combined in discussing the effect
on patient management of the eye diagnosis.
2. Functional Definition of Low Vision
Low vision should properly be defined functionally: impaired

visual function in individuals with damage to the visual sys-
tem. Neither visual acuity nor visual field should be used to
describe function.
The healthy eye and brain do a remarkable job of inte-

grating their many functions: detailed acuity, peripheral field
awareness, spatial judgements, color perception, contrast
judgements, light and dark adaptation, protecting against
glare by pupillary reactions and eyelid contraction.
In low vision one or a combination of these functions are

affected by disease and each function must be tested, evalu-
ated and compared to a norm. This results in a profile of
anticipated performance which is unique for each person.
However, in the clinical evaluation of patients one soon

realizes that eye diseases seem to have features in common
that influence plans for rehabilitation. These conditions fall
naturally into three general categories based on visual field
description: diseases that produce a field defect, those with
central field defects, and those diseases that result in periph-
eral field constriction (F AYE, [I D.
97
Conditions that do not cause field defects can be further

divided into diseases that affect contrast and those that
affect image resolution. Each group has different visual
symptoms.
Contrast reduction. Opacities of the cornea, lens and

vitreous adversely affect the image-forming quality of the
eye. When light rays are deflected by opacities, the random
scatter results in a reduction of contrast perceived by the
patient as hazy, blurred vision. Constricted pupils further
reduce total luminance reaching the retina, while dilated
pupils increase glare.
Patients, when requested to describe symptoms, complain

of diffuse blur and glare sensitivity for both distance and near
vision. It is important in this group of conditions to elicit dif-
ficulties patients have that are related to contrast. The
extensive list of difficulties includes reduced vision in twi-
light and on bright hazy days, navigating curbs, uneven pave-
ment, going down a flight of stairs, seeing signs and print,
loss of texture perception, and most of all a fear of unexpect-
ed hazards in travelling around.
These answers are significant when combined with the

results of function tests, particularly acuity and contrast.
3. Function Tests
The tests in common use are contrast sensitivity acuity and

grating test interferometry Potential Acuity Meter and glare
tester.
Visual acuity charts in three percentages of contrast have

been developed providing letters of 95%, 9% and 3% (REGAN
and NEIMAN, [2]). The chart is designed to be viewed at 10
feet. A simple grating test developed for clinical use
(GINSBURG, [3]) provides 5 rows of targets which cover a
range of 1.5 cycles per degree (c/deg) to 18 c/deg providing 3
orientations and contrast from 45% to 0.5%. The distance
test is presented at 3 m for normally sighted persons and 1 m
for low vision patients with adjustment of the c/deg at this
distance of 0.5 c/deg to 6 c/deg. The near test can be suc-
cessfully presented to low vision patients at the suggested
98
test distance of 45 cm to provide comparable data. Contrast

tests do not replace Snellen acuity or reading acuity tests,
but provide additional valuable information on the individual's
range of response to low, middle and high spatial frequencies.
Contrast tests are useful to follow the progress of diseases in
pre-retinal structures, particularly cataracts, keeping in mind
that these tests do not differentiate between optical and
neural disorders without further tests.
Case 1. A patient with posterior subcapsular cataracts

and 6/12 (20/40) visual acuity in each eye complained of diffi-
culty driving at night against highway lights and headlights.
The right eye showed mid and high frequency loss on contrast
testing. Following cataract extraction, the patient's curves
were in the high normal range.
4. Treatment Considerations
Surgery must always be ranked as the most important treat-

ment consideration. Cornea, lens, iris and vitreous may be
replaced or altered in most cases. However, surgery may be
delayed for many reasons including medical problems, fear
and poor results in the fellow eye. Visual rehabilitation
should always be considered as an alternative.
Refraction and contact lenses should always be tried,

particularly since contrast sensitivity function (CSF) is
refraction sensitive.
Tinted or antireflective lenses may improve acuity or

contrast and reduce luminance.
It should be emphasized that magnification is not always

appropriate for persons with pre-retinal eye disorders. Since
the retinal image is already degraded, the further reduction
of contrast by the optical device defeats the purpose of mag-
nification. In these situations, we rely more on non-optical
aids such as large bold print, writing aids, typoscopes, yellow
acetate filters, and simple glare reducing devices such as
wide brimmed hats or visors.
Correct illumination considers not only appropriate lumi-

nance but reduction of surface reflection and selection of a
99
broad band spectrum that enhances contrast; for example, the

emphasis on the yellow band of the spectrum in incandescent
bulbs may be superior to the blue fluorescent lights. The
newer low wattage daylight spectrum fluorescent bulbs should
be investigated for use in low vision. Neodymium 60 watt and
100 watt bulbs provide a daylight spectrum with reduced yel-
low component and may have limited use in pre-retinal disor-
ders.
Lenses to control light are numerous: tinted lenses

including gradient tjfts, uv ~bitors, ph~ochromic lenses
including Photogray ,Photosun , and CPF 511, polarizing
lenses and antireflective coatings (mirror, single and multi-
layer coatings). One of the latest coatings on the commercial
market is a multilayer antireflective coating which has been
particularly valuable in high myopia when the corrective lens-
es have plano front curves.
Unfortunately the selection of the best tinted or reflec-

tive lens is still dependent on patient preference under actual
outdoor conditions.
5. Image Resolution
This second group with reduced image resolution is included

under conditions that do not produce a field defect. It is a
small but significant group whose most prevalent conditions
are macular dysplasia in all types of albinism, amblyopias,
and macula edemas (cystoid, diabetic retinopathy).
It should be emphasized in this group that reduced image

resolution is not the same as reduced contrast. Resolution is
a neuroretinal function. If macular cones are defective
genetically or affected by suppression or localized edema, the
remaining retina transmits a less detailed image which
becomes increasingly less defined toward the periphery. This
image is difficult to simulate optically or photographically
because objects viewed by the para macular retina are not
hazy or unfocused but simply less detailed.
Central Scotoma: Retinal or neurologic conditions that

cause loss of central detail, or scotomas, are the commonest
causes of retinal dysfunction. The most prevalent are age-
related atrophic macular disease and disciform or
100
hemorrhagic/serous maculopathy. Other causes are chronic

cystoid edema, macular holes, inflammatory lesions, laser
photocoagulation for neovascular membranes, trauma, drugs
and demyelinating optic nerve disease.
Diminished function as perceived by patients with the

condition emphasizes difficulty with any type of reading task,
followed by facial recognition which is a social interaction
that has been underestimated; reading signs is a recurrent
problem in distance vision, as is fear of driving and uncertain-
ty in color recognition. On a positive note, peripheral vision
is retained almost without exception.
Patients mayor may not report actual scotomata. In the

atrophic stage, central vision may be blurred or indistinct but
scotomas usually signal either advanced disciform scarring or
the post laser photocoagulation scar.
Function tests have centered around mapping scotomas or

identifying viable retina by scanning. The Amsler grid, Gold-
mann and automated perimetry, tangent screen, contrast sen-
sitivity function tests (CSF), laser scanning and color tests
are currently undergoing more critical evaluation particularly
since no single test provides the answers we need to evaluate
function.
Amsler grid: The Amsler grid has become a household

word in its recent playing card sized version for home testing.
Whether it is sensitive enough or performed correctly is not a
matter to debate here. It has had an unexpected positive
effect on consumer awareness which has created more of a
demand for low vision services. As an alternative to the grid
I have photocopied crossword puzzles for maximum contrast
and asked patients to monitor their central vision for distor-
tion or wavy lines. They find it easier to fixate the center of
the puzzle. Another version of the Amsler grid which I have
developed is the customary grid on a large enough dark back-
ground to allow the positioning of a white oval the size of
blind spot in the position required for a test distance of one
third of a meter. If the patient is co-operative, it is possible
to maintain central fixation by keeping the white target in
the blind spot. If the white target pops out or appears at the
edge, the patient is alerted to the fixation task. Then the
observer can trap the scotoma with a small test wand, mark-
ing the disposable sheet. This scotoma is always larger than
the scotoma described by the patient. The chart is reversed
for the fellow eye.
101
Tangent screens are not valuable for mapping central

scotomas unless the same blind spot fixation can be main-
tained.
Automatic perimetry with macular programs are useful in

early age-related macular disease to follow the central area
in gray tones.
Contrast sensitivity function tests: To follow subtle

changes in macular disorders, contrast sensitivity tests may
prove to be valuable in identifying inability to discriminate
gratings in low, mid and high frequencies. The patient must
be wearing an accurate refractive correction with a compen-
satory lens for the test distance if required. Lighting stan-
dards should be consistent so that test results are compara-
ble.
CSF is surprisingly useful not only in monitoring progress

of disease but in documenting the modulation transfer func-
tion of low vision lenses, helping the clinician select the most
efficient optical aids, either convex lenses or telescopes.
Contrast sensitivity in four types of macular disease:

Contrast sensitivity function (CSF) tests have been carried
out for 18 months on patients with several types of drusen,
retinal pigment epithelial (RPE) atrophy, subretinal neovascu-
lar membrane with serous detachment, and finally disciform
scar. The Vistech 6000 test was used exclusively after com-
parisons with other tests and revealed comparable results
(CORWIN and RICHMAN, [4]). The average time taken for
each eye and a binocular reading was four minutes, using a
separate chart for each test. The time factor as well as ease
of administration encouraged the use of the near contrast
sensitivity test.
Non-confluent drusen: This degenerative condition of

Bruch's membrane is considered to be relatively benign.
Case 2. A 76 year old white female with drusen in both

eyes has normal low frequency CSF curves but low normal
mid and high frequency CSF curves. Distance acuity is 6/12
(20/40) OU and she wears a reading addition of +2.75 D OU.
In view of the lack of clear cut risk factors for eventual
development of neovascular membranes, we may eventually
uncover useful data if we follow drusen over time with con-
trast tests.
102
Mineralized drusen: Mineral deposits in drusen are also

considered to be, if not benign, relatively non destructive.
Case 3. An 81 year old female with 6/9- 1 (20/30- l ) OU

has a normal CSF curve in the left eye. The right eye has
normal low frequency, reduced mid frequency and no response
between 12 and 18 c/deg. The left eye requires a +3.00 D
reading lens, the right +5.00 D for comparable acuity. She
accepted +3.50 D lenses QU.
Confluent drusen: Confluence has been identified as a

risk for eventual development of neovascular membranes.
These drusen are different in appearance from the discrete
lesions mentioned above. They have ill defined borders and
are pale.
Case 4. A 72 year old male has 6/6 (20/20) OD, 6/18

(20/60) OS. The fundus OD reveals a few isolated drusen
while OS has confluent drusen over 15 degrees of the posteri-
or pole. The CSF in the right eye is within the normal range
with a mid frequency dip. The OS on the other hand falls
below the normal from 1.5 to 12 c/deg. The high frequency
patches were not seen. This patient also described distortion
on the Amsler grid and was found on retinal evaluation to
have an area of serous detachment of the RPE temporal to
the macula.
Retinal pigment epithelial atrophy: Truly an age-related

macular change present in most persons over 70; visual acuity
may be normal or remain only slightly reduced for years.
These individuals are extremely responsive to adequate illu-
minance levels and moderate reading additions of +3.00 to
+5.00 diopters.
Case 5. A 72 year old female has no perception of a visu-

al problem although she has advanced RPE atrophy OD. Her
acuity is 6/12 (20/40) OD, 6/7.5 (20/25) OS with a normal
CSF. The CSF curve OD is reduced at 6 and 12 cldeg with no
response above 12 c/deg. Binocular vision is "normal" for
reading with a +3.00 diopter addition.
All of these patients have been followed at 6 month

intervals for 18 months without significant changes either in
CSF or perceived function.
103
The following case illustrates progression of a neovascu-

lar membrane.
Case 6. A 59 year old male with a subfoveal neovascular

membrane OS, 6/9 (20/30) Snellen acuity and no response at
the 18 c/deg targets, has been followed for one year by CSF
and fluorescein angiography. At the last follow up in June
1986 the acuity was noted to be less accurate than usual.
CSF revealed inability to discriminate the 12 c/deg targets
seen at the last visit. Amsler grid was perceived as normal.
A small hemorrhage was noted under the left fovea. A con-
sultation is ongoing regarding the possibility of laser photo-
coagulation to arrest the neovascular membrane.
Binocular testing should be part of the CSF test.
Case 7. A 68 year old male had a visual acuity of 4/70 OD

or 6/105 (20/350) but only 2 spatial frequency responses on
CSF. The OS at 4/80 or 6/120 (20/400) had three spatial fre-
quency responses. Monocular reading acuity at 40 cm OD
12M, OS 4M; but binocularly 2.5M at 40 cm. The patient
rejected a monocular correction OS for binocular prism half
eye glasses (+8.00 D with 10.6 base in OU).
Disciform disease: When patients develop disciform

lesions with gliosis and organization of scar tissue, the scoto-
ma may reach 20 degrees. In some cases the CSF response is
zero which indicates profound low vision and minimal useful-
ness of magnification lenses. CCTV, on the other hand, may
be helpful. In any case, if the response is less than three spa-
tial frequencies and the patient is sensitive only to high con-
trast levels, very high powered reading lenses are required.
6. Treatment Considerations
In spite of discussion about high tech visual aids, on the

everyday clinical level over 90% of the corrective lenses are
convex spectacles, hand magnifiers or illuminated stand mag-
nifiers.
The ideal illumination is bright without being hot (low

wattage), mounted in flexible or adjustable sturdy lamps, and
with filters to provide the best spectrum. Fluorescent, halo-
104
gen or other high intensity bulbs may be uncomfortable for

the low vision patient with macular disease.
Non-optical aids such as writing guides, large print, bold

ink pens and reading stands are useful supplements to optical
devices.
Tinted lenses for the visual difficulties of macular dis-

ease are not a pressing need. Patients with disciform disease
who rely on peripheral retina tend to prefer amber filters
that inhibit UV.
If a patient with macular disease has a cataract that

reduces contrast, consideration should be given to extracap-
sular lens extraction with a posterior chamber lens implant.
Pre-operative acuity can be evaluated best with the SITE or
laser interferometer. The patient must understand that the
operation is being done to restore peripheral vision and to
facilitate the response to magnification.
Case 8. An 81 year old male with nuclear and posterior

subcapsular cataracts and known maculopathy was progres-
sively visually disabled by cataracts. The pre-operative acui-
ty of 3/200 or 6/400 (20/1322) OS was "improved" to 3/70 or
6/140 (20/466) with an intraocular lens. The patient could
once again travel efficiently and use his 12X microscope lens
to read 1M print. He requested surgery on the OD as well,
with the same results. The OS remained the preferred eye
for reading.
CSF and reading aids: Another specific use of the Vistech

6000 near test is to provide an objective measure of low
vision reading aids. Pa tients can view the lines of decreasing
contrast and increasing spatial frequency with a variety of
aids, and curves can be constructed for each aid. For dis-
tance telescopes, use the Vistech 6500 at 3 m.
105
7. Peripheral Field Constriction
The third group consists of diseases that impair the peripheral

field, most significantly for this discussion, retinitis pigmen-
tosa and glaucoma. However, hemianopias, optic nerve dis-
ease, traumatic field defects and peripheral retinal disease
should not be discounted.
Retinitis pigmentosa: Having observed and questioned

patients for many years, I have established an informal cutoff
point for functional disability at about 5 degrees. The reason
for this may be that a field of this size represents the fovea
only, with no peripheral retinal input. People do not orient
themselves rapidly in space which limits their independent
mobility particularly in unfamiliar surroundings. They are
night blind with only a cone-packed fovea. They scan con-
stantly to accumulate central retinal information which they
must process to fill in the missing peripheral information. It
is a laborious process that goes on constantly and takes con-
siderable energy and concentration.
Function tests for this group include ERG, Amsler grid

(or tangent screen), glare tests, visual acuity with refraction,
color and hearing tests. Contrast sensitivity is normal if
foveal retina is normal, but may be useful in following a
young person who is at risk of developing posterior subcapsu-
lar cataracts, or cystic maculopathy.
Treatment tends to be limited to adequate illumination,

Corning's CPF 527 and 511 lenses, CCTV and other magnifi-
cation devices that do not magnify excessively. Minifiers and
prisms have been of theoretical rather than practical value.
Glaucoma: Chronic glaucoma is the second most preva-

lent cause of impaired vision in the USA. Is this inherent in
the disease, or do we fail to diagnose field changes early
enough, even with current diagnostic instruments? In spite of
treatment, the drop-out of the nerve fiber layer seems subtle
and inexorable. This results in loss of contrast and field
defects that may not be detectable in conventional field
tests.
Automated perimetry with 30 degree gray tone threshold

tests reveal more than previous perimeters as do the gray
tone macular tests. However, multiflash campimetry dis-
106
cussed elsewhere in this volume by BR USSELL et aZ. seems to

offer promise for earlier detection.
Contrast sensitivity tests may help to confirm suspects

who already have field defects on automated perimeters, but
may have presented with normal visual acuity and borderline/
normal intraocular pressures. Following patients' response to
medication with CSF as well as perimetry may ultimately
improve therapeutic management.
Individuals with nerve fiber layer damage are difficult to

help with magnifying aids. Because of their decreased neural
sensitivity, they often do not see the magnified image as
clearly as they see a high contrast normal-sized image. The
response of the macula is often subnormal as demonstrated by
threshold fields.
Mobility in retinitis pigmentosa is clinically different

from advanced glaucoma. The former are quite adept at
scanning and can function visually in good lighting levels.
Glaucoma patients seem to have their "rheostats" turned
down permanently. They are sensitive only to high contrast
targets, often responding to only two spatial frequencies.
Many of these persons must have support services such as

orientation/mobility, daily living skills, counselling and voice
output aids such as talking books and calculators, computers
and reading devices.
In conclusion, we have developed great skill over the past

two decades in the mechanics of visual rehabilitation. Now
we must try to understand the psychophysical and functional
aspects of low vision to move beyond simply prescribing aids
and devices by traditional methods.
107
8. References
1. E.E. Faye: Clinical Low Vision, 2nd ed. (Little, Brown,

Boston 1984) pp. 171-196.
2. D. Regan, D. Neima: Low-contrast letter charts as a

test of visual function. Ophthalmology (Rochester) 90,
1192 (1983)
3. A.P. Ginsburg: A new contrast sensitivity test chart.

Am. J. Optom. Physiol. Opt. 61,403 (1984)
4. T.R. Corwin, J. Richman: Three clinical tests of spa-

tial contrast sensitivity function: A comparison. Am.
J. Optom. Physiol. Opt. 63,413 (1986)
The Role of X and Simple Cells in the Contrast Transducer
Function of Low Vision and Normal Observers
Teri B. Lawton
1. Introduction
The defining property of retinal X-like cells and cortical sim-

ple cells is that they exhibit a null phase at which grating
stimuli produce little or no response [1-5]. It follows that for
such "linear summation" cells a masking stimulus at the null
phase should have no effect on detection of a stimulus at the
optimum phase (90 degrees from the null phase). When the
stimuli are in phase, however, we expect the masking stimu-
lus to reduce sensitivity according to the power law of the
contrast discrimination function [6]. Thus, if the contrast
transducer function were determined exclusively by either
retinal X-like cells or cortical simple cells, the degree of
masking should be markedly affected by the phase of the
background relative to the test.
We examined this hypothesis in low VlSlon and normal

observers by presenting brief contrast increments, for a 2
octave range of test frequencies, at 0 degree and 90 degree
phase added to a steady background grating, consisting of sin-
gle or multiple spatial frequencies at a background contrast
of between 0 - 20%. We measured the contrast sensitivity at
different spatial frequencies of low vision observers having
age-related maculopathies (ARM) or myopic maculopathies
and of normal observers to provide a baseline for compari-
sons.
109
2. Characteristics of Maculopathies
Exudative or disciform maculopathies are characterized by

fluid or scar tissue accumulating beneath the macular retina,
where acuity is 6/60 (20/200) or worse [7]. The central 5
degrees of the visual field is often no longer functional in
many ARM patients, and is reduced in patients with myopic
maculopathies. The ARM patient has to detect patterns in
the periphery, outside of the fovea where visual acuity is
highest. Patients with maculopathies show a substantial con-
trast sensitivity loss for intermediate and high spatial fre-
quencies [8-13]. Thus, patterns appear blurred and distorted
[14,15]. Commonly the clinician limits the use of the term
maculopathy to instances where the retinal abnormalities
cause a decrease in visual acuity [7].
ARM patients detect contrast differences using similar

types of visual processing as that used by normal observers in
the periphery. ARM patients, who have little or no central
vision, show a loss in sensitivity to intermediate and high spa-
tial frequencies [8-13]. A loss in contrast sensitivity to inter-
mediate and high spatial frequencies is found as normal
observers discriminate between patterns presented in the
periphery [16-18], compared to discriminations made using
central vision. Once the contrast of a grating is above detec-
tion threshold, however, contrast discrimination in the
periphery is the same as in the fovea [16].
3. Contrast Sensitivity Measurements in Patients with

Maculopathies
Maculopathies are detected earlier and the severity is pre-

dicted more accurately by measuring an observer's contrast
sensitivity function than by using other noninvasive tests of
visual function [8-11] such as Snellen acuity. The peak of the
contrast sensitivity function is the most important parameter
for characterizing the severity of the visual impairments of
patients with maculopathies [10,13]. Patients with maculopa-
thies tend to show a peak sensitivity to 1 c/deg patterns,
their contrast sensitivity decreasing as they detect higher
spatial frequency patterns. The contrast sensitivity function,
measured using portable, analytic, and automated instrumen-
110
tat ion, provides a more complete estimate of the ability of

the patient with maculopathy to cope with a variety of visual
tasks than is provided by other visual acuity tests [12]. The
contrast sensitivity function changes depending on the loca-
tion in the retina that the patient with maculopathies uses to
view the pattern [12], indicating the sensitivity of this meas-
ure to changes in the macula.
4. Use of Inverse Filtering and Magnification in Image

Recognition
If patients with maculopathies process contrast differences

similarly to normal observers in the periphery, then inverse
filtering which would increase the contrast of intermediate
spatial frequencies above the detection threshold of ARM
patients should significantly improve contrast discrimination.
Filtering of the image by the inverse of the patient's contrast
sensitivity function to increase the contrast of the less visible
spatial frequency components should improve image recogni-
tion for the visually impaired [19]. Inverse filtering should
be particularly effective for ARM patients since the contrast
sensitivity function predicts losses in visual function at dif-
ferent levels of processing more accurately than any other
clinical tests of visual function. Magnification will shift the
pattern frequencies towards the low spatial frequencies which
are significantly more visible for the ARM patient. There-
fore, both inverse filtering and magnification were used to
correct the visual transfer function so that image acquisition
was optimized for the ARM patient.
Psychophysical measurements find a shift in the peak

sensitivity to lower spatial frequencies as more eccentric fix-
ations are used for both ARM patients [12] and normal
observers [16,17]. Single unit recordings, however, find that
there are cortical cells tuned to approximately the same
range of spatial frequencies in the fovea and at 5 degrees
eccentricity [20]. The neurophysiology suggests that it is pos-
sible for inverse filtering to improve image recognition for
ARM patients following sufficient practice.
If patterns are magnified for normal observers as they

are presented at eccentric locations, then patterns detected
in the periphery are seen as well as patterns detected in the
111
fovea [21-23]. Since ARM patients must use eccentric view-

ing to detect patterns, it is likely that magnification would
aid image recognition for ARM patients. The reduced con-
trast sensitivity of patients with maculopathies to intermedi-
ate and high spatial frequencies can be partially compensated
for by increased magnification to aid image recognition
[9,10]. Too much magnification, however, will produce such
large letters that reading performance deteriorates [24].
Magnification alone is not sufficient for transmitting recogni-
zable information to the visual cortex [25]. Only by enhanc-
ing the less visible components can the image be restored to
the sharpness seen by normal observers [26].
5. Experimental Methods
5.1 Apparatus
The Visual Testing and Training Instrument (VTl) consists of a

PDP 11-23 laboratory computer, interface software and hard-
ware, and a high resolution display [27]. All the functions
needed to display the proposed stimulus variations are already
implemented in software. The VTl enables the test and back-
ground patterns to be varied independently across space and
over time. The number of patterns in a test interval and the
spatial extent, spatial frequencies, contrast, position, move-
ment, and duration of the test and background pattern com-
ponents are varied independently by the VTl. Each stimulus
parameter can be input from the keyboard by the experimen-
ter or automatically from stimulus datafiles.
The VTl displays any number of pattern intervals, for any

duration containing vertical mUltiple spatial-frequency sinu-
soidal gratings, having a test frequency in any phase relative
to a steady or moving single or multiple frequency back-
ground having a contrast and mean luminance input by the
experimenter. Each stimulus parameter is independently con-
trolled by the VTl. The VTl displays one or two patterns hav-
ing a variable width so that variable pattern widths at differ-
ent eccentricities can be used to measure contrast thresholds
for contrast, spatial position and movement discrimination,
using either a temporal or spatial 2 Alternative Forced-
Choice (AFe) procedure.
112
Multifrequency gratings were displayed on a high resolu-

tion HP 1332a CRT display. The contrast of the pattern
increased gradually reaching a maximum at the center of the
pattern presentation and then decreases gradually, for the
gradual pattern presentations. The contrast of the pattern
jumped from 0 to a fixed level for the abrupt pattern presen-
tations. The contrast of different frequency gratings and dif-
ferent size letters were calibrated using a Pritchard photom-
eter.
5.2 Stimuli
Gratings were composed of spatial frequency components

spanning a 3 octave range to examine the types of interac-
tions that exist between different spatial frequency compo-
nents as different types of visual processing are used for con-
trast discrimination. A 2 octave range of test frequencies
were added to single or multifrequency backgrounds. Multi-
frequency backgrounds consisted of intermediate spatial fre-
quencies that repeat over a wide area. These backgrounds
have been shown to facilitate spatial-phase discrimination
[28]. Multifrequency backgrounds consisted of several cosinu-
soidal components all in zero degree phase (Fig. 1). The posi-
tion of the test grating was either shifted to the right or not
at all before being added to the background. A 360 degree
phase difference corresponds to the spatial period of the fun-
damental frequency of the test or background grating, whi-
chever is highest [26]. The ~ean luminance of test and back-
ground gratings as 17 cd/m • The visual field subtended 4
degree visual angle.
5.3 Procedures
The contrast thresholds for discrimination of a brief contrast

increment at 0 degree and 90 degree phase were measured
using a 2 AFC contrast discrimination task, for a 2 octave
range of test frequencies added to a steady single or multi-
frequency background grating. The contrast increment was
presented for 250 msec using a gradual onset and offset, 250
msec after the onset of a 750 msec background. The two pat-
tern intervals wer~ separated by 500 msec. The mean lumi-
nance was 17 cd/m .
113
LfFT -SHIFTED SINEWAVE
SPATIAL POSITION LEFT-SHIFTED STIMULUS-
, .,
+
.. ,.---=BA
= CK;;
G::..:
ROU= N;::
D_ - ,
RlGHT-SHIn-ED STIMULUS·
SPATIAL POSITION LEFT -SHIFTED STIMULUS'

LEGEND
o. 3 c/dog TEST GRATING
PHASE-SHIFTED 1/B PERIOD
TO LEFT.
b. 6 + 7 c/dog BACKGROUND
FREQUENCIES IN PHASE.
c. 3 c/deg TEST GRATING
PHASE-SHIFTED 1/B PERIOD
TO RIGHT.
d. 3.5 c/dog TEST GRATING PHASE- ..
SHIFTED 1/B PERIOD TO LEFT.
ADDED TO 6 + 7 c/dog
BACKGROUND FREQUENCIES.
o. 3.5 c/deg TEST GRATING PHASE-
SHIFTED 1/B PERIOD TO RIGHT.
ADDED TO 6 + 1 c/dog
BACKGROUND FREQUENCIES
2411l1li
RIGHT -SHIFTED STIMULUS'
-THESE ItICTURES AllIE ONLY THE CENTER STRIP Of THE STIMULI
Figure 1. Phase-shifted 3- and 3.5- c/deg test gratings

added to a 6 + 7 c/deg background.
The observer discriminated whether the test grating seen

as a brief contrast increment was added to the background in
the first or second pattern interval. The contrast increment
was added either in-phase or 90 degrees out-of-phase for 250
msec in the center of the pattern interval. The contrast
increment is either presented gradually on a 750 msec back-
ground at a stimulus onset asynchrony (SOA) of 250 msec or
abruptly on a 250 msec background at an SOA of 0 msec. The
observer's task was to identify the pattern interval containing
the contrast increment.
114
Contrast thresholds were measured using a 2 AFC task by

varying the contrast between each pair of pattern intervals.
The contrast was decremented, from the level determined
during the practice run, one step each time the observer cor-
rectly identified the pattern interval containing the test grat-
ing. A staircase method [29] was used, following the first
incorrect response, to measure the minimum contrast needed
to correctly identify the left-shifted or incremental stimulus
79% of the time. This forced choice procedure minimizes the
variance of observer responses and maximizes the sensitivity
and repeatability of threshold measurements [30].
Threshold measurements were to be obtained interactive-

ly by the VTI. The order of different values of the indepen-
dent variable, such as the contrast of the background, were
determined randomly to minimize the effects of adaptation
and response bias. Approximately 35-40 pairs of test and
background patterns were presented to determine one thresh-
old measurement. Approximately 2 threshold measurements
were averaged for each data point that is plotted. Approxi-
mately 100-200 data points were collected for each stimulus
parameter that is varied.
Disciform ARM patients learned to focus the pattern in

the periphery (eccentric fixation) where the contrast sensitiv-
ity is greater than in the macula where little or no visual
information is detected. Training an ARM patient to use
eccentric fixation is usually a difficult, time-consuming task
[25,31]. However, using the Visual Testing and Training
Instrument (VTI) developed by Lawton [27,28] eccentric view-
ing is easily learned by ARM patients.
5.4 Observers
The ARM patient, who is 71 years old, has disciform ARM

with a corrected Snellen acuity of 6/60 (20/200) in the right
eye and 5/400 (6/480 or 20/1600) in the left eye. Laser pho-
tocoagulation treatment was used to prevent the acuity of
the right eye from regressing to the acuity of the left eye.
The left eye of this ARM patient was patched to remove pat-
tern distortions caused by the large refractive difference
between the two eyes. The patient with myopic maculopa-
thies, who is 28 years old, has a corrected Snellen acuity of
6/9 (20/30) in both eyes. The normal observer, who is 33
years old, has an acuity of 6/6 (20/20) and has extensive prac-
tice discriminating contrast differences.
115
Patients with maculopathies and normal observers used simi-

lar discrimination processes when discriminating brief con-
trast increments added to either single or multifrequency
backgrounds. For either patients with maculopathies or nor-
mal observers [32] no difference was found between the con-
trast sensitivity for 0 degree and 90 degree increments at any
combination of spatial frequencies or background contrasts
(Figs. 2-6). The defining property of retinal X-like cells and
cortical simple cells, linear summation cells, is that they
exhibit a null phase at which grating stimuli produce little or
no response [1-5] (Fig. 7). It follows that a background grat-
ing at the null phase should have no effect on the detection of
a stimulus at the optimum phase (90 degrees from the null
phase). When stimuli are in phase, however, we expect the
background stimulus to reduce sensitivity according to the
power law of the contrast discrimination function [6]. Thus,
if the psychophysical contrast discrimination function is
determined exclusively by linear summation cells, then the
degree of masking should be markedly affected by the posi-
tion of the background's peak luminance relative to the test
stimulus' peak luminance. Since no difference was found
between the contrast sensitivity for 0 degree or 90 degree
increments at any combination of spatial frequencies or back-
ground contrasts, it is likely that cells exhibiting nonlinear
summation are used to detect contrast differences for both
macular degeneration and normal observers. This indicates
that magnocellular processing, mediated by Y-like or complex
cells [3,5], determines the visual response at low contrasts.
The SOA was reduced from 250 msec to 0 msec to

present stimuli that are detected more easily by non-linear
spatial summation cells, than by the linear spatial summation
cells. The results for the ARM patient and a normal observer
are presented in Fig. 6. For these patterns, no difference was
found between the contrast sensitivity for O.,degree and 90
degree increments. These results show that reducing the pat-
tern duration from 750 msec to 250 msec and reducing the
SOA from 250 msec to 0 msec lowered contrast thresholds.
The results for both long and short duration patterns show
that an observer's contrast sensitivity for detecting contrast
increments is phase-independent. Therefore, retinal X-like
and cortical simple cells can be involved in contrast detection
only with some type of pooling between cells in quadrature
phase. Both threshold and supra threshold contrast sensitivity
116
conform to behavior expected for magnocellular processing

resulting from V-like or complex cell mediation.
60.0 r--'-r'V-,----,----.---r--,
50.0
40.0
30.0 l--- ----~...... ... ~ . . ,;"",;

c
~
200
~
0.
15.0
ci
..J
0
:I: 10.0
~ 9.0
a: 8.0
:I: 7.0
I- 6.0
l-
V> 5.0
..:
a:
I- 4.0
Z
0 3.0
u
>-
u
fE
::0
2.0
8a:
u-
I- 1.0
~ 0.9
I- 0.8
0.7
0.6
0.5
0.4
0.3
o 2.5 5.0 10.0 20.0
BACKGROUND CONTRAST, percent
-SPATIAL FREQUENCY = 1 eye/dog
; : ~~;;~~i:OF-PHASE FOR ARM OBSERVER
;, : ~~(t~~i:OF-PHASE FOR NORMAL OBSERVER
---SPATIAL FREQUENCY' 7 eye/deg
~~(t~~i:OF-PHASE
o: ~~(t~~i:OF-PHASE
; : FOR ARM OBSERVER
FOR NORMAL QBSERVER
Figure 2. Contrast discrimination on low and intermedi-

ate frequency backgrounds.
Differences between the contrast sensitivity functions of

patients with maculopathies and normal observers provides a
sensitive test of losses in visual function for a wide range of
spatial frequencies. When detecting low spatial frequencies,
such as I c/deg test grating on a I c/deg background, the con-
trast discrimination thresholds for the disciform ARM patient
was approximately 6 times higher than those for a normal
observer (Fig. 2). When detecting intermediate spatial fre-
quencies, such as a 7 c/deg test grating on a 7 c/deg back-
117
60.0~---."~~--~-~--~~
50.0
40.0
~ 30.0
~
~
Cl
20.0
...J
o 15.0
I
~
~ 10.0
I- 9.0
to ~g
~ 6.0
\;; 5.0
8 4.0
>-
~ 3.0
:0
~ 2.0
a:
u.
...J
<t
i=
1.0
;t 0.9
V)
O.B
f- 0.7
~ 0.6
f-
0.5
0.4
0.3
20.C
-SPATIAL FREQUENCY· 1 eye/deg
; : ~~~~~~:~F-PHASE FOR MYOPIC OBSERVER
;,. : ~~~P~~i_EOF_PHASE FOR NORMAL OBSERVER
---SPATIAL FREQUENCY' 7 eye/deg
~ ~6~P~~.;:aF-PHASE
o: ~~~P~0'i:OF-PHASE
: FOR MYOPIC OBSERVER
FOR NORMAL OBSERVER
Figure 3. Contrast discrimination on low and intermedi-

ate backgrounds.
ground, the contrast discrimination thresholds for the disci-

form ARM patient was 8-30 times higher than those for a
normal observer (Fig. 2). ARM patients show a loss in con-
trast sensitivity to all spatial frequencies. A loss in contrast
sensitivity to all spatial frequencies was also found for the
patient with myopic maculopathies (Fig. 3). However, the
thresholds were only 2 times higher than those for a normal
observer, instead of 6-30 times higher as found for the disci-
form ARM patient.
The contrasts needed to detect test gratings on different

contrast backgrounds for patients with maculopathies (Figs.
118
60.0 r-----T-----.---,.----.---.--.
50.0
40.0
i 30,0
ci 20.0
o-'
I
~
~ 10.0
t- 9.0
t- B.O
(.f.) 7.0
;i 6.0
~ 5.0
8 4.0
-------!----~----~
>-
~ 3.0
:::J
~ 2.0
a:
u.
~
~ 6',g
~ O.B
~ 0.7 Q..~~~.Iti
0.6
0.5 f-- • ---~----1
0.4
0.3 '---JJ'v---'--~'-------'-,---'---'
-BACKGROUND FREQUENCY = 1 eve/deg
;, : ~~(;P~~~~OFPHASE FOR ARM OBSERVER

;,. : ~~~P~~~~OF-PHASE FOR NORMAL OBSERVER
---BACKGROUND FREQUENCIES' 6 + 7 eye/dog
~ ~~(;P~~i~OF-PHASE
o:
: FOR ARM OBSERVER
~~o"~~~~OF-PHASE FOR NORMAL OBSERVER
Figure 4. Contrast discrimination on single and multifre-

quency backgrounds.
2-4) show a similar type of processing as that found for nor-

mal observers in the periphery [16]. Once the observer's con-
trast threshold was reached there was a gradual increase in
the detection threshold as the background contrast was
increased. There is a more gradual change in contrast sensi-
tivity as the contrast increments are detected on different
contrast multifrequency backgrounds that repeat over a wide
area for normal observers than for increments on single fre-
quency backgrounds, and on both single and multifrequency
backgrounds for ARM observers.
119
60.0 r - - - ' I - - - - , - - - , , - - , - - - - - - - ,
50.0
40.0 ~~g~g~g~~g g~~:~~g~: i~; msee
30.0
i
c
20.0
g'
0
-
'8:8
~
~
a:: B.O
J: 7.0
I- 6.0
I-
5.0
~~ii
U)
..:
a::
I- 4.0
~----~/
Z
0 3.0
u
>-
u
~ 2.0
:::>
£a::
~
~
-
5
I-
1.0
0.9
O.B
0.7
0.6
0.5
0.4
0.3
1.0 20 30 4.0
lEST SPATIAL FREQUENCY, eye/deg
-BACKGROUND FREQUENCY = 1 eye/deg
~ : ~~;,P~~i:OF -PHASE FOR ARM OBSE RVE R

~ : ~~~P~~i:OF-PHASE FOR NORMAL OBSERVER
---BACKGROUND FREQUENCIES' 6 + 7 eye/deg
~ ~~~P~~i:OF-PHASE
o: ~~~P~~i:OF-PHASE
: FOR ARM OBSERVER
FOR NORMAL OBSERVER
Figure 5. Contrast discrimination for a 2-octave range of

spatial frequencies on single and multifrequency
backgrounds.
It is important to determine whether increasing the con-

trast of intermediate background frequencies that repeat
over a wide 1 degree region will lower the contrast discrimi-
nation thresholds. This will provide a test of whether magni-
fication and inverse filtering will be effective in improving
image recognition. We found that ARM patients see a con-
trast increment on a multifrequency background composed of
intermediate spatial frequencies that repeat over a 1 degree
area at much lower contrasts than for an equal contrast I
c/deg background composed of a single spatial frequency
component (Figs. 4-6). Lowering the fundamental frequency
120
40.0
30.0
BACKGROUND CONTRAST' 10%
BACKGROUND DURATION' 250 moee
20.0
c
~
c. 10.0
A
ci 9.0
...J 8.0
0 7.0
:J:
'"
w
a:
6.0
5.0
:J:
>- 4.0
>-
'"a:<t 3.0
z>-
~---l
0 2.0
u
>-
u
~
::;)
1.0
~ 0.9
a: 0.8
~
0.7
§ 0.6
>- 0.5
0.4
0.3
0.2
TEST SPATIAL FREQUENCY, eye/deg
-BACKGROUNO FREQUENCY' 1 eye/dog
~ : ~~~P~~i~OFPHASE FOR ARM OBSERVER

;,. : ~~~P~~i~OF-PHASE FOR NORMAL OBSERVER
---BACKGROUND FREQUENCIES' 6' 7 eye/dog
~ : ~~~,P6~i~OF-PHASE
o: ~~:~~i~OF-PHASE
FOR ARM OBSERVER
FOR NORMAL OBSERVER
Figure 6. Contrast discrimination for a 2-octave range of

spatial frequencies on single and multifrequency
backgrounds.
of the multifrequency backgrounds magnifies the area over

which the background repeats. Lowering the fundamental
frequency yet keeping the intermediate background frequen-
cies in a 1/2 octave bandwidth enables measuring the effects
of magnification and inverse filtering separately. The results
from this study indicate that inverse filtering combined with
magnification may significantly aid image recognition for dis-
ciform ARM patients.
121
IN-PHASE (0°)
DETECTION
FIELD --~=--+-~---:~---
TEST
MASK
MASKED TEST
FIELD
MASK
DETECTION --o;::--""-H<--~--
FIELD
TEST
MASK
MASKED
FIELD TEST
MASK
Figure 7. Contrast detection using "linear summation"

cells.
7. Acknowledgements
The research described in this paper was carried out by the

Jet Propulsion Laboratory, California Institute of Technology,
under a contract with the National Aeronautics and Space
Administra tion.
122
8. References
1. D.H. Hubel, T.N. Wiesel: Laminar and columnar distri-

bution of geniculo-cortical fibers in the macaque mon-
key. J. Compo Neurol. 146,421 (1972)
2. C. Enroth-Cugell, J .G. Robson: The contrast sensitivi-

ty of retinal ganglion cells of the cat. J. PhysioZ:
(Lond.) 187, 517 (1966)
3. P.H. Schiller, J.G. Malpeli: Functional specificity of

lateral geniculate nucleus laminae of the rhesus mon-
key. J. Neurophysiol. 41, 788 (1978)
4. D.H. Hubel, T.N. Wiesel: Receptive fields, binocular

interaction, and functional architecture in the eat's
visual cortex. J. Physiol. (Lond.) 160, 106 (1962)
5. E. Kaplan, R.M. Shapley: X and Y cells in the lateral

geniculate nucleus of macaque monkeys. J. Physiol.
(Lond.) 330, 125 (1982)
6. G.E. Legge: A power law for contrast discrimination.

Vision Res. 21,457 (1981)
7. The Framingham Eye Study Monograph. Surv.

Ophthalmol. (Suppl.) 24, 335 (1980)
8. J. Sjostrand, L. Frisen: Contrast sensitivity in macular

disease. Acta Ophthalmol. (Copenh.) 55, 507 (1977)
9. B. Brown: Reading performance in low vision patients:

Relation to contrast and contrast sensitivity. Am. J.
10. G.S. Rubin, G.E. Legge: Predicting low-vision reading

rates from measures of contrast sensitivity. Presented
to the First meeting of Noninvasive Assessment of the
Visual System sponsored by the Optical Society of
America, Incline Village, Nevada (1985)
11. M. Wolkstein, A. Atkin, I. Bodis-Wollner: Contrast sen-

sitivity in retinal disease. Ophthalmology (Rochester)
87, 1140 (1980)
123
12. R.F. Hess, R.J. Jacobs, A. Vingrys: Central versus

peripheral vision: Evaluation of the residual function
resulting from a uniocular macular scotoma. Am. J.

14. R.C. Watzke: Acquired macular disease. In Clinical

Ophthalmology, Vol. 3, Chapt. 23, ed. by T.D. Duane
and E.A. Jaegar (Harper and Row, Philadelphia 1976)
15. A.M. Fine, M.J. Elman, J.E. Ebert, P.A. Prestia, J.S.
Starr, S.L. Fine: Earliest symptoms caused by neovas-
cular membranes in the macula. Arch. Ophthalmol.
104, 513 (1986)
16. G.E. Legge, D. Kersten: Contrast discrimination in

peripheral vision. Invest. Ophthalmol. Vis. Sci. (Suppl.)
27, 225 (1986)
17. J. Rovamo, V. Virsu, R. Nasaren: Cortical magnifica-

tion factor predicts the photopic contrast sensitivity
of peripheral vision. Nature 271, 54 (1978)
18. R.F. Hess, J .S. Pointer: Differences in the neural basis

of human amblyopia: The distribution of the anomaly
across the visual field. Vision Res. 25, 1577 (1985)
19. E. Peli, T. PeU: Image enhancement for the visually

impaired. Opt. Eng. 23,47 (1984)
20. R.L. De Valois: Encoding of spatial frequency, orienta-

tion, and contrast in striate cortex. Optic News 11,
119 (1985) and personal communication at ARVO
(1986)
21. J. Rovamo, V. Virsu: An estimation and application of

the human cortical and magnification factor. Exp.
Brain Res. 37, 495 (1979)
22. G. Westheimer: Scaling of visual acuity measurements.

Arch. Ophthalmol. 97, 327 (1979)
124
Z3. C.W. Tyler: Space-time metric of function organiza-

tion in human peripheral vision. Optic News 11, 113
(1985)
Z4. G.E. Legge, D.C. Pelli, G.S. Rubin, M. Schleske: Psy-

chophysics of reading II. Low vision. Vision Res. 25,
Z53 (1985)
Z5. G.L. Goodrich, E.B. Mehr: Eccentric viewing training

and low vision aids: Current practice and implications
of peripheral retinal research. Am. J. Optom. PhysioZ.
Opt. 63, 119 (1986)
Z6. D.B. Gennery: Determination of optical transfer func-

tion by inspection of frequency-domain plot. J. Opt.
Soc. Am. 63, 1571 (1973)
Z7. T.B. Lawton: The effect of phase structure on spatial

phase discrimination. Vision Res. 24, 139 (1984)
Z8. T.B. Lawton: Spatial frequency spectrum of patterns

changes the visibility of spatial-phase differences. J.
Opt. Soc. Am. (A) 2, 1140 (1985)
Z9. G.B. Wetherill, H. Levitt: Sequential estimation of

points on a psychometric function. Br. J. Math. Stat.
PsychoZ. 18, 1 (1965)
30. K.E. Higgins, M.J. Jaffe, N.J. Coletta, R.C. Caruso,

F.M. de Monasterio: Spatial contrast sensitivity:
Importance of controlling the patient's visibility cri-
terion. Arch. OphthalmoZ. 102, 1035 (1984)
31. Y.L. Yap, H.E. Bedell, P.L. Abplanalp: Blind spot "fix-
ation" in normal eyes: Implications for eccentric view-
ing in bilateral macular disease. Am. J. Optom. Physi-
oZ. Opt. 63, Z59 (1986)
3Z. T.B. Lawton, C.W. Tyler: The role of X and Simple

cells in the contrast transducer function. Invest.
OphthaZmol. Vis. Sci. (Suppl.) 27, 340 (1986)
Differential Retinal Structural Damage Exhibited by
Image Enhancement of Fundus Photographs
Sunanda Mitra, Steven Whiteside, Thomas Krile
1. Introduction
The use of fundus photographs and fluorescein angiographs

aids significantly in differential diagnosis of retinal diseases
[1]. The quality of fundus photographs is, however, often poor
and therefore early detection of retinal damage is not always
an easy task. To restore and enhance information contained
in degraded photographs such as aerial photographs or satel-
lite images, high speed digital computers have been used
since the early 1960s [Z,3]. Such image processing operations
have been recently applied to analyze fundus photographs for
early detection of retinal structural damage [4]. Until now,
most image processing algorithms such as adaptive binariza-
tion, spatial and spectral filtering, point to point mappings as
well as image registration and normalization of illumination,
have been performed with mainframe or minicomputers. The
use of such techniques as a clinical research tool necessitates
development of a user-friendly, low-cost image processing
system with menu-driven specific software for easy operation
and documentation. Such a fundus image processing system
(FIPS) has been recently developed with specific image pro-
cessing capabilities comparable to existing large scale sys-
tems. The applicability of such capabilities in quantitative
estimation of retinal structural damage arising from macular
degeneration and glaucoma is presented.
126
2. Method
Image analysis of fundus photographs was performed using

two image processing systems: (1) a COMTAL VISION ONE/20
interfaced to a VAX 11/780 computer and (2) a recently
developed microcomputer based fundus image processing sys-
tem (FIPS) [5]. Slides of fundus images were directly digi-
tized with an image capture device consisting of a CCD or
vidicon camera and a frame grabber. The digitized images
were stored in the computer prior to applying any preprocess-
ing algorithms for image normalization, registration and
enhancement. For rapid processing time, a 256x256 subimage
was chosen for image enhancement by spectral filtering.
Spatial domain processes such as contrast stretching and line
scan of gray level were usually performed on a chosen quad-
rant of the digitized image. Similar subimages were also cho-
sen when performing image normalization and registration
algorithms. Fundus photographs of two groups of patients
afflicted with either macular degeneration or glaucoma were
analyzed.
3. Results
3.1 Analysis of Fundus Photographs of Glaucoma Suspects
Figure 1 demonstrates how a poor quality fundus image of a

glaucoma suspect exhibits nerve fiber layer streaks after
digital image enhancement when no structural difference is
observed in the original photograph. The VAX-COMTAL sys-
tem was used in the above analysis. Figure 2a shows the fun-
dus image display of the FIPS when a high-contrast red-free
photograph of a glaucoma suspect was used. Figure 2b shows
the result of an enhanced subimage of nerve fiber layer.
3.2 Analysis of Fundus Photographs of Patients with

Macular Degeneration
Figure 3 shows the image of a normal macula and its gray

level linescan dis piayed by FIPS. Figures 4 and 5 show
enhanced fundus image displays of two patients having differ-
ent types of macular degeneration and the corresponding line-
scans.
127
a b
.~
.. t.
').0
",
c d
Figure 1. (a) A 512x512 sampled image of a poor quality

fundus photograph of a glaucoma suspect. (b)
Fourier transform of a 256x256 subimage
including the optic nerve head and a section of
nerve fibre layer. (c) Inverse Fourier transform
of the subimage (b) after the image enhance-
ment operations of a high pass filter and con-
trast stretching. The above operations highlight
the optic cup/disk contours and the nerve fiber
layers. (d) A contrast reversal of the image (c).
128
b
Figure 2. (a) A displayed fundus image on the mono-
chrome monitor of the microcomputer based
fundus image processing system (FIPS). (b) An
enhanced image of (a) on the FIPS monitor.
129
Figure 3. (a) A normal macula displayed on FIPS monitor

(b) A horizontal linescan of the gray levels
across the region shown by the scale on Fig. 3(a)
130
b
Figure 4. (a) Enhanced macular region of patient
suspected of having a macular hole. (b) The hor-
izontal line scan across the region of interest
indicate a loss of reflectivity as compared to a
normal macula thus suggesting structural
changes in the foveal region.
131
Figure 5. (a) Enhanced macular region of a patient also

suspected of having a macular hole. However
the enhanced image suggests the existence of a
cyst rather than a hole. (b) The corresponding
line scan shows a higher reflectivity in the cyst
region.
132
4. Discussion
The results presented here demonstrate the feasibility of

digital image processing techniques as a non-invasive method
of assessing early retinal damage caused by macular degener-
ation and glaucoma. However, the fundus photographs have
to be carefully preprocessed for normalization of illumination
and registration (geometric alignment) so that an artifact-
free evaluation of retinal damage can be obtained by image
subtraction [6,7].
The high speed microcomputer based image processing

system FIPS [5] developed recently may be quite valuable as a
clinical research tool for evaluating retinal damage and for
early diagnosis of glaucoma thus reducing the risk of glauco-
ma induced blindness. Further research has to be carried out
to analyze a large number of fundus photographs of patients
of each disease group in order to establish the validity of this
approach. A correlation between the actual structural dam-
age in the retina and the functional loss in vision as assessed
by visual field loss has to be demonstrated also. At present
clinical perimetry is capable of detecting visual field defects
sometimes not before five years after detection of nerve
fiber layer anomaly in glaucoma [8]. Other perimetric meth-
ods [9,10] may be considered to detect early loss in vision for
correlating functional loss with the structural damage
observed.
5. Acknowledgements
This research was supported by a grant from Microcraft Inc.,

Dallas, Texas. The hardware and operating system software
have been developed by Microcraft Corp. The fundus photo-
graphs have been supplied by Dr. Z. Shihab of Texas Tech
Health Sciences Center, Lubbock, Texas.
133
6. References
1. J .D.M. Gass: Stereoscopic Atlas of Macular Disease,

2nd ed. (Mosby, St. Louis 1977)
2. H.C. Andrews, B.R. Hunt: Digital Image Restoration

(Prentice-Hall, Englewood Cliffs 1977)
3. W.K. Pratt: Digital Image Processing (Wiley, New

York 1978)
4. J. Kern, B. Schwartz: Digital image processing in the

diagnosis of glaucoma and ocular disease. In Applica-
tions of Digital Image Processing, ed. by A.G. Tescher,
Proceedings of the Society of Photo-Optical Instru-
mentation Engineers Vol. 119 (SPIE, Bellingham 1977)
pp.274-7.
5. S. Whiteside, S. Mitra: Digital image processing in

early diagnosis of glaucoma. Paper to be presented at
the 8th Annual Conference of IEEE Engineering in
Medicine and Biology Society, Forth Worth, TX,
November 7-10, 1986.
6. 0.1. Barnea, H.F. Silverman: A class of algorithms for

fast digital image registration, IEEE Trans. Comput.
C-21, 179 (1972)
7. A. Ling, T. Krile, S. Mitra, Z. Shihab: Early detection

of glaucoma using digital image processing. Invest.
8. A. Sommer, H.A. Quigley, A.L. Robin, N.R. Miller,

J.Katz, S. Arkell: Evaluation of nerve fiber layer
assessment. Arch. Ophthalmol. 102, 1766 (1984)
9. S. Mitra: Spatial contrast sensitivity in macular disor-

der. Doc. Ophthalmol. 59, 247 (1985)
10. C.W. Tyler: Specific deficits of flicker sensitivity in

glaucoma and ocular hypertension. Invest. OphthaZ-
moZ. Vis. Sci. 20, 204 (1981)
The Visual Requirements of Mobility
Denis G. Pelli
1. Introduction
Most of us use our eyes to great advantage in getting about

the world. Visual guidance of walking is a very useful skill.
While the totally blind can and do learn to travel without
sight, they do so using different cues, and travel more slowly
and less safely than the sighted. However, relatively few
people - only half a million in the U.S. - are totally blind,
whereas several million have low vision. Low vision is handi-
capped vision, i.e. visual function is sufficiently restricted to
seriously impair the peron's ability to do important tasks [1].
The two most important of those tasks, judging by the com-
plaints of people with low vision, are reading and mobility.
The visual requirements of reading have been described else-
where [2-4]. Here we consider the visual requirements of
mobility.
The low vision population is mostly and increasingly eld-

erly. It is very common for those with low vision to cease
travelling independently because they feel they can no longer
do so safely. This makes them dependent on others for shop-
ping and for visiting anyone outside the home. One motiva-
tion of our research is the hope that a better understanding of
the visual requirements of mobility might lead to help for
these people.
Our method is empirical. We artificially restrict the

vision of normally sighted subjects, and measure their per-
formance at specific mobility tasks. There are several
advantages in using artificial restriction of normal vision,
rather than real low vision. The artificial restrictions are
well understood, stable, and repeatable from observer to
observer. They can be varied over the full range of 0 to 100%
of normal. They allow parametric investigation of the effect
of one variable without confounding variation of the other
parameters. We can retest the same subject at many degrees
of restriction so the subjects act as their own controls. This
allows us to discount the effects of subject variables such as
motivation and athletic skill.
135
We are also testing low vision subjects on the same

mobility tasks. We do this because we are interested in the
mobility of people with low vision. It is possible that a long-
term visual restriction is either less disabling {as a result of
long-term learning} or more disabling {as a result of long-
term deprivation from important environmental cues, or,
more likely, as a result of loss of confidence and motivation}.
MARRON and BAILEY [5] measured the mobility and

vision of nineteen low vision subjects. They found that area
of visual field and peak contrast sensitivity were the best
predictors of mobility performance and that acuity was not a
good predictor.
We controlled and measured three parameters of V1S10n:

field, resolution {i.e. acuity}, and contrast. These three
parameters provide a first-order description of any optical
imaging system, such as the human eye.
Recently there has been considerable interest in measur-

ing contrast sensitivity function. Figure 1 illustrates the
reader's contrast sensitivity function. Any patch of the fig-
ure is a sinusoidal grating. As you direct your gaze gradually
upward the contrast falls until the grating eventually disap-
pears. As you go from left to right the grating gets finer
until the grating is too fine to resolve. The outline of visibili-
ty, an inverted U, is the reader's contrast sensitivity function.
This function is characterized by its vertical position on the
contrast axis and by its horizontal position along the spatial
frequency or resolution axis. In our experiments we restrict-
ed contrast, corresponding to the vertical axis of Fig. 1, and
resolution, corresponding to the horizontal axis of Fig. 1.
136
Figure 1. A sinusoidal grating. The spatial frequency

increases from left to right. The contrast
decreases from bottom to top. The outline of
visibility of the grating is the viewer's contrast
sensitivity function. The diagram is based on a
similar figure by Campbell and Robson.
2. Methods
The subject's other eye was occluded so all viewing was

monocular. Thus all subjects were denied the benefit of bino-
cular parallax, a cue to depth and three dimensional struc-
ture. This was done for two reasons. Firstly, it is technically
difficult to maintain alignment of devices worn on two eyes
as the subject walks about. Secondly, clinically, diseases pro-
ducing low vision rarely affect both eyes equally, and a per-
son is classified as having low vision on the basis of the vision
in the better eye, so it is not clear how much benefit the
worse eye provides. F or these reasons the possible role of
binocular parallax in mobility is beyond the scope of this
study. It is worth pointing out, however, that the subjects
137
move through the test environments and thus do benefit from

motion parallax, which offers very similar information to
binocular parallax, yielding a strong impression of three
dimensional depth, as every movie goer knows.
We use three kinds of artificial visual restriction: field,

contrast, and resolution. All the restrictions are designed to
fit in standard trial frames.
We restrict field by a truncated paper cone, base in (Fig.

2). This produces a peripheral field loss. Note that the field
restriction is fixed to the head, unlike a scotoma, which is
fixed to the eye. To scan the world our subjects must move
their heads, whereas people with scotomas scan the world by
moving their eyes. We used six field diameters from 1 degree
up to 60 degrees.
Figure 2. Subject wearing trial frames and paper cone to

restrict visual field.
We restrict contrast by contrast reduction disks (Fig. 3).

They consist of 0.5 11 m diamonds suspended in a clear plastic
138
disk the size of a trial lens. Looking through a contrast

reduction disk is like looking through a fog. The contrast
reduction factor was controlled by making disks with various
concentrations of diamonds. Contrast sensitivity measure-
ments have shown that these disks reduce the contrast of the
retinal image by the same factor at all spatial frequencies
from 0.5 to 30 c/deg. Many diseases can cause loss of con-
trast sensitivity. In particular, intraocular scatter (e.g. due
to cataract) will reduce the contrast of the retinal image.
We use six contrast factors from 0.3% up to 100% of normal
contrast.
Figure 3. Subject wearing trial frames and blurscope to

restrict resolution.
We restrict resolution (i.e. acuity) by a blurscope (Fig. 4).

This is a modified unit-power telescope incorporating a diffu-
sion screen. (A complete description will be published else-
where.) We tested mobility at eight resolutions corresponding
to acuities of 6/2400 (20/8000) to 6/6 (20/20).
139
Figure 4. Subject wearing trial frames and contrast

reduction disk to restrict contrast.
In addition to the above restrictions, each normally sight-

ed subject's performance was also measured while blindfolded
to establish the level of nonvisual performance.
We measured mobility performance in two environments:

a laboratory maze and a shopping mall.
The maze was a long corridor (15 m long by 2 m wide)

cluttered with 19 vertical foam rubber columns, each about 2
m tall and 0.5 m wide. Walking through this maze is like
working your way through a stationary crowd of people. This
design is very convenient, offering a safe controlled mobility-
testing environment. While other studies have used indoor
mazes or obstacle courses to test mobility [5] they have all
been of a fixed design, and it is difficult to make the maze
long enough to get a significant number of errors in a single
passage by the subject. A particular advantage of the foam
rubber columns is that they are easily moved. We randomize
the positions of all the columns before every test. This
allows us to retest each subject many times, each time in a
brand new maze.
140
The other test environment is a shopping mall (Fig. 5a).

The subjects begin at one end of the L-shaped mall and are
asked to walk to the other end, a total distance of about 250
m.
For each environment we measure time and bumps. Time

is how long it takes them to do the task; bumps is the number
of contacts with obstacles plus any full stops.
3. Results
Figure 5a shows the shopping mall which we used for testing

mobility. Figure 5b shows the effect of restricting field to 7
degrees. Figure 5c shows the effect of restricting resolution
to 1.6 c/deg, equivalent to an acuity of 6/150 (20/500). Fig-
ure 5d shows the effect of restricting contrast to 1% of nor-
mal contrast.
Figure 5a: The shopping mall used for testing mobility.

141
Figure Sb: Field restricted to 7 degrees.
Performance was measured for the entire range of each

type of restriction. Blindfolded travel in the maze was much
slower (average of 80 s vs. 20 s travel time) and less accurate
(17 bumps vs 0.5 bumps). In both environments the level of
performance was nearly unimpaired for degrees of vision
from normal down to very restricted vision. Beyond this
point, performance worsened quickly to the level of blindfold-
ed travel. We summarize our results by this critical point,
the severest restriction at which performance is only slightly
impaired. Performance in the maze is only slightly impaired
by the following restrictions (critical points): 10 degree field,
resolution of 6/600 (20/2000), and 4% of normal contrast.
The corresponding critical points in the mall are 4 degree
field, resolution of 6/600 (20/2000) and 2% of normal con-
trast.
We wondered whether our subjects were making use of

auditory cues. It is well known that blind travellers can learn
to use echo location to sense nearby objects. Several of our
subjects volunteered that in the shopping mall they could ori-
ent themselves by the sounds of rustling bags of the people
142
Figure 5c: Resolution restricted to 1.6 c/deg, equivalent to an

acuity of 6/150 (20/500).
walking along the length of the mall. They said these audito-
ry cues were very helpful. To determine how much benefit
the subjects were obtaining from auditory cues we made a
pair of headphones which produce loud uncorrelated white
noise in the two ears, making the wearer effectively deaf.
We retested performance at each of the critical points in the
maze and mall with and without the headphones and found no
significant difference in time or bumps.
In order to help relate our results to actual low VlSlon

travel we did several supplementary experiments. Even if
one can walk up and down a mall there is little point in doing
so unless one can find specific stores in the mall and find spe-
cific products in the store. We have measured the visual
requirements (critical points) for finding stores at the mall.
They are similar to those for just walking up and down the
mall. The visual requirements for shopping in a supermarket
are higher, but still well below the criteria for legal blind-
ness.
143
Figure Sd: Contrast to I % of normal contrast.
4. Conclusions
Our results show that very little VISIon is required to walk

through indoor environments with reasonable accuracy and
speed.
Since the critical points are so low, far below the criteria
of legal blindness,. a remaining puzzle is why people who pre-
sumably have enough vision to travel with reasonable accura-
cy and speed in our tests still complain of mobility problems.
These people may be reluctant to travel partly because they
consider the risk of injury to be excessive. Our experiments
do not allow us to estimate the risk of injury, as no potential-
ly dangerous events occurred. This is partly because we
intentionally excluded the greatest dangers: drop offs (i.e.
stairs and curbs) and moving vehicles. We speculate that
many people with low vision could travel, but choose not to
simply because they feel it is too dangerous. If this is true, it
would be very important to determine whether this risk
144
assessment is accurate. People are notoriously bad at esti-

mating the probability of rare events. For example, many
people avoid flying, yet fail to wear their seat belt while driv-
ing, which is much more dangerous. There may be an impor-
tant role here for low vision mobility instruction in teaching
how to assess the safety of travel by the client in various
environments.
Although it is difficult to assess the probability of injury

faced by the low vision traveller, the danger is certainly real.
Walking off an unseen drop off or being hit by an unseen car
can be serious or fatal, especially for an elderly person, as
are most people with low vision. There are three approaches
to minimizing this danger: training, aids and environmental
modification. Our research suggests that training may be
important in helping the person assess risk. Of existing aids,
the most important are the telescope (used only while stand-
ing still because it is disorienting to walk with) and the long
cane (detects drop offs).
Are there any simple changes to the environment that

would make it safer for the low vision traveller? A striking
and unexpected result of the simulations is that self luminous
objects, such as neon signs and street lights, are exceedingly
resistant to loss of contrast and loss of resolution (Figs. 5c
and 5d). It seems likely that the danger of drop offs could be
greatly reduced or eliminated by marking the upper edge with
light bulbs. It is hard to imagine installing lights outdoors
along every curb of every city block, but in indoor public
spaces, such as shopping malls, it may be worthwhile to
install a strip of lights along the top edge (of just the top
step) of every stairway. Once the traveller is aware that a
drop off is nearby he or she can slow down and negotiate
safely.
5. Acknowledgements
Some of these results have been reported at the annual meet-

ings of the Association for Research in Vision and Ophthal-
mology and of the Optical Society of America. Jim Serio,
Lois Applegate, and Preeti Verghese acted as research assis-
tants, running most of the subjects, and suggesting some of
the experiments. Dan McLaughlin, Orientation and Mobility
145
Supervisor of the Syracuse Lighthouse, was a consultant to

this project and made many useful suggestions. This research
was supported by NIH grant EY04432.
146
6. References
1. A. Colenbrander: Dimensions of visual performance.

Trans. Am. Acad. OphthaZmoZ. OtoZaryngoZ. 83, 332
(1977)
2. G.E. Legge, D.G. Pelli, G.S. Rubin, M.M. Schleske:

Psychophysics of reading I. Normal vision. Vision Res.
25, 239 (1985)
3. G.E. Legge, G.S. Rubin, D.G. Pelli, M.M. Schleske:

Psychophysics of reading II. Low Vision. Vision Res.
25, 253 (1985)
4. J.A. Marron, I.L. Bailey: Visual factors and

orientation-mobility performance. Am. J. Optom.
Physiol. Opt. 59, 413 (1982)
5. D.G. Pelli, G.E. Legge, M.M. Schleske: Psychophysics

of reading m. A fiberscope low-vision reading aid.
Invest. OphthaZmol. Vis. Sci. 26, 751 (1985)
Visual Acuity Deficits and Chromatic Aberration
in Pseudophakia
Stanislaw J. Rag, Charles W. White, Pierre Simonet
1. Introduction
Several previous studies have assessed the chromatic aberra-

tion of the human eye. For example, HOWARTH and
BRADLEY [1] found that the longitudinal chromatic aberra-
tion of a normal eye varied from approximately -1.5 D at 420
nm to 0.5 D at 660 nm relative to zero at 578 nm. These data
are consistent with those found by some other researchers
[2,3,4]. However, MORDI and ADRIAN [5] found somewhat
less chromatic aberration in the blue end of the spectrum
than previously reported, which they attributed to an artifact
of averaging across different ages in the previous studies.
MILLODOT [3] had already reported that chromatic aberra-
tion decreases with age. By testing aphakes in different age
groups, he showed that changes in the crystalline lens could
not account for the age-related changes in chromatic aberra-
tion.
Most studies of human chromatic aberration restricted

testing to the visible spectrum between 400 and 700 nm, but
WALD and GRIFFIN [6] extended their investigation to cover
the range from 365 to 750 nm. The chromatic aberration
found at 365 nm was -2.63 D and 0.62 D at 750 nm.
Most previous studies examined the combined chromatic

aberration of the crystalline lens and cornea in phakic observ-
ers, but IVANOFF [7] tested an aphakic observer and reported
that the longitudinal chromatic aberration was -0.43 D at 483
nm and 0.38 D at 712 nm with data normalized at 589 nm.
The chromatic aberration function may be computed for a
schematic aphakic eye. Figure 1 illustrates such a theoreti-
cal function calculated using Cornu's formula for the disper-
sion of the cornea and the ocular media.
The normal human eye possesses several mechanisms that

minimize deficits in visual acuity due to short-wavelength
148
0.5 ; - - - - - - - - - - - - - - - - - - - - - - ,
0-.-0-
0.0 - <>-0-<>-" _
..•.........•........................ .....~-.•............... ......................... _.
!V'~
0/
-0.5 ·········7 o .c.......•....
/,./
ChromatIc -1.0
•
························07"········· .......... _.............•...
AberratIon 0/
t==~-===::-------
(Dlopters) -1.5
-2.0
-2.5
300 400 500 600 700

Wavelength (nm)
Figure 1. Theoretical longitudinal chromatic aberration

of a schematic aphakic eye.
chromatic aberration: the crystalline lens absorbs near-

ultraviolet, the macular pigment absorbs short wavelength
visible light, and blue-absorbing cones are largely absent from
the foveal region of highest acuity. However, the aphakic
eye lacks the first of these mechanisms, the macular pigment
transmits relatively more highly in the near ultraviolet, and
the red- and green-absorbing cones have appreciable sensitiv-
ity in the near-ultaviolet spectral region. The case is the
same for pseudophakes with UV-transmitting IOLs.
WHITE, ROG and WILLIAMS [8] demonstrated that the

visual acuity of the pseudophakic eye was impaired by ultra-
violet light. They used two different light sources to illumi-
nate the visual acuity tests. Figure 2 compares the spectral
energy distributions of the light sources with a typical day-
light spectrum. As illustrated, the unfiltered Xenon arc lamp
approximates the daylight distribution, while the UV filter
cuts off the near-ultraviolet component below approximately
400 nm.
149
100 , ..... ----_........ -- ........ _................. _-_....... _-- ............ _.... -----_ ... - ...... -
.'
-'",
"..
~1.-.=.t'"-:",, " - ...
...."...-
10
.
I
I
>-
01
I
I
I
,,•
L..
<II
C
UJ
daylight
<II
> xenon
..... xenon & UV filter
t:I
<II 0.1
Il:
0.01
300 400 500 600 700
WAVELENGTH (nm)
Figure 2. Relative spectral energy distributions of typi-

cal daylight and a Xenon arc, with and without
a UV-blocking filter.
Pseudophakic observers were tested under lighting condi-

tions that either included or excluded the near-ultraviolet
spectral region. The results of a vernier acuity experiment
are shown in Fig. 3. Notice that for phakic eyes, with intact
crystalline lenses, the average vernier acuity is nearly the
same in the filtered and unfiltered conditions. That is, the
presence of ultraviolet illumination did not impair phakic
acuity. For the pseudophakic eyes, however, vernier acuity
thresholds were significantly reduced in the filtered condi-
tion. That is, blocking the ultraviolet apparently improved
performance on the vernier acuity task for pseudophakic
observers.
Visual acuity was also tested with a conventional eye

chart for the same observers. The magnitude of the ultra-
violet visual deficit was approximately one line on a Sloan
chart when the chart was illuminated by a mixture of UVand
visible light. One possible explanation for the observed acui-
ty deficit is longitudinal chromatic aberration, because eyes
150
Pseudophakic eyes
-
'0 Phakic eyes
c
2
-E
filtered unfiltered filtered unfiltered

Figure 3. Vernier acuity thresholds of pseudophakic and
phakic eyes, tested with and without a UV-
blocking filter. Lower vernier acuity thresholds
represent better visual acuity.
with UV-transmitting IOLS should exhibit relatively large

chromatic aberration in the near-UV spectrum.
151
2. Method
2.1 Observers
Three subjects aged between 66 and 68 years of age partici-

pated in this study. All observers had posterior chamber IOLs
implanted with intact posterior lens capSUles. None of the
observers who are reported here had UV-absorbing IOLs. All
implants were done by the same surgeon who performed com-
plete ophthalmological exams prior to the study. One of each
observer's eyes was tested.
2.2 Apparatus
Figure 4 illustrates the optical apparatus. A computerized

Badal type optometer with a 10 D quartz lens and a 3 mm
artificial pupil was used with a Xenon arc lamp and monoch-
romator. The movable target consisted of 2 thin vertical fila-
ments in an open frame.
CID Tllrget
Diffusion
Shutter Blldlli Screen Mono- Protective
" Lens
Artificilll
Pupil _
-"- I \ Chromator...--_.....,
Xenon Arc
Ltlmp
Computer
Figure 4. Schematic diagram of the computer-automated

Badal optometer. Observers adjusted the target
along the optical axis to the point of sharpest
focus.
152
2.3 Procedure
All observers were tested from 340 to 660 nm in random

increments of 20 nm. The data reported here include from 2
to 8 measurements at each wavelength for each observer.
The observer was positioned securely in a chin and fore-

head rest. A shutter prevented continuous exposure of the
target, while a laboratory microcomputer controlled a step-
ping motor to adjust the monochromator to a pre-selected
wavelength. The observer then pressed a switch to open the
shutter and moved the target along the optical axis to the
point of clearest focus. Closing the shutter ended the meas-
urement and enabled the computer to proceed to the next
wavelength. A computer mouse device was attached to the
movable target, which allowed the computer to register the
position of the target after each trial and to convert it to
diopters. Each measurement was printed immediately after
each trial.
3. Results
The individual results for three pseudophakic observers are

shown in Fig. 5. The large variability between observers at
certain wavelengths can be best explained by measurement
error due to the SUbjective uncertainty and the low number of
trials for some data points.
The average results for the three observers are compared

to several other functions in Fig. 6. The calculations for the
schematic aphakic eye were computed using Cornu's formula.
The single observation made by IVANOFF [7] agrees with the
theoretical function. Figure 6 compares the longitudinal
chromatic aberration measured in the three pseudophakic
eyes to the computed aphakic values and also to WALD and
GRIFFIN's average data for phakic eyes [6]. The pseudophak-
ic data were normalized at 580 nm. Notice that the chromat-
ic aberration measured in the pseudophakic eyes falls
between the theoretical aphakic values and the average phak-
ic data.
Longitudinal chromatic aberration in pseudophakia is

approximately the same as in normal eyes at long wavel-
153
Longitudinal Chromatic Aberration
1.00
0.50
0.00
... 51
-0.50
Diopters ..... 52
reo 580 -1.00
.. 53
-1.50
-2.00
-2.50 +--r---r---r---t---t---t---t---r--r--r--r--t--t---t---t--t
340 380 420 460 500 540 580 620 660
Wavelength (nm)
Figure 5. Longitudinal chromatic aberration of three

pseudophakes.
engths, but is significantly reduced at wavelengths below

approximately 500 nm. The residual aberration, in conjunction
with the ultraviolet transmissivity of most intraocular lenses,
may account for the reported ultraviolet visual acuity deficit.
At 340 nm, the theoretical longitudinal chromatic aberra-

tion of the aphakic eye was approximately -1.53 D, while the
pseudophakic value was -1.82 D. The results suggest that the
intraocular lenses examined in the present experiment exhib-
ited relatively less chromatic aberration than the crystalline
lens would have. Chromatic aberration remains a possible
explanation for ultraviolet visual acuity deficits in pseudo-
phakia, because the pseudophakic eye lacks the previously
mentioned mechanisms that tend to reduce the ultraviolet
impairment.
154
1.0 -
--
C
c
0.5
0
0
::i -0.5
~
L.
L.
eI)
.Q
-1.0
< ..... Pseudophaklc Eyes
U -1.5
::i .••• Theoretical Aphakic Eye
~
E
0
-2.0 - Average Phaklc Eye
L. o AphakiC Eye (Ivanoff)
.c -2.5
U
300 400 500 600 700

Wavelength (nm)
Figure 6. Mean chromatic aberration, with standard

error bars, of pseudophakic observers, compared
to aphakic and phakic data.
4. Acknowledgements
The current research project was supported by an operating

grant to the second author from the Natural Sciences and
Engineering Research Council. The authors express their
appreciation to Dr. Gordon Balazsy for selecting the pseudo-
phakic observers for the study.
155
5. References
1. P.A. Howarth, A. Bradley: The longitudinal chromatic

aberration of the human eye, and its correction.
Vision Res. 26, 361 (1986)
2. R.E. Bedford, G. Wyszecki: Axial chromatic aberra-

tion of the eye. J. Opt. Soc. Am. 47, 564 (1957)
3. M. Millodot: The influence of age on the chromatic

aberration of the eye. Graefes Arch. Clin. Exp.
Ophthalmol. 198, 235 (1976) ,
4. C. Ware: Human axial chromatic aberration found not

to decline with age. Graefes Arch. Clin. Exp. Ophthal-
mol. 218, 39 (1982)
5. J.A. Mordi, W.K. Adrian: Influence of age on chromat-

ic aberration of the human eye. Am. J. Optom. Physi-
ol. Opt. 62, 864 (1985)
6. G. Wald, n.R. Griffin: The change in refractive power

of the human eye in dim and bright light. J. Opt. Soc.
Am. 37, 321 (1947)
7. A. Ivanoff: Les Aberrations de l'Oeil: Leur Role dans

I'Accommodation (Revue d'Optique Theorique et
Instrumentale, Paris 1953)
8. C.W. White, S.J. Reg, T.T. Williams: Ultraviolet visual

acuity with intraocular lenses. Invest. Ophthalmol.
Vis. Sci. (Suppl.) 25, 450 (1984)
Measurement of Central Fields Following
Macular Degeneration
Stephen G. Whittaker, Roger W. Cummings
1. Introduction
In a recent presentation and following discussion among low

vision practitioners, we were rather surprised to find a near
consensus that central fields of someone with advanced
maculopathy provided little useful information in prescribing
visual aids and planning a rehabilitation program. Admittedly,
by the time Age Related Maculopathies (ARM) or fundus fla-
vimaculatus have advanced to the formation of an absolute
central scotoma, central fields are of limited diagnostic use.
However, other practitioners have argued that central fields
are valuable in predicting functional capabilities [1,2]; argu-
ments that have been supported in the laboratory by the
recent work of LEGGE et al.[3], who identified the state of
central fields as the best predictor of peak reading rate
among their diverse sample of low vision individuals.
Accurate central fields would be useful to the low vision

specialist in estimating maximum expected visual perform-
ance for tasks such as reading. Estimated maximum visual
performance would provide either a basis for discharging
patients who have achieved near optimum performance, or
justification for recommending continued rehabilitation servi-
ces for those who fall significantly short of optimum per-
formance. However, before recommending field testing, it
would be worthwhile to examine if such information might
represent an improvement over existing assessments.
Visual acuity has been found to be a significant predictor

of reading performance among a partially sighted population
[4]. Proper measurement of visual acuity is already a neces-
sary part of the low vision examination as this information is
valuable in selection of the appropriate visual magnification
aid [5,6,1]. Visual acuity also must relate to size of a central
scotoma. As average scotoma size increases, subjects are
forced to use increasingly peripheral retina for resolving
157
detail and it has been well known that visual acuity is

inversely related to retinal eccentricity of a target image [7].
In light of the expected correlation between visual acuity and
scotoma size, one might suspect that independent measure-
ment of central fields would add little to our ability to pre-
dict reading rate.
In this paper, we will show that contrary to this expecta-

tion, scotoma size can significantly enhance the reliability
with which the practitioner estimates prognosis and pinpoints
specific visual limitations to efficient reading.
A failure to appreciate the prognostic and diagnostic val-

ue of central fields in the clinic reflects serious methodologi-
cal problems in measuring the central fields of someone with
a central scotoma. Unstable fixation poses the major obsta-
cle to measuring reliable fields. Secondly, we will summarize
previously published objective measurements of eye move-
ments made while individuals with central scotoma attempt
to fixate a target and will discuss the limits fixation perform-
ance might impose on the reliability of central fields. Final-
ly, we will recommend methodological improvements in
assessing central fields of people with unstable fixation sec-
ondary to central scotoma that might significantly enhance
the clinical value of central field testing.
2. Methods
Following informed consent, a total of 39 eyes from 30 sub-

jects with macular degeneration were successfully evaluated.
Subjects had macular pathologies that were previously diag-
nosed by ophthalmoscopy to include fundus flavimaculatus,
atrophic ARM, exudative ARM or macular holes. Subjects
described herein had bilateral maculopathies, were usually
co-operative and motivated and had received rehabilitation
services. The sample was thus biased to present "best case"
performance.
Fixation eye movements were measured while subjects

attempted to view a variety of fixation targets projected at
1.13 m. Some subjects were asked to fixate different size
characters and to center the scotoma in a large, high con-
trast, luminous fixation cross on separate trials. The fixation
158
cross was intended to typify a customary method used to con-

trol fixation during field testing [6,8,1]. Using an objectively
calibrated search coil in a magnetic field eye tracking system
[9,10], horizontal and vertical eye position could be resolved
to better than 3 minarc over a ~O degree positional range.
Luminous characters (48-64 cd/m , 99% contrast) projected
on a dark screen were sized in accordance with the subject's
best visual acuity.
During the same testing ~ession, central fields were

measured (1 isopter, 206 cd/m , 4 minarc target) using a
dynamic technique that differed in two significant ways from
standard tangent screen procedures. Subjects directly con-
trolled whether a target position was stored as seen or not
seen by pressing a button. Secondly, an image stabilizing sys-
tem moved the projected target position on the tangent
screen in order to compensate for eye movements during field
testing [9,10]. The examiner could thus control the retinal
image position of the target independently of eye movements
and generate higher resolution fields than has been previously
possible with this population.
Visual acuities used for the data analysis were character

recognition acuities determined at a 1.13 m test distance.
The rear projected Snellen characters used as fixation targets
were also used to measure visual acuity.
The Pepper Visual Skills for Reading Test (VSRT) [11] was
administered to most of these subjects (21 eyes). In this test
subjects read aloud 13 lines of unrelated words or symbols and
reading rate (correct words/minute) measures were recorded.
The appropriate print size was chosen by presenting subjects
increasingly larger print sizes until they correctly identified
the single characters in the first line of the VSRT. Subjects
were tested monocularly and with their habitual visual aid for
reading.
In order to identify the relative importance of visual acu-

ity and scotoma size for predicting reading rate, a stepwise
multiple regression was performed. First, visual acuity was
correlated with reading rate and this source of variance
removed. Then scotoma size was entered into the regression
equation and the relative contribution of both variables to
total variance of reading rate was determined.
159
3.1 The Value of Accurate Central Fields
-
120
•
-
100
E
Co
~
80
W
t-
••
<C •
a: 60 • • •
~
z 40 •
C
<C •
w ••
a: 20 • • • ••
• •• • • • •
• •
0
0.0 1.0 2.0
VISUAL ACUITY (logmar)
Figure 1. Reading rate in correct words per minute as

determined by performance on Pepper Visual
Skills for Reading Test (VSRT) is plotted in
relation to visual acuity described as minimum
angle of resolution in LOG minarcs for a group
of eyes with various maculopathies. Zero read-
ing rate indicates an inability to correctly iden-
tify words and characters printed on the test
card.
A moderate correlation between scotoma size (log

minarc 2) and v~ual acuity (log minarc minimum angle of res-
olution) was found (r = 0.68). Unexplained variance could
probably be attributed to irregularities in scotomata shape;
islands or peninSUlas of functioning central retina or the vari-
ed effects of retinal pathology on peripheral retinal function.
160
120 Y
W 100
I-
::;)
Z
~
--
CI)
0
80 X
a:
0
;: •
iii
60 X
• • • ••
l-
« • Stargardt1s
a:
(!) 40
• atrophic S.M.D.
exudative 8.M.D.
z
C
« •• •X macular hole
•• •
w no scotoma
••
..• .
a: 20
• •~
I • I • * I
0 10 100 1000 AREA(deg 2)
~f
• ~o
I
50
I I 0 I 0
10 0 20 30 DIAMETER
(equivalent circle)
SIZE OF CENTRAL SCOTOMA
Figure 2. Rate of reading aloud unrelated words or char-

acters in correct words per minute (using the
Pepper Visual Skills for Reading Test) is plotted
in relation to scotomi size described as log sco-
toma area in degree. Individual points repre-
sent different eyes with maculopathy. The
diameter of a circle having the area indicated
on the lower abscissa provides an estimate of
average diameter of the scotoma.
Figure 1 illustrates that a higher correlation was

observed between scotoma size and reading rate than the cor-
relation between visual acuity and reading rate (Fig. 2).
When visual acuity was separately considered, it accounted
for 9.5% (r = 0.31) of the total variance in reading rate and is
similar to a correlation (r = 0.24) reported earlier for patients
with maculopathies [4]. These low correlations are less sur-
prising if one considers that the reading tests were adminis-
tered after the text was enlarged to compensate for reduced
visual acuity. When scotoma size was added into the regres-
sion equation, it accounted for an additional 30% of the vari-
ance reducing the variance accounted for by visual acuity to
less than 1%.
161
Accurate estimate of scotoma size itself would therefore

significantly enhance the practitioner's ability to predict
expected reading rates and the cost (at least in terms of time
and effort) of recovering higher reading rates. Moreover,
measurement of central fields and fixation performance
would, in some cases, significantly reduce rehabilitation costs
by pinpointing some aspect of fixation performance or a sco-
toma characteristic that is the primary performance limiting
factor.
A 20 degree scotoma size apparently marks an abrupt

increase in fixation variability [12] and also an increasing fre-
quency of low reading accuracy [10]. Problems in fixation
control have been overcome by at least some subjects and
might be alleviated in others by appropriate training proce-
dures. Identification of large (>20 degree diameter) scotoma
would permit identification of patients who are more likely to
benefit from training procedures intended to improve reliable
eccentric fixation.
In other cases, accurate measurement of central fields

have permitted diagnosis of specific causes for unexpectedly
low levels of performance [10]. For example, identification
of a central island of vision has explained high visual acuity
but very poor reading performance in one subject who habitu-
ally used this central island to fixate words but was unable to
see more than a few characters at a time. In other subjects
we have identified habitual fixation positions that position
the scotoma in the right field, obscuring text to the right.
These limitations could be overcome by teaching subjects dif-
ferent eccentric gaze angles for reading.
3.2 Limits on Central Fields Accuracy Imposed by Eye

Motion
As a customary clinical practice, the perimetrist relies on the

patient to maintain steady fixation in order to obtain reliable
visual fields. However, since the retinal image position of a
steady target will move to the same extent that the eye
moves, the variability in estimating an isopter point can be no
less than the variability of fixation. Attempts have been
made to help subjects hold their eye still, usually by asking
them to center a large fixation cross in their central scotoma
[6,8,2]. We examined whether use of a large fixation cross in
fact stabilized fixation.
162
Nine subjects performed various fixation tasks. These

subjects attempted to fixate the center of a large fixation
cross with their scotoma, or on other trials fixate different
size characters so that they could be seen most clearly. No
systematic differences in the variability of eye position dur-
ing these 12 second fixation trials were evident. Although
these different targets might have produced subtle, statisti-
cally reliable, differences in fixation variability, they were
neither large nor reliable enough to be of practical signifi-
cance.
Our measure of fixation variability was a descriptive sta-

tistic that characterized the total retinal area where the tar-
get image spent the highest relative amount of time accumu-
lating to 68% of the fixation period [9,10]. Fixation
variability increased w~h scotoma size, up to an area approx-
imately 3,200 minarc with scotoma sizes less than 20
degrees average diameter [12]. Frequently during fixation,
the target image positions of highest dwell times clustered on
generally contiguous retinal areas forming a reasonably well
defined locus of fixation. When scotoma sizes exceeded 20
degrees, fixation variability sharply increased upwards to
10,000 minarc ,and did not utilize a well defined retinal area
for fixation.
When a subject used a single well defined area on the ret-

ina for fixation, the linear extent of this dwell area was usu-
ally 1-2 degrees. Some subjects, however, would reliably
position a target image in two or more retinal areas that
were several degrees apart. If this occurred and the same
areas were consistently used, our descriptive estimate of the
variability might increase only slightly. The range or linear
extent of this dwell area describes these fixation shifts (Fig.
3). Apparently, even among our generally highly co-operative
subjects, the eye often shifts by more than 10 degress during
fixation. Figure 3 illustrates an estimated large linear extent
of these fixation dwell areas were exhibited by patients with
exudative ARM, and Fundus Flavimaculatus, pathologies that
had created the larger (10-20 degrees) central scotoma.
These subjects were not, however, specifically told to hold
their eyes still. Normally sighted subjects with simulated
central scotoma up to approximately 20 degrees exhibited
very similar eye position variability when asked specifically
to hold their eyes as still as possible. They did not, however,
exhibit the large fixation shifts that were common among our
low vision subjects. We found that during field testing: some
patients, who exhibited more than one reliable fixation angle,
163
EYE POSITION RANGE
1000
U)
o
I
"-
tIS
c: 100
E
10+-----~------~----~------~----~
Atr. Exu. M.H. Star.
Maculopathy
Figure 3. Linear extent of eye position changes (minarc)

during fixation of a single character at the acu-
ity limit, so that it was seen most clearly. Sub-
jects are grouped according to type of maculo-
pathy. To identify the preferred fixation areas
the retina was divided into a matrix (e.g. with
20 X 20 minarc cells) and the relative time the
target image spent in each cell was computed.
Cells containing the highest relative times were
tagged until 68% of the fixation period was
accumulated. The range is determined by the
greatest distance between tagged cells.
easily followed instructions to maintain one preferred eye

position. We suspect, therefore, that large shifts in fixation
angle could be significantly reduced if fixation eye move-
ments were monitored and subjects instructed to maintain a
preferred fixation angle.
Drift, or slow component of the fixation nystagmus were

also measured and found to be less than approximately 2
degrees/sec [12]. Some subjects exhibiting close to 2
degrees/sec drift velocities might have difficulty maintaining
164
visibility of a target during prolonged fixation, but only if the

target happened to drift toward poorly functioning retina.
These drift velocities were not high enough to degrade visual
acuity [12]. Saccadic suppression associated with more fre-
quent saccades necessary to compensate for higher drift
velocities might add some variability to static perimetry
data.
Our data suggest that the practitioner might be able to

obtain reasonably reliable single isopter fields in patients
with scotoma less than 20 degrees average diameter without
the currently commercially unavailable image stabilizing
equipment necessary to compensate for unsteady fixation.
Image stabilization, however, is necessary for full threshold
fields, or more psychophysically sophisticated assessments of
localized retinal function [13] that require repeated retinal
stimulation within an approximately 5 degree area.
3.3 Improved Measurement of Central Scotoma
Firstly, it is necessary to continually monitor fixation. This

can, of course, be accomplished with an eye tracker, but
attentive monitoring of a magnified view of the pupil using a
telescope with cross hairs would suffice to pick up about a 1
degree shift in gaze. Shifts in head position pose a probem
with this technique since they might mask real gaze shifts or
produce apparent gaze shifts when the pupil or corneal specu-
lar reflection is tracked. The latter is more common. Head
movements can be nearly eliminated by testing patients in a
supine or reclined position. Testing could be interrupted and
the subject corrected every time an apparent shift in fixation
is detected. If capable, the co-operative subject will quickly
learn to maintain a gaze position. This general approach can
be facilitated by use of a Canon non-mydriatic ophthalmo-
scope perimeter or a scanning laser ophthalmoscope, as these
devices allow one to directly and continuously monitor the
retinal image position of the targets.
Our fixation data suggest that this approach might yield

fields that are, at best, accurate within approximately 2 - 4
degrees with a co-operative subject who is capable of main-
taining a single fixation position (e.g. has central scotoma of
less than 20 degrees diameter). Since accuracy is limited by
fixation, one could use a bowl perimeter such as the Hum-
phery Vision Analyzer or Goldmann Perimeter that is
165
equipped with a fixation monitoring telescope, without a sig-

nificant loss in resolution.
Secondly, we recommend that, rather than fixation cross-

es, subjects be asked to fixate high contrast single characters
so that they are seen most clearly. The characters could be
positioned to the side so that the scotoma would project cen-
trally on the perimeter screen. Although there are no data to
date that indicate more stable fixation would result such a
fixation target field testing might produce more valuable
clinical information. This procedure would indicate the gen-
erally preferred fixation position each patient might adopt.
Moreover, failure to obtain reliable fields would indicate a
functionally significant breakdown in fixation control; a prob-
lem that should be given priority in the rehabilitation pro-
gram.
In summary, our data and that of others indicate that

precise measurement of central fields might yield informa-
tion that would be quite useful to the low vision practitioner
and investigator. Unfortunately, fixation eye movements of
patients with absolute central scotoma significantly limit the
accuracy of central fields. We look forward to the first rea-
sonably priced, commercially available, field testing system
that measures and compensates for eye movements. In the
meantime, our findings suggest that the skilled perimetrist
can significantly improve field reliability with currently
available equipment.
166
4. References
1. J.E. Lovie-Kitchin, K.J. Bowman: Senile Macular

Degeneration: Management and Rehabilitation (Butter-
worth, Boston 1985)
2. I.L. Bailey: Visual field measurement in low vision.

Optom. Mthly. 69, 697 (1978)

Psychophysics of reading II: Low vision. Vision Res.
25, 253 (1985)
4. C.C. Krischer, M. Stein-Arsic, R. Meissen, J. 2ihl:

Visual performance and reading capacity of partially
sighted persons in a rehabilitation center. Am. J.
5. E.E. Faye: Clinical Low Vision. (Little, Brown, Boston

1976)
6. E.B. Mehr, A.N. Freid: Low Vision Care (Professional

Press, Chicago 1975)
7. T. Wertheim: Uber die indirekte Sehscharfe. Ztschr.

Psychol. 7, 172 (1894)
8. J .E. Kitchin: Assessment of visual functions of

patients with senile macular degeneration. Aust. J.
Optom. 64, 176 (1981)
9. S.G. Whittaker, R.W. Cummings: Redevelopment of

fixation and scanning eye movements following the
loss of foveal function. In Development of Order in
the Visual System, ed. by S.R. Hilfer, J.B. Sheffield
(Springer, Berlin, Heidelberg 1986)
10. R.W. Cummings, S.G. Whittaker, G.R. Watson, J.M.

Budd: Scanning characters and reading with a central
scotoma. Am. J. Optom. Physiol. Opt. 62, 833 (1985)
11. J. Baldasare, G.R. Watson, S.G. Whittaker, H. Miller-

Shaffer: The development and evaluation of a reading
test for low vision macular loss patients. J. Visual
Impairment & Blind. 80, 785 (1986)
167
12. S.G. Whittaker, J.M. Budd, R.W. Cummings: Eccentric

fixation with macular scotoma. Invest. OphthaZmoZ.
Vis. Sci. (under review)
Detection of Visual Field Defect Using
Topographic Evoked Potential in Children
Peter K.H. Wong, Roberto Bencivenga, James E. Jan, Kevin

Farrell
1. Introduction
The interpretation of visual evoked potentials (VEP) suffers

from a lack of objectivity due to several causes. First there
is as yet no good biological model explaining the relation
between the physiological activities occurring in the brain
and the voltage variations measured during the YEP. This
means that no mathematical standard for a "normal" reading
has been set, nor a normal range of variation against which to
test a YEP under investigation. There are certain features
whose presence or absence are considered relevant to the
analysis, like peak morphology and latency features, but these
are difficult to code mathematically and it is not clear at all
what their statistical distributional properties are, both in the
healthy population and in specific types of illnesses.
Other features which may seem more amenable to math-

ematical coding, like the latency and amplitude of the main
occipital positive peak, need a subjective element for their
identification and, moreover, have proved to have variable
discrimination power in the study of patients with visual
field defects [1]. Increasing the number of scalp electrodes
has provided additional information of the relationship of
electrical activity between different brain regions, and has
been found useful [l].
In this study we have therefore tried to concentrate on

methods which may be justified on a theoretical basis,
accepted on an empirical one, and could be treated in a man-
ner as objectively as possible. The basis of comparison was
both the clinical localization of the defect and the visual
interpretation of the VEPM. It is hoped that a statistical pro-
cedure could be developed which does not require any subjec-
tive judgement.
169
2. Materials and Methods
The patient population consisted of 12 children (age 1 to 10

years, median 3 years). All were examined with particular
emphasis on their neurological and visual systems, including
visual field examination at the bedside. All 12 patients had
unilateral homonymous hemianopia on clinical examination.
The etiology of the field defects varied among the patients.
For three of the patients we had mUltiple VEPMs performed
at various points during the clinical recovery. These data
were only used to further test the techniques analyzed.
The control group consisted of 23 normal subjects (age 6

to 18 years, median 10.8 years) with no history of neurologi-
cal disturbance or visual defect.
The subjects were not sedated and during testing laid

supine in a darkened room. The strobe unit was placed 10
inches in a direct line of sight from the patient's closed eyes.
Continuous effort was made to ensure alertness. Collodion
electrodes were used, with impedence less than 3K Ohm.
Twenty simultaneous channels (International 10-20 system,
including Oz) were available to acquire data for 512 ms after
each flash stimulus. The input was led to a 21 channel
electroencephalogram (Nihon Kohden Corporation model
4221, Irvine, California) and its output digitized at 500 Hz by
a dedicated microcomputer {Bio-logic Systems Corporation
model Brain Atlas, Northbrook, lllinois}. Two hundred stimuli
were averaged with automatic artifact rejection to form a
single average evoked potential. Visual interpretation was
done without prior knowledge of any identification or clinical
information. The VEPM data was transferred into an ASCII
file and analyzed with the methods described below by using
the Systat statistical package (Systat Inc., Evanston, Illinois).
The problem that seemed most tractable for our study is

that of symmetry, namely, whether the two sides of the brain
react in a similar way to the incoming symmetric stimuli.
Several statistics were constructed.
170
2.1 Correlation Coefficients:
The Pearson correlation coefficient was computed for the

data from the S homologous electrode pairs: F p 1-!"'PZ, F3-F4,
F7-FS, C3-C4, T3-T4, P3-P4, TS-T6, and 01-0Z' The control
group was used to construct one sided 9S% normal confidence
intervals. Values below the lower limit of such interval were
considered abnormal.
2.2 Multiple Correlation:
Multiple correlation coefficients were computed for each of

the lateral occipital channels with respect to the adjacent
channels (01 with respect to TS, P3, Pz and 0z; 0z with
respect to PZ, P4, T6 and Oz. A twosided 9S% confidence
interval was constructed using the normal subjects and then
the values obtained for the patients were tested against such
interval. If a strong local source was present near 011 say, a
high spatial gradient would exist, giving a lower correlation.
Conversely, a higher correlation value would result from a
low spatial gradient as a result of volume conduction from
distant generators.
2.3 Mahalanobis Distance:
Each EP series of ZS6 points was divided into S time bins of

64 ms, with a single average voltage computed per bin. This
produced response vectors from which the mean value and the
covariance matrixes were estimated, separately for the
patient and control groups. The Mahalanobis distance [3] of
each response from both groups was then computed and the
YEP classified to belong in the group with the smaller dis-
tance. In order to respect the independence and the meaning-
fulness of the data we used only the left channel (01) for the
control subjects and the abnormal side for the patients.
2.4 Cross-correlation Analysis:
This type of analysis was limited to the 01-0Z pair. The cor-
relations between 01 and 0Z, in this order, were computed at
lags varying from -ZO to +ZO (i.e. -40 to +40 ms) and the fol-
lowing set of statistics was constructed for each YEP:
171
- the lag at which maximum correlation occurred;
- the value of the correlation at lag 0;
- the ratios of the correlation at lag 0 to those at lag -20

and +20 independently (skewness ratios).
These were all taken as non parametric statistics, as no

assumptions were made about their distribution. The two rat-
ios presented an additional problem, namely that for some
EP's the correlation at one of the three lags considered was
negative. In order to maintain interpretability we assigned to
these ratios a value of 0 whenever the correlation at lag 0
was negative and a value of 100 whenever the correlation at
lag 0 was positive but was negative at the corresponding
extreme.
3. Results
Visual interpretation resulted in 11/12 (92%) VEPMs being

classified as abnormal, while 1 was classified as normal.
3.1 Correlation Coefficients:
The set of correlation coefficients provided some indication

of the nature of the VEP investigated. Of the 23 control sub-
jects 13 had all correlations within the confidence limit, 7
had one abnormal value, 2 had two abnormal values and 1 had
three. Of the 12 patients 1 had 1 abnormal value, 1 had them
all abnormal and the remaining had between 3 and 7 abnormal
values, mostly in the occipital region (see Table 1).
Multiple correlation produced disappointing results. The con-

fidence interval obtained from the control group was too wide
and close to unity to provide any discrimination value for the
patient group (mean 0.987, S.D. 0.011, 95% interval 0.965 to
1). Of the 46 values for the control group 2 were classified as
abnormal. Of the 24 values for the patient group only 5 were
classified as abnormal and two of these referred to the side
which, on clinical examination, had proved less abnormal.
172
Table 1
Summary of statistics for correlation coefficients.
Ch. ~ Mean ~ Lower limll JI. of ~ below limit

CONTROL PATIENT
Fjl1- Fp2 .975 .028 .928 2/23 5/12

3- F 4 .925 .088 .780 3/23 2/12
F 7 -F 8 .776 .249 .366 1/23 5/12
T 3 -T 4 .827 .148 .583 2/23 10/12
C3 -C 4 .923 .071 .806 1/23 6/12
P 3 -P 4 .891 .100 .726 1/23 9/12
T 5 -T 6 .812 .122 .611 2/23 10/12
°1-°2 .923 .082 .788 2/23 10/12
The Mahalanobis distance method provided only one misclas-

sification in the control group, but only 7 of the 12 patients
were correctly classified as abnormal.
3.4 Cross Correlation:
Based on the values obtained from the control group we set

the following ad hoc discrimination rule. A reading was con-
sidered abnormal if:
- the maximum correlation was at a lag lower than -5 or

higher than +5 or,
- the correlation at 0 was lower than 0.6 or,
- one of the skewness ratios was lower than 1.5.
With this criterion all control subjects were classified as

normal and all patients in the study group as abnormal except
one (92%), the same accuracy as visual interpretation. The
exception was caused by a type of abnormality not revealed
173
by the procedure, namely the patterns of the two channels

were well matched in phase, but quite different in amplitude.
4. Discussion
We were initially concerned about the age difference between

the patient and control group. However we believed that the
type of features under study (mainly correlations) would not
be affected by age, unlike some latency values or morphologi-
cal features. In fact an inspection of all the data revealed no
systematic difference in the quantities analyzed between low
and high age subjects.
4.1 Correlation Coefficient:
While correlation coefficients are the first choice for an

analysis of this type and despite the fact that they did prove
useful, some technical considerations would suggest intrinsic
limitations.
First of all, for a normal subject one would expect the

value of each such coefficient to be close to 1. In this case
the distribution of the sample corerlation, even under the
usual assumptions of normality and independence of the data,
tends to normality very slowly [4]. In fact it is not asymptot-
ically normal if the true coefficient is exactly 1. This means
that one may not correctly use normal confidence intervals to
test individual readings. This theoretical fact was confirmed
by our data: the histogram of the correlation coefficients for
the control group was quite skewed and normal confidence
intervals failed to create a convincing division between the
two groups.
It is perhaps worth mentioning that even if we had used a

non-symptotic distribution theory for these coefficients, an
"abnormal reading" would tell us very little about the nature
of the abnormality, since this may be due to a phase shift,
reversal, or unilateral low amplitude reading. Similarly, an
abnormal result with large slow waves and missing peaks,
which would present as a featureless reading of normal ampli-
tude, may provide a high correlation coefficient, thus mask-
ing a striking abnormality together.
174
There is always the possibility of spurious findings on the

basis of repeated or multiple statistical tests applied to the
data [5]. This difficulty may be overcome by the use of a
proper multiple comparison procedure [6].
The lack of usefulness of multiple correlation coefficients is

likely caused by the great variability present in the data and
by the variety of causes which may generate large values for
this coefficient. This problem, which is already noticeable in
the simple correlation case, became overwhelming when deal-
ing with multiple correlations. No simple solution was found.
The method based on calculation of the Mahalanobis distance

can be, in our opinion, quite effective. Several technical fac-
tors limited its use in the current study. First it requires
estimation of the covariance matrix for both control and
patient groups. Hence if we want to look at a fine time axis
subdivision {i.e. bins of 8 ms or less} a large number of cases
is needed in each group in order to obtain reasonable esti-
mates. With the current number of cases we could only
divide the total epoch into 64 ms bins and therefore some
precision was lost. Short of obtaining a much larger number
of VEP, one could limit attention to part of the VEP, say 80 -
200 ms. or chose as response variables some of the morpholo-
gical features of the VEP, suitably quantified. This last
option however would reintroduce the element of SUbjectivity
that we are trying to eliminate.
Finally, it would be better to use one set of data to esti-

mate mean vectors and covariance matrices and another to
test the reSUlting procedure, but again this was not possible
with the number of cases available.
175
4.4 Cross Correlation:
The analysis of cross-correlations proved very effective,

despite the degree of arbitrariness that it required. It uti-
lizes correlation coefficients in a non parametric, and so
more acceptable way, and was developed based on the follow-
ing considerations. In an ideal normal subject one would
expect a fairly high positive correlation at lag 0, due to sym-
metry, and rapidly decreasing values as the two series are
shifted with respect to one another, due to the richness of
features of the EP. On the other hand if one side is delayed
with respect to the other one would notice a maximum corre-
lation at some lag different from 0, corresponding to the
phase (latency) shift. Further, a unilaterally low amplitude
result (e.g. unilateral cortical destruction) would generate
low values at all lags, while a symmetric but featureless EP
{which may indicate a bilateral disorder}, would decrease
quite slowly at either side of O.
The choice of the value 100 for the case of negative cor-
relation at one extreme should not cause any concern, again
because we are treating these ratios non parametrically; a
value of 30 is just as positive a finding as 60, not half as
much.
The problem of the lack of sensitivity to amplitude asym-

metries has to do with the calculation of cross-correlation.
Local variations of slope {i.e. small peaks or troughs} are de-
emphasized. We believe that these shortcomings may be cor-
rected by the use of further statistics which emphasize local
morphological features.
If the main occipital peak {DEF component}[7] is the

dominant peak in the EP, a maximum correlation achieved at
a large negative lag accompanied by a low but positive first
ratio, may be a strong indication of a left side abnormality,
based on a larger latency. This however requires subjective
know ledge of the EP morphology.
More generally, the use of all four variables considered

may provide a better understanding of the nature of the
asymmetry than the correlation coefficient alone.
The statistics we have considered were aimed at detect-

ing asymmetries and none could clearly identify the abnormal
side. However we believe that the Mahalanobis distance has
176
the potential to do so, once the optimal division of the series

is identified and a sufficient number of cases is available.
5. Conclusion
Our preliminary study suggests that a useful approach to the

objective interpretation of a YEP is in terms of cross-
correlations. In the small sample studied, this method gave
the same accuracy as visual interpretation of the VEPM. The
statistics we chose seemed to discriminate quite effectively
between control and patient groups. Further work along
these lines and a better understanding of the statistical prop-
erties of the variables involved seems worthwhile. In particu-
lar it will be crucial to validate that the set limits do in fact
represent threshold values, and are not artifacts of our small
set of data.
177
Table 2
Mahalanobis distance.
Individual controls' distance from: Individual patients' distance from:
control Patient control Patient

group group Classif. group group Classif.
2.410 3.450 n 3.810 4.600 n

6.100 48.990 n 4.290 4.990 n
3.100 22.110 n 66.740 10.000 a
7.500 129.890 n 29.330 9.780 a
7.290 46.460 n 11.810 9.950 a
4.910 17.920 n 4.550 4.920 n
4.480 7.260 n 68.700 7.960 a
3.100 48.530 n 6.310 8.420 n
9.290 568.860 n 22.130 8.590 a
13.790 392.200 n 6.880 5.290 a
1.400 12.540 n 6.810 7.970 n
3.790 3.070 a 10.680 5.530 a
3.550 11. 620 n
7.590 45.470 n
6.630 82.310 n n ~ 5/12
19.870 142.760 n
9.430 486.660 n a 7/12
8.700 47.060 n
13.470 158.030 n
15.380 172.970 n
10.820 185.250 n
10.920 47.130 n
2.470 76.300 n
n = 22/23
a ~ 1/23
n = individual is closer to the control group.
a = individual is closer to the patient group.

178
Table 3
Cross-correlation study
Lag of max. Correlation Ratio with Ratio with

correlation at lag 0 corr. at -20 carr. at +20
CONTROL GROUP
o 0.958 2.047 2.777

-1 0.975 1.509 1.908
o 0.928 1.657 2.812
o 0.976 1.852 2.509
o 0.736 3.242 3.472
o 0.916 5.234 1. 722
o 0.857 12.243 100.000
o 0.874 28.194 16.491
2 0.954 100.000 954.000
o 0.949 100.000 100.000
-1 0.929 1.621 2.617
o 0.989 12.519 6.774
o 0.972 2.467 1.873
o 0.973 1.954 1.900
-2 0.922 1.592 2.499
-1 0.984 1.922 2.491
o 0.939 1.912 1.940
2 0.898 10.090 4.157
o 0.961 1.806 2.164
1 C).958 1.808 1. 797
-1 0.969 16.424 100.000
o 0.971 1.994 2.111
1 0.648 2.455 7.714
.•
PATIENT GROUP
17 0.185 1.063 0.564

o 0.296 10.963 100.000
*
·•
-20 0.387 0.531 1.155
-10 0.500 1.018 100.000
-1 1.925
·••
0.928 1.291
-2 0.862 1.626 1.523
20 • -0.248 0.000 0.000
20 • -0.190 0.000 0.000
-20 * -0.230 0.000 0.000
-20 0.279 0.398 100.000
-20 0.079 0.140 100.000
-17 0.476 0.815 23.800
The asterisks indicate values beyond the set limits.

179
6. References
1. D. Regan: Evoked Potentials in Psychology, Sensory

Physiology and Clinical Medicine. (Chapman and Hall,
London 1972) p. 173.
2. K.A. Kooi, R.E. Marshall: Visual Evoked Potentials in

Central Disorders of the Visual System. (Harper and
Row, New York 1979) pp. 3-5.
3. D.F. Morrison: Multivariate Statistical Methods, 2nd

ed. (McGraw-Hill, New York 1976) p. 235.
4. M.G. Kendall, A. Stuart: Advanced Theory of Statis-

tics, Vol. 2, 3rd ed. (Hafner, New York 1969) p. 341.
5. B.S. Oken, K.H. Chiappa: Statistical issues concerning

computerized analysis of brainwave topography. Ann.
Neural. 19,493 (1986)
6. K. Godfrey: Comparing the means of several groups.

N. Engl. J. Med. 313, 1450 (1985)
7. R.E. Dustman, E.C. Beck: The effects of maturation

and aging on the wave form of visually evoked poten-
tials. Electroenceph. Clin. Neurophysiol. 26, 2 (1969)
Preliminary Study of Topographic Visual Evoked Potential
Mapping in Children with Permanent Cortical
Visual Impairment
P.K.H. Wong, K. Farrell, J.E. Jan, S. Whiting
1. Introduction
The assessment of cortical visual impairment in handicapped

children is often difficult and the diagnosis may be delayed
for several years, particularly when the child is difficult to
examine [1]. The striate cortex receives and transmits visual
information to the association areas. Thus it governs the
ability to resolve fine detail (visual acuity) and acts as a
gateway to conscious visual analysis. More complex visual
interpretation occurs in the association areas. Total destruc-
tion of the striate cortex results in loss of vision for con-
scious visual analysis. Destruction of the association areas
results in visual agnosia and such individuals see without
being able to recognise. Because of the similarities in their
visual behaviour, it is difficult clinically to distinguish
between profound visual agnosia and severe loss of visual acu-
ity in multihandicapped children.
A clear role has not been established for flash visual

evoked responses (VER) in children with cortical visual
impairment (CVI). Although there have been reports suggest-
ing that VER may be useful in the diagnosis of cortical blind-
ness in children [2,3], these studies were uncontrolled and
only 6 patients were examined in each study. In contrast, the
VERs in a group of 30 children with CVI did not differ signifi-
cantly from a control group of 31 children with normal vision
[4].
The above reports have described VERs recorded largely

from the occipital electrodes. Because vision may be affect-
ed by abnormalities in other regions of the brain, we com-
pared the VER recorded from the occipital electrodes with
the topographic visual evoked response map (VEPM) recorded
from electrode positions in 23 children with CVI.
181
2. Method
A diagnosis of CVI was made in 23 children (14 males and 9

females) who were first seen in the Visually Impaired Pro-
gram of the British Columbia's Childrens Hospital in 1985.
The criteria for diagnosis included severe visual loss, normal
or minimal ocular findings, and clinical, electrophysiological
and computed tomography (CT) evidence of a post-geniculate
abnormality. All children had a multidisciplinary evaluation
including examination by a pediatric ophthalmologist and neu-
rologist, and electroencephalogram (EEG), CT, VER and
VEPM studies. The 13 control children had a mean age of 8.0
years, no visual abnormalities, normal intelligence and a nor-
mal EEG.
The ages at time of the VEPM ranged from 1 to 13 years,

mean 7 years. The presumed etiology of the cortical visual
impairment related to perinatal abnormalities in 13 patients
(Table 1). All of the children had serious associated cognitive
and motor dysfunction (Table 2). Measurement of visual acui-
ty by conventional methods was impossible in most cases
because of delayed mental development and visual inatten-
tion. Hence visual acuity was classified as (a) no apparent
vision, (b) light perception or (c) ability to see objects at a
defined distance (Table 3). CT scan abnormalities were local-
ised and classified by a pediatric neuroradiologist (Table 4).
None of our patients had pure lesions of their striate cor-

tex or association areas. Nevertheless, based on the history
obtained from the parents and on clinical examination, it was
apparent that some children had more residual vision but very
limited visual interpretive skills while others were the oppo-
site. This suggested that the degree of dysfunction of the
striate cortex and association areas varied with each child.
The VEPMs were obtained using a special purpose micro-

computer (model Brain Atlas, Bio-logic Systems Corporation,
Northbrook, Illinois) connected to a 21 channel electro-
encephalograph (model 4221, Nihon Kohden Corporation,
Irvine, California). The 19 routine international 10-20 elec-
trode positions were used, in addition to Oz. Silver-silver
chloride disc electrodes were applied with coUoidon, referred
to linked mandibles and the impedance was kept below 3K
ohms. The high filter setting was 70 Hz and time constant
was 0.3 sec. The built-in strobe unit was placed 10 inches
182
Table 1
Etiology of permanent cortical visual impairment.

Perinatal
Toxemia and antepartum hemorrhage 2
Perinatal asphyxia 9
Intracerebral hemorrhage 2
Acquired
Trauma 2
Shunt failure 2
Degenerative disease 2
Venous thrombosis 2
Cardiac arrest 1
Meningitis 1
Table 2
Neurological abnormalities in 23 patients with cortical visual

impairment.
Mental retardation 16
Developmental delay 4
Learning disability 2
Epilepsy 16
Cerebral palsy 15
Obstructive hydrocephalus 4
directly in front of the eyes. Two runs each of 200 flashes

occurring l/sec were averaged using on-line artifact rejec-
tion. Patients were unsedated and kept as alert as possible,
183
Table 3
Visual acuity in cortical visual impairment.

No apparent vision 2
Light perception 3
Vision within 3 feet 8
Vision beyond 10 feet 7
Table 4
CT abnormalities in 23 children with cortical visual impair-

ment.
Occipital infarction
Bilateral 9
Bilateral plus temporal/parietal 1
Unilateral 1
Unilateral plus temporal/parietal 3
Periventricular leukomalacia
Bilateral 3
Bilateral plus temporal/parietal 1
Other
Temporal/parietal cyst 1
Normal 3
Not done 1
with eyes closed during stimulation. Data was digitized at

184
500 Hz for the 512 ms epoch. Calibration end correction of

interchannel differences of DC baseline and amplifier gain
was performed automatically. Linear interpolation and color
coding allowed creation of color topographic maps [5-7]. This
permitted a rapid sequential display of the entire epoch of
256 maps in chronological order.
The VEPM data were coded, mixed randomly with VEPM

data from patients with other disorders and interpreted with-
out any clinical information by an electroencephalographer.
The VEPMs were scanned to exclude recordings with techni-
cal problems. The following parameters were then noted:
latency, morphology and symmetry of the major components
[2], pattern of anterior spread of occipital activity particular-
ly from 120 - 200 ms; and inter-electrode amplitude and
phase relationships during this anterior progression.
The VEPM were classified into 3 categories according to

latency and amplitude. In type 1 complete the activity
recorded at the 0h 02 and Oz electrodes was absent or of
lesser amplitude than that recorded at the mid parietal, and
bilateral central and parietal electrodes reflecting volume
conduction. In type 1 incomplete, the occipital activity had
abnormal latency or amplitude, beyond 2 standard deviations
of control. In type 2, the occipital activity was of normal
amplitude and latency but the activity in the non-occipital
electrodes, particularly the parietal and central electrodes,
was absent or abnormal.
The R value, a measure of relative occipital activity, was

based on the ratio of averaged rectified activity at the 3
occipital electrodes to that of the mid parietal and mid cen-
tral electrodes. The R value, which is dimensionless and
independent of amplitude, is calculated by measuring the area
under the curve at each signal trace, with the same start and
stop times as that of the beg~ing and end of the main occi-
pital positive component (DEF) • The values obtained at each
electrode were expressed as a summated voltage and the
ratio calculated as follows:
R=[(OI + 02 + Oz}/3] 1 [(P z + C z ) 12]

185
3. Results
Analysis of the occipital data alone revealed abnQrmalities in

only 10 (43%) of the 23 patients. When data from all head
regions were analysed, the VEPMs were abnormal in all 23
patients.
Plates I, II, ill and IV show maps of the composite mean

data from the control group at 90, 100, 150 and 190 ms after
the flash showing the start, growth and anterior progression
of the main occipital positivity (DEF component). Plate V
shows the integrated (summated) map of the control group
from the beginning to the end of the occipital DEF compo-
nent.
Plate VI shows an integrated map calculated similarly

from a patient with a type 1 complete abnormality. This boy
had profound visual impairment and a mild cognitive dysfunc-
tion which related to severe perinatal asphyxia. Although he
had difficulty seeing stationary objects at more than 1 foot,
he could name colors, could follow but not describe moving
Objects and had good stereognosis. The CT showed bilateral
infarction of the striatal cortex. Topographically the type 1
complete group had generalised low activity.
Plate VII shows an integrated map from an infant with

adrenoleukodystrophy, a progressive neurodegenerative disor-
der. The CT scan was normal. The child appeared to have
normal visual acuity but almost complete visual agnosia. The
VEPM shows that most of the activity is in the occipital
region and the parietal/central activity is relatively less than
that seen in the control group.
As there were only 4 cases of pure VEPM type (two each

of 1 complete and 2), data of mixed type were grouped
according to their dominant type. Visual acuity and CT
abnormalities correlated with type 1 complete (p < 0.005) and
type 2 (p < 0.005) VEPM abnormalities (Tables 5 and 6).
The R values of the predominantly type 1 complete and

type 2 both differed from the control group when the non-
parametric Mann-Whitney U test was used (Table 7).
186
Table 5
Relationship visual acuity to type of VEPM abnormality.

Type 1 complete Type 2
No apparent vision 2
Light perception
Vision within 3 feet 6 1
Vision within 10 feet 1 4
Table 6
CT findings according to type of VEPM abnormality.
Predominantly Type 1 complete

Bilateral occipital infarcts 8
Bilateral periventricular leukomalacia 1
Right temporal/parietal cyst 1
Predominantly Type 2
Bilateral periventricular leukomalacia 1
Unilateral occipito-parietal infarct 1
Normal 2
Not done 1
187
Table 7
R values of patients with Type I and IT abnormalities.

NUMBER MEAN (SD) MIN MAX
TYPE 1 complete 9 0.95 (0.36) 0.62 1.83
TYPE 2 5 4.64 (0.86) 3.43 5.52
NORMALS 13 2.81 (0.73) 1. 64 4.00
4. Discussion
The diagnosis of CVI in children is often difficult. Because

they have multiple handicaps and poor visual attention, for-
mal visual testing is not possible in the majority of these chil-
dren [1]. In addition, children with involvement of the visual
association areas may have profound visual cognitive dysfunc-
tion but exhibit few of the classical features of blindness.
Thus many parents are unaware of the severity of the visual
impairment prior to assessment [1].
Traditional VER measures the amplitude and latency of

the response to a photic stimulus at the occipital electrodes.
Analysis of the occipital data alone revealed abnormalities in
only 10 (43%) of the 23 patients. This observation supports a
previous controlled study in which flash VEPs were not signif-
icantly different in children with CVI [4]. In contrast to the
traditional VER results, when data from all head regions were
analysed, the VEPM were abnormal in all 23 patients. In
addition, the type of VEPM abnormality correlated with the
anatomic localization in 17 (74%) cases and not one child had
a VEPM incompatible with the anatomic localization.
We believe that the type 1 abnormality occurs in patients

with injury to the optic radiation or to the striate cortex. It
appears that the VEPM can not distinguish between injury to
the optic radiation and the striate cortex. Furthermore,
when the damage to the striate cortex is profound, the
absence of occipital activity does not permit the normal
anterior progression to occur and damage to the visual associ-
188
ation areas can not be evaluated. The type 2 abnormality

appears to be associated with injury to the visual association
areas. On the other hand, the same "abnormality" may be
seen during drowsiness in normally sighted children. Thus a
type 2 abnormality is sensitive to the alertness of the patient
and must be interpreted with caution.
The R value is a measure of the relative activity at the

occipital electrodes as compared to the mid-parietal and mid-
central electrodes in response to a flash stimulus. As long as
the major positive occipital peak can be measured, R can be
calculated. We were able to identify the major positive occi-
pital peak in 22 of the 23 children. The R value permitted
the differentiation of patients with either type 1 complete or
type 2 abnormalities from controls.
In conclusion, the VEPM was more sensitive to cortical

visual impairment than the traditional VER method. The
VEPM appeared also to be helpful in demonstrating visual
association dysfunction when the optic radiation and striate
cortex were not severely affected. Finally, this method can
be used in children with multiple handicaps where the diagno-
sis of CVI may be difficult to establish on clinical grounds.
5. References
1. S. Whiting, J .E. Jan, P .K.H. Wong, o. Flodmark, K.

Farrell, A.Q. McCormick: Permanent cortical visual
impairment in children. Dev. Med. Child Neurol. 27,
730 (1985)
2. M.S. Duchowny, I.P Weiss, M. Heshmatolah, A.B. Bar-

net: Visual evoked responses in childhood cortical
blindness after head trauma and meningitis. Neurology
(NY) 24, 933 (1974)
3. A.B. Barnet, J.I. Manson, E. Wilner: Acute cerebral

blindness in childhood. Neurology (NY) 20, 1147 (1970)
4. Y. Frank, F. Torres: Visual evoked potentials in the

evaluation of cortical blindness in children. Ann. N eu-
rol. 6, 126 (1979)
189
5. P.K.H. Wong, R.E. Ramsay: Topographic EEG analysis

as a test of cortical function in cerebral ischemia.
Electroencephalogr. elin. Neurophysiol. 52, 5114
(1981)
6. D.L. Gregory, P.K.H. Wong: Topographic analysis of

the centrotemporal foci in benign Rolandic epilepsy of
childhood. Epilepsia 25, 705 (1984)
7. F.H. Duffy, J.L. Burchfiel, C.T. Lombroso: Brain elec-

trical activity mapping: A new method for extending
the clinical utility of EEG and evoked potential data.
Ann. Neurol. 5, 309 (1979)
Prescribing Magnification:
Strategies for Improving Accuracy and Consistency
Ian L. Bailey
1.
All low VlSlon practitioners develop their own individual

approaches to prescribing magnifying devices. The simplest
is the "cafeteria" approach in which the patient is presented
with an array of low vision magnifiers and asked to make a
selection. In time, the cafeteria will develop a corner delica-
tessan atmosphere with the proprietor proudly giving advice
based on the experience and comments of previous customers.
"Try this one, it seems stronger." "Try moving closer, that
often helps." "See if it looks better through the bifocal part
of your glasses." In contrast, trained professionals prefer to
use systems based on measurements of resolution ability,
magnification effects, tests of performance and judgements
of comfort and convenience.
Figure 1 presents a model that is virtually universal to all

systematic prescribing of low vision devices. The starting
point is always the case history which identifies the patient's
needs for magnifying devices. The history establishes the
resolution goals. For distance vision, resolution goals are
usually expressed in angular terms, such as visual acuity
scores, and these involve a specification that depends on both
the size of detail and the viewing distance. The clinician
makes estimates of the visual acuity required to perform par-
ticular tasks such as reading bus numbers from a reasonable
distance, reading signs on the other side of the street, seeing
faces on stage, etc. Specification of resolution goals for dis-
tance viewing tasks is made in terms of angular size because
usually angular size cannot be readily changed; it is not usual-
ly convenient to create significant changes in viewing dis-
tance.
For near vision, when resolution goals are considered, the

common clinical practice is to primarily consider the linear
size of task detail rather than its angular dimension. At near,
191
------~)I SET RESOLUTION GOALS I~(---------
points)
~--?
t----fail--
Figure 1. A basic system model for prescribing magnifi-

cation.
angular size can usually be changed significantly by simple

changes to the observation distance. Resolution goals are
typically expressed in terms of print size and the size of print
may reasonably be designated in either "M units" or "points".
The more commonly used near vision charts use typeset print
and they present sets of words, sentences or paragraphs.
Compared to letter charts, typeset charts present the patient
with much more congested tasks and they provide a much
more representative simulation of the kind of near vision
tasks that will be of prime importance to most patients.
192
Refraction is a vital procedure in the prescribing of mag-

nifying devices for either distance or near vision. The image
presented to the patient by any optical system should be rea-
sonably well focused for the observer's eye. While many low
vision patients are often not as sensitive to defocus, they are
usually much more sensitive than one would predict from con-
siderations of the tolerable blur circle diameter that would be
expected from the visual acuity scores. Having established
control of the proper focus of the retinal image, the clinician
proceeds to work with measures of resolution and degrees of
magnification in order to attain the resolution goals estab-
lished in the history taking.
In prescribing magnifiers for distance VlSIon tasks, the

visual acuity is first measured. Then the magnification is
calculated from the ratio of the measured visual acuity to the
established resolution goal. For example, it may be estimat-
ed that 6/12 (20/40) would be adequate for reading street
signs and so a patient with 6/60 (20/200) would require 5X
magnification. The clinician should next verify that a 5X tel-
escope does indeed afford a visual acuity of 6/12 (20/40).
When testing visual acuity with telescopes, the clinician
should be certain that either the telescope has an appropriate
range of focus adjustment or, if not, use special collimation
procedures such as placing the chart 4 m from the subject and
holding a +0.25 D lens over the objective of the telescope.
Next, the clinician should conduct real-world trials to ensure
that the patient can actually perform the required task with
the chosen level of magnification. Then it remains to choose
a telescope that first provides adequate magnification and
second provides an appropriate combination of other optical
factors (field size, exit pupil size, focus adjust ability, image
quality, and optical adaptability) and non-optical considera-
tions (weight, portability, convenience, appearance and cost).
For near vision tasks, the broad procedure is essentially

the same. After the history and refraction, resolution ability
is measured typically by determining the smallest typeset
print that can be read when the patient views the chart from
a convenient near distance under conditions that provide an
in-focus retinal image. The observation distance must be
known. Then, the magnification needed can be expressed by
the change in conditions required to achieve the resolution
goal. For example, for a patient reading 5 m print at 40 cm
there might be an established resolution goal of 1 m print.
The change in viewing conditions required may be expressed
193
as equivalent viewing distance so that an EVD of 8 cm (1/5 of

40 cm) is indicated here. Alternatively equivalent viewing
power needs to be increased from the starting point of 2.50 D
by a factor of 5 to achieve an EVP of 12.50 D. If one prefers
to use a traditional magnification rating, the initial EVP of
2.50 D is divided by 4 to give the initial magnification rating
of 0.63X, and this must be increased by a factor of 5 so that
the indicated magnification rating is 3.2X.
It should be verified that the patient does achieve the

desired resolution goal when the indicated change in viewing
conditions is provided. Typically this is done using spectacle
lenses to enable a clear focus to be obtained when the chart
is held at a distance that is closer by the appropriate amount.
At near, comparisons should be made of the relative merits of
high addition reading spectacles, hand held magnifying glass-
es, stand magnifiers, near vision telescopes, and video mag-
nifiers. Each of these has various advantages and disadvan-
tages.
Accuracy and consistency of prescribing come from hav-

ing sound systematic ways of predicting performance as
patients change from one observation system to another. In
the measurement of distance visual acuity, the clinician gains
best control and understanding if confident predictions can be
made about the resolution that the patient should achieve
with a given magnifier used under given conditions. Many of
the more widely used visual acuity charts have features that
restrict predictability. For many charts, moving the chart
closer to the patient or giving the patient a telescope changes
the nature of the task so that the patient is asked to read a
group of letters that may be larger in number, has different
spacing relationships and the size ratios between adjacent
rows may chaJlge substantially. These factors can have quite
significant effects on the visual acuity score obtained.
For systematic measurement of visual acuity, the task

presented to the patient should be the same at each size lev-
el. This means the number of optotypes or symbols per row
should be constant, the spacing between adjacent letters and
between adjacent rows should be proportional to letter size,
the size progression should use a consistent ratio, and the sets
of symbols should be equally difficult at each size level. Size
should be the only significant variable. Considering the
charts available today, only 3 satisfy the requirements for
good measurement [1-3]. The Bailey-Lovie, "ETDRS" and the
194
University of Waterloo charts all have five letters per row,

the spacing between rows and between letters are proportion-
al to the print size and the size progression follows a constant
ratio (i.e. the progression is logarithmic or geometric) (Fig.
2).
For grading visual acuity, counting fingers tests could be

used, as could a collection of samples of different size print
cut from yesterday's newspaper. But these are not really
standardized tasks. The so called "standard Snellen" chart
with the traditional "E" at the top and the fairly popular
Feinbloom chart both use symbols that have not been stan-
dardized for legibility and they grade vision rather than
measure it. Both of these charts also have erratic size pro-
gressions and the number of symbols differ from one size lev-
el to the next. The Sloan chart [4] made significant advances
towards achieving an optimal measuring system. Sloan stan-
dardized her family of 10 letters and insisted on a logarithmic
progression, but she failed to standardize the spacing and the
number of letters per row. The 1982 chart from the New
York Lighthouse is, at first glance, something like the Sloan
chart but it uses an irregular and idiosyncratic progression.
Figure 3 illustrates the size progression used in many of
today's charts on a logarithmic scale. The Bailey-Lovie,
ETDRS, and University of Waterloo charts follow a systemat-
ic size progression. The lines joining the dots in this figure
indicate that the tasks are essentially the same. All other
charts have irregular size progressions. As an example, the
Lighthouse chart has 5 rows in the 6/60 (20/200) - 6/36
(20/120) range, but if the chart were to be brought to 1.5 m
(or a 4X magnifier were to be used), this acuity range would
shift to the 6/15 (20/50) - 6/9 (20/30) region where there are
only 3 rows to cover the equivalent range.
Irregular size progression and tasks that vary from one

size to the next preclude systematic measurement. It is like
trying to measure distance with a ruler which has short "inch-
es" in some sections and long "inches" in others. Figures 4a,
b, c and d show four different pairs of charts. For each pair,
one chart is an enlargement (approximately 2.5X) of the other
and the two black dots indicate limits of equivalent opera-
tional ranges. These illustrate the rather profound changes in
the nature of the task that can occur with changes in magni-
fication unless the chart has the appropriate design features
as is the case in Fig. 4d.
195
..
160
EZHPV 09
38
125
DPNFR 08
30
100
RDFUV 07
2.
80 URZVH 06
19
" HNDRU 05
15 50 ZVUDN O.
12 40 V PHD E 03
9532
620 _ _ _ _ __
PVEHR "
7525 E H V DF 0-1
NUZF E ... _ 0
.::.: -=:;:::-
b) ~=D S R K N··
CKZOH
ONRKD
KZVDC
VSHZO
HDKCR
CBRHN
...
~i-----
SVZDK
NOVOZ
"MaDY
...... w
•.
~
_ _ _ _- 0 0
ii
Figure 2a-b. Visual acuity letter charts that use a geometric

(constant ratio) size progression and standardize
the task. (a) Bailey-Lovie (b) "ETDRS"
There are practical advantages that come from using sys-

tematic measurements of visual acuity. Changing the view-
ing distance or changing the magnification will alter thresh-
196
-
I S
- -
.A..
100
-i-
0
lO
l]f
C D
.=.
l2.
j 20
_!!i~~ .
HZ.
- - - !:!~~
K V RHN :::i
D H v::'
°D
1
I
K Nc
0 ••
~il
80
S o
0 I
R eN,
H
I R V V
RDr
00'6
c ~ 15
D K - 19
I V C Z
R
6
6
24
N -
30
I D 12
-..
6
li
I K
-
6
6
60
Figure 2c. (c) University of Waterloo
old size, but for best predictability one should use a

systematic measuring system with the task remaining con-
stant and the steps of size progressing so that each successive
row is smaller by a constant ratio. With appropriate charts,
the same amount of magnification will improve acuity by the
same number of rows, regardless of the patient's visual acui-
ty. It is often useful for clinicians to be aware of the differ-
ence between threshold size for maximum efficiency and the
threshold size for maximum resolution. Appropriate chart
design ensures that, if there is a three row difference
between the threshold for maximum efficiency and the
threshold for maximum resolution before magnification, then
there will always remain a three row difference after magni-
fication, regardless of the amount of magnification.
197
,
5 10
I , 40 8020
I
400 800
I
160 200
I I I I I I I
Feel
• ••••••••••••• B-L
• • • • • • • ......
•••••••••••••• ETDRS,UW
• • ••• • .....
••• • • F
••
• •••••• • •
·
• • • • • .............
•
LH
AMA
AO-E
I I I I I I
,
I I I I I I I I I I I MeIers
3 6 12 24 60 120 240
Scaling on Distance Visual Acuity Charts
Figure 3. Comparing size progression used in many of

today's charts. Lines connecting dots indicate
that the task is essentially the same across size
levels. B-L = Bailey-Lovie; ETDRS = chart
from Ferris et al.; UW = University of Waterloo;
F = Feinbloom chart from Designs for Vision
Inc.; L.H. = New York Lighthouse; AMA = AMA
chart from Bausch and Lomb; AO-E = American
Optical Chart 1937~
Testing VISlon at near most commonly involves test

charts with typeset print. There are some clinicians who base
their initial magnification estimate on a measure of letter
chart acuity. Some predict near vision magnification needs
from distance letter chart acuity, while there are others who
begin with a letter chart acuity measurement at near. For
many patients, especially those with macular disorders, there
is not good concordance between the acuity scores obtained
with letter charts and those obtained with typeset reading
charts. Magnification requirements at near should be based
on tests that involve typeset material, especially if the reso-
lution goal is related to the reading of printed matter.
Many of the near vision charts used in low vision today

have irregular size progressions. There are only three, the
Bailey-Lovie Word Reading Chart, the Keeler, and the North
Carolina charts [5-7] that use a regular geometric size pro-
198
-... ....-
--,'~ ..
a) ,·N C Z S"
H S N E ~E Z N~:NCZS-
·H S N E
· V K CO" -H C R·· V K CO·
z·.
'NRSDZ'
• N R S 0 e -S 0 . O··CSEOHK.
-N C V H·· ....... .
D.HKNR
C S E 0 H K .. _R K 0 E .: ::~: "M_
e E H K N R ,. • _ . ~_~~~
Z N S ".-C .......
--=::-~ '.~.' ' .
• VRNO MCDe "
ZVSN KRHV
KVRO CZEH
".HZ f(COO
b)
E
.0 B
DLF
"rZBDE
PTEO E-
CB"
DLF
OFLCTB PTJ:O
F • • DB
TPJ:OLI'DZ orLO".
Figure 4a-b. lllustrating the effect of magnification (here

2.5X) on 4 different charts. For each chart
pair, dots have been added to identify equiva-
lent operational ranges (selected arbitrarily).
For charts a, band c, magnification changes the
number of letters and spacing arrangements in
199
the observer's task. The design features of

chart d avoids this as the task is essentially the
same at all size levels.
gression. These are shown in Figs. Sa, band c. Most charts

use text or meaningful sentences, and there are a few, such
as the Bailey-Lovie Word Reading Chart and the Lebensohn
Chart, that use unrelated words. When meaningful text is
used as the test material, the length of the passage and the
contextual difficulty varies from one size level to the next,
but provided the typeface and the spacing arrangements are
consistent, the difficulty of the task can usually be consid-
ered to be reasonably equivalent from one size level to the
next.
Many near V1SIon charts used in low V1SIon work create

difficulties because of their limited range of sizes. Figure 6
illustrates the size progression used in seven different near
vision charts. Both the Keeler and Sloan charts were
designed for low vision work and while they include very large
print (10m), their smallest print is only equivalent to news-
print (1 m or 8 points). The Sloan reading cards do not follow
a systematic progression [8]. Since most patients, with
appropriate optical assistance, are enabled to read 1 m print,
they will be able to read all print on these charts and conse-
quently their resolution limit will not be determined. On the
other hand, the American Optical and Bausch and Lomb
charts shown in Fig. 7 have been designed for routine
ophthalmic clinical testing and while they do include smaller
print, they are very limited in the larger sizes and conse-
quently they are not well suited for use in the early stages of
assessment in patients with poor reading acuities. Charts
that use a constant size progression ratio ensure a consistent
number of rows between the size threshold for maximum
reading efficiency and the size threshold for maximum reso-
lution. Providing magnification devices or changing the view-
ing distance to improve the resolution threshold by two lines
will automatically create a two line improvement in the
threshold for maximum reading efficiency when such charts
are used.
The constant ratio size progress charts [5-7] shown in Fig.

5 also provide especially useful guidance about the magnitude
of the change in viewing condition required to achieve a cer-
200
c)
9 2 5
47 86 73
4 8 726-
.-
-62 3 4
935-
1467-
726· 51623
4 3 0 9 7
9 35· 615412
-8467· 4295317
~F N P R Z
d)
.... EZHPV
' ,
,.
11251
DPNFR
...,.
, ,
RDFUV
1'01" ·URZVH "iF N P R Z.
,OJ'
"=== HNDRU
ZVUDN ! DPNFR
"VPHDE R D F U v. .
PYEHR • URZYH
7!,~~' _ _ _ _ _ _ _ EuHM:,~F _ _ :. ==== HZ~~:NU =====
, 'MDI
..
: : , - - :.; . / - -
Figure 4c-d.
201
a)
Ridges .could
The)' could .... wisp of smoke of (~i"y
thousand acres on (he ocher. The sky
Amona UM: I reel and • StOuP of buildin,p rar

OU1
which IS not -...... _.__._-"-"' -

lhey Mid it
......... ,...
WI:S lheft: lhar lhe ttaiUlt """,I be
... _..................... ........... . .

............
Look down from
-_----_. _--....------_
..... _. "_ _ __ _ ....n._, .. _ ...., __
.... _ _1t
.... _- .. -_---
...... _ " ........ ~
.... _ _ fII .....
on either side of -. .. _. . .. - ...
II ;:.--..:;,:::=-:: II
~~_"~~W~
Clu _ _
The road Them upon marshes

where snowy reeds on
are track Sighed and ice crackled
the duck in the red sun
Ran mostly
.... . ICI[ ~ "'"
Quacked on the winter air. It

was a very strange sound. There
along high Wos a manh on one side of the ridge

and a forest which appeared to be on
Figure Sa. Reading charts that use a geometric (constant

ratio) size progression. (a) Keeler
tain degree of improvement in resolution. If the clinician

wishes to provide a 2 line improvement in resolution, then the
EVD, the EVP or the magnification rating of the viewing sys-
tem should be changed by two steps from within the same
number sequence. On these three charts the number
sequence of the size labels follows the geometric progression
1.0, 1.25, 1.60, 2.0, 2.6, 3.2, 4.0, 5.0, 6.3, 8.0, 10, 12.5, 16, 20,
etc. The multiplier is about 5/4. As an example, consider a
patient who can read 1.6 m print at 20 cm when using a 5 D
addition. To resolve 1.0 m print, a 2-step improvement is
required as this is 2 rows on the chart. From the sequence of
size labels on the chart, it can be seen that changing 2 steps
closer from the initial EVP of 20 cm would bring the equiva-
lent viewing distance to 12.5 cm. Alternatively, increasing
the EVP from 5 D by 2 steps brings about a new EVP of 8 D.
Clinicians who prefer to use magnification ratings would find
that the initial magnification rating of 1.25X (5/4) should be
increased by 2 steps to 2X.
202
perfection keel .- -- ,- _ ...

deep quietly ~~=";~:':';:..$
inquest yolk
lamentable meaning half
caution watchmaker mil
omit needlessly serious
------
~-=lD -::-2.':=:.T:...~
~":--=="-
Everyone worD on a farm
Dogs bark loudly.
Birds fly.
Figure 5b-c. (b) Bailey-Lovie Word Reading Chart; (c)

North Carolina Chart
• • • • • • • • • • • e--e--e e--e--e B-l

••••••••••• K
• •••••• • • S
••••••••••• • NC
• • • • • ••• •• • • l
• • ••••• AO
•• ••••••
I I I I I
Bal
1
0.2 0.4 O.B 1.0 2.0 4.0 B.O 10 20 M units
Scaling on Reading Charts
Figure 6. Comparing size progression of many of today's

reading charts. B-L = Bailey-Lovie; K = Keeler;
S = Sloan; NC = North Carolina; L = Lebensohn;
AO = American Optical 11970; B&L = Bausch
and Lomb.
203
mille cm READING CARD (Metric)

1.1
_-
4
It .... the month of May, The clear warm - .___ _"__ .__ "_ .. _..... ___0-,---.....
lIIIIl8bine lay upon the mOUlltain whieh bad ---'- ._.... _--_.... -'''- '- ''- - . _--..
tumed greeD again. Tbe laat IDOWII bad
----.----- • __ __._ ... _..•.
~ ~~
till
cY~",,:rl'!kf,,~...'7,.~.i ----- __ __"'- --

-'---"---""'-- --- _._--_
. . . .. -.
. . ... .... ... ....... ......--- -
.. _---_ .. --"
-~
_ neodleo to make room lor the "... briaht

~ 0DeI that weft 1000 to deck out the tree. ia ----_._._- . . ...-,_ ..__..._---
_ _ 1 _ _ _ _ _ _ _ _ _ 1 _ _ _ _ _ _ -*
an tM ~
---. . _. ...___
_. ._--._ . . _1 _.._. ____
__ . ...,.,--
-- .-~----.....
----
'I'M ..... mmbirw lit. lip tM hu.t. aDd
.... ,... -... -_ . .. _--...

-:.
~~~ ~.~~=:.tt:! :::~b!n~~
bod)' ... bit ....tbd t.bn.a tbt Yillaae 00 biI way.
- -.----- 7
~
Ull
4__ .11_ . ._"

AllJlWl:ldWtM~lI"w:ItkIpIt . . . tridlt . t:bt~ 8
lQIIII. udlOClllltM'""'~~Wdvp~~.
:=~~::".:'r:!!:.:.dt!:..~
,.,....1_. _ _ _ _
" " " " _ . . _ _ I . . . ... _ . . ." """"'l _ _
-._ ~ ~ ........... ...
• __ ", ~ - _-__ IoL ................ _ ~ ...... . _ _
tl
II JI
==-"'a.M~~==-:-,=::::-~:: -1I1 _ _ _ _ "' __ I .... ..
=.;."!:=-..:r~~"t::1.-.::.""""'''' NdIIo
II_~ . lI<t""" - Ho: ""_M_ I _
......
. . Ia W.ItfI!ttott.'""'"II>M
-..~ _ _ .. _ _ ID ........
....."W1IiIIII
- '_ _ _ ' l " . . . il_
.. ~""oJJI'_
_ ru ...
II~I w.'
n
~~€~~.a..~::= 10
T-..dlrr d ioMo:#. 1 "',. . . . . wI.,a-t ........ o.&It; .n4 ~ 1111 .., lit
Iifu' ri1 ,..
.
.111
....,.nI I
! 1IIt t~ Ow dlJl'_11llNl1l1AM _ wIlli, ...... ftlU .. h.d thl'
vlo.IIPwi" . ... .1nlI:1I1 .bftu ..
~'!;~:;'-:""~~=·"'.:I$'1':.~~,,,*,,~--= III~ rt:lNlhtklllllMl ~ "",,"
11
In _ , ~.I ~tllc wllon:ltt:llt.c&IO)'tarrM&npkl-.llhemInIJlo
• quqmi .....tId jllo cit')' ....c.,1'op t"'<I:"cI"", _u oIkmi ..... I laW .... 1111
JlAln t~ inb.bi" l'Iu""'liil'li I" ... ud ........lcpun::tt. ttin. tlwuPll"'iMlll'l ..
Figure 7. Reading chart designed for routine ophthalm-

ic examinations. (a) American Optical Chart
11970; (b) Bausch and Lomb Chart 713568
Control and consistency will be optimized if constant size

progression ratios are used. Clinicians tend to be relatively
conservative and resistant to changes in their routine but log-
ic dictates that, in time, near vision prescribing will routinely
be based on charts that use constant ratio size progressions.
The clinician dealing with magnifying devices should have

a good understanding of how to quantify and control the mag-
nification effects. For distance vision, consideration of mag-
nification effects involves telescopes and the process of pre-
scribing magnification is quite straight forward. However,
for near vision, there are alternative ways of quantifying
magnification effects and furthermore, different kinds of
optical considerations must be made when considering the
magnification effects provided by the different kinds of mag-
nifying devices. The clinician should have facility in deter-
mining the equivalent viewing distance (or alternatively the
equivalent viewing power or the magnification rating) of each
viewing system presented to the patient.
204
1.1 High Addition Reading Glasses
High addition reading glasses achieve an enlarged retinal

image by effectively forcing the patient to adopt a closer
viewing distance. For presbyopic patients the Equivalent
Viewing Distance is simply the viewing distance measured
from the spectacle plane. The Equivalent Viewing Power is
essentially the power of the addition. The magnification rat-
ing is the power of the addition divided by 4. For pre-
presbyopic patients, accommodation may be exerted and the
EVD is simply the viewing distance. The EVP is the sum of
the addition and the accommodation and this is the reciprocal
of the viewing distance. The magnification rating (by the
most commonly used method) is equal to the EVP divided by
4. In prescribing high addition reading glasses with lens pow-
ers from 10 to 18 diopters, cataract aspheric lenses are advi-
sable; for additions over 2.0 D special series lenses are indi-
cated. Binocular performance may often be achieved for
reading additions of 12. diopters or less and in such cases spe-
cial consideration must be given to lens decentration to avoid
difficulties with convergence [9].
1.2 Hand-Held Magnifying Glasses
Held-held magnifying glasses are very versatile. If used so

that there is a long eye-to-lens distance, presbyopic patients
will be required to use their distance correction while young
observers should ideally relax accommodation. For most pur-
poses, the EVD is about equal to the focal length of the mag-
nifying lens, the EVP is equal to the power of the lens, and
the magnification is equal to the power of the lens divided by
4. For very short eye-to-Iens distances the presbyope will
achieve more magnification if the reading addition is used,
and young people will receive more benefit if they accommo-
date. If the magnifier is held close to the spectacle plane,
the EVD will be approximately equal to the reciprocal of the
sum of the reading addition (or the accommodation) and the
power of the magnifying lens. The EVP will be approximately
equal to the sum of the addition and the lens power. Again
the magnification will be equal to the EVP divided by 4 [10].
205
1.3 Stand Magnifiers
For virtually all stand magnifiers, the image is located quite

close to the lens. The clinician should know the distance sep-
arating the image from the lens. Also the equivalent power
of the lens system should be known. The image formed by the
magnifier will be larger than the original object, and the
enlargement ratio can be calculated by taking the sum of the
lens power and the emerging vergence and dividing this by the
emerging vergence. For this calculation, the positive or neg-
ative signs should be ignored. The clinician should take spe-
cial note of the eye-to-Iens distance. The accommodation
demand is determined by the sum of the eye-to-Iens distance
and the lens-to-image distance. In predicting resolution per-
formance with stand magnifiers, the clinician should consider
the accommodation demand and the enlargement ratio. The
EVD is calculated by dividing the eye-to-image distance by
the enlargement ratio. The EVP is obtained by multiplying
the accommodation demand (in diopters) by the enlargement
ratio. The magnification rating can be obtained by dividing
the EVP by 4 [11].
1.4 Near Vision Telescopes
The clinician prescribing near V1S1on telescopes should be

acquainted with the front vertex focal length of each tele-
scope because this sets the viewing distance. It is also useful
to know the magnification of the telescope component of the
system. If the near vision telescope is created by adding a
lens cap to an afocal telescope, then the power of the lens
cap will determine the front vertex focal power of the sys-
tem. All near vision telescopes can be considered to consist
of an afocal telescope combined with additional lens of posi-
tive power on the objective to achieve a near focus. The EVD
is obtained by taking the anterior vertex focal length and
dividing by magnification of the telescope component. The
EVP is determined by multiplying the front vertex focal pow-
er by the power of the telescope. The magnification again
can be deduced by dividing the EVP by 4 [12].
206
1.5 Video Magnifiers
In order to predict resolution achievable with video magnifi-

ers the clinician should know the transverse magnification
and the viewing distance. The viewing distance indicates the
accommodation demand. The EVD will be equal to the view-
ing distance divided by enlargement ratio (or the transverse
magnification). The EVP will be equal to the product of the
accommodation demand and the enlargement ratio. The mag-
nification rating will be given by the EVP divided by 4.
In the prescribing of near vision optical magnifiers, the

clinician determines the Equivalent Viewing Distance, Equiv-
alent Viewing Power, or magnification rating that will
achieve the patient's resolution goal. Typically, the clinician
will then present a series of near vision magnifiers that
should achieve the resolution goal as all should provide about
the same EVD, EVP or magnification rating. This will allow
the patient to compare the different kinds of aids and to
evaluate and appreciate the restrictions that each may place
on field of view, depth of field, working distance, working
space, lighting needs, image quality, postural flexibility, man-
ual dexterity, endurance, portability, convenience, cost,
appearance and weight.
2. Summary
For most accurate and most consistent results in prescribing

magnifiers the clinician should have good control of the
measurements of resolution performance and a good under-
standing of the optical systems being employed. To optimize
prescribing consistency the following conditions should be
met.
1. The charts used for the assessment of resolution and

efficiency should have a task that is standardized at
all size levels, and the size progression should be geo-
metric following a consistent ratio.
2. Changes in magnification should be made in steps

whose sizes follow the same number sequence as the
letter sizes labeled on the chart.
207
3. Clinicians should understand the optical principles of

the systems that they use. The clinician should have
facility in performing the calculations that allow the
determination of EVD, EVP or magnification ratings
of various optical systems after giving appropriate
consideration to the conditions of use.
4. Clinicians should know their own inventory of low

vision devices. They should be familiar with the key
parameters of all the optical systems that they are
likely to present to their patients.
3. References
1. I.L. Bailey, J.E. Lovie: New design principles for visu-

al acuity letter charts. Am. J. Optom. Physiol. Opt.
53, 740 (1976)
2. F.L. Ferris, A. Kassoff, G.H. Bresnick, I.L. Bailey:

New visual acuity charts for clinical research. Am. J.
Ophthalmol. 94,91 (1982)
3. G. Strong, G.C. Woo: Distance visual acuity chart

incorporating some new design features. Arch.
Ophthalmol. 103,44 (1985)
4. L.L. Sloan: New test charts for the measurement of

visual acuity at far and near distances. Am. J.
Ophthalmol. 48, 807 (1959)
5. I.L. Bailey, J.E. Lovie: The design and use of a new

near vision chart. Am. J. Optom. Physiol. Opt. 57, 378
(1980)
6. C.H. Keeler: Visual aids for the pathological eye

(excluding contact lenses). Trans. Ophthalmol. Soc.
U.K. 76, 605 (1956)
7. H.T. Lewis: UNC near vision chart. J. Vision Rehab.

3(1), 8 (1985)
8. L.L. Sloan, D. Brown: Reading cards for selection of

optical aids for the partially sighted. Am. J. Ophthal-
mol. 55, 1187 (1963)
208
9. I.L. Bailey: Centering high addition spectacle lenses.

Optom. Mthly. 70, 523 (1979)
10. I.L. Bailey: Combining hand magnifiers with spectacle

additions. Optom. Mthly. 71, 458 (1980)
11. I.L. Bailey: The use of fixed-focus stand magnifiers.

Optom. Mthly. 72(8),37 (1981)
12. I.L. Bailey: Principles of near vision telescopes.

Optom. Mthly. 72(9), 32 (1981)
The Amorphic Fresnel Prism Trioptical System
Richard L. Brilliant, Sarah D. Appel, Robert J. Ruggiero
1.
Severe visual field constriction, characteristic of conditions

such as retinitis pigmentosa and end stage glaucoma creates
significant limitations on an individual's mobility and activi-
ties of daily living. In such cases, reverse telescopic systems
have been utilized in an attempt to expand the visual field
[1,2]. These image minification systems focus a larger field
area on the remaining viable retinal tissue. The practicality
of such systems is limited, however by the significant reduc-
tion in visual acuity resulting from the 360 degree minifica-
tion. A -2.0X magnification system, for example, providing
2X expansion in field will reduce visual acuity by a factor of
O.SX. Also problematic are perceptual difficulties arising
from vertical and horizontal displacement of the visual envi-
ronment. Depth perception, so critical during mobility tasks,
is greatly impaired by minification of the vertical meridian.
Problems in correctly judging the position of stairs, curbs and
irregularities on the road surface render the system impracti-
cal for continuous viewing while in motion.
In an attempt to minimize visual acuity degradation as

well as vertical displacement of the visual environment, a
meridional minification system, the amorphic lens was
designed by Dr. William Feinbloom [3]. By properly spacing
axis 90 degrees positive and negative cylindrical lenses, mini-
fication is limited only to the horizontal meridian. This
results in the creation of a retinal image unchanged in the
vertical meridian and compressed in the horizontal meridian
by a factor of the minifying power of the lens. Full field
amorphic lens systems were made available in powers of
-1.2X, -1.4X, -1.6X, -1.8X and -Z.OX.
210
2. Previous Study
Initial evaluations of the full field amorphic lens systems

were conducted at the William Feinbloom Vision Rehabilita-
tion Center as well as in a number of low vision facilities
throughout the country (1984). The criteria for patient selec-
tion included the following:
1. Visual field constriction to 20 degrees or less with no

peripheral islands of vision remaining.
2. Visual acuity better than 6/60 (20/200).
3. History of significant mobility problems.
Thirty two patients were evaluated at the William Fein-

bloom Vision Rehabilitation Center. While 28 patients (88%)
appreciated an expansion of their visual field area, 7 patients
(22%) reported immediate dizziness and disorientation and
were withdrawn from further evaluation. Throughout testing
there was a preference (84%) for the -1.6X and -l.8X
amorphic lens systems. The remaining 25 patients were given
instruction by orientation and mobility specialists in the cor-
rect utilization of the full field amorphic system during b9th
indoor and outdoor mobility tasks. At the completion of the
instructional sequence, the full field amorphic system was
loaned for evaluation at home. After a two week trial period,
4 patients (12%) reported improved mobility skills and elected
to continue utilizing the full field amorphic lens system for
mobility purposes. The remaining 21 patients (66%) preferred
head and eye scanning movements to utilization of the
amorphic lens system during activities involving mobility.
Follow up interviews conducted one year later indicated that
only 2 individuals were utilizing the full field amorphic lens
system on a regular basis.
Sixty nine patients were evaluated in the national study

[3]. Thirty patients (44%) immediately felt dizzy and diso-
riented during the initial mobility evaluation and were there-
fore withdrawn from the study. Of the remaining 39 patients
who continued the study and utilized the lenses for two
weeks, 14 reported improved mobility as they continued to
use the lens system. Complaints voiced throughout both stud-
ies included depth of field confusion, distortion and the need
for decreased head movements during mobility tasks. In
reviewing these results it became apparent that adaptation to
211
a full field amorphic design was difficult due to the necessity

of continuous viewing through the amorphic lens. The na tion-
al study found that only 14 of the 69 patients evaluated
reported the full field amorphic lens system to have a benefi-
cial effect on independent mobility.
3. Present Investigation
In order to improve patient acceptance of and adaptation to

the cylindrical minification system, Dr. Richard Brilliant and
Designs for Vision Inc. designed a bioptic amorphic lens sys-
tem. The amorphic lens was redesigned into a rectangular
shape and mounted onto the superior half of a carrier lens.
Such a design enabled the individual with constricted fields to
alternate between the cylindrical minifier and the carrier
lens while in motion. One can thus gradually increase the
amount of time spent viewing through the amorphic lens until
a comfortable limit is achieved. Adaptation is thereby facili-
tated.
This design also allows for the introduction of Fresnel

prisms onto the temporal aspect of the carrier lens in order
to enhance an individual's peripheral field awareness [4-6].
Fresnel prisms shift the apparent position of visual targets
occurring in the area of field loss onto the primary visual
axis. An individual utilizing this prism system can thus sub-
stitute scanning eye movements for gross head movements in
order to protect peripheral objects in his environment.
Use of the amorphic Fresnel trioptical system during

travel involves a continuous sampling of the visual environ-
ment through each optical lens. Safe and purposeful mobility
requires an awareness of one's immediate environment as well
as an orientation to the general surroundings. For example,
while walking in a crowd, scanning into the field of the Fres-
nel prism allows the traveler to detect individuals or obsta-
cles positioned at large angles of eccentricity. This includes
individuals walking alongside or slightly in front of the low
vision traveler. In this manner immediate mobility hazards
can be avoided. The amorphic lens is utilized when an
expanded picture of the visual environment is necessary such
as viewing an intersection, finding an individual in a crowd, or
negotiating a narrow aisle in a supermarket. When the triop-
tical system is used correctly, gross head movements are
212
Figure 1. Front view of the amorphic Fresnel trioptic

system.
Figure 2. Side view of the amorphic Fresnel trioptic sys-

tem.
minimized thereby relieving the dizziness and disorientation

213
that was reported with the full field design.
A group case study evaluating this trioptical system was

conducted at the William Feinbloom Vision Rehabilitation
Center in 1985. Patient selection criteria remained the same
as in the 1984 study. Six patients were evaluated with the
amorphic Fresnel trioptical system. All 6 had been previously
diagnosed as having typical retinitis pigmentosa, confirmed
by ERG testing. Five patients also had histories of Ushers
Syndrome with accompanying bilateral neurosensory hearing
loss.
As in the 1984 study there was a marked preference for

the intermediate range of minification power. Four preferred
-1.6X and 2 preferred -1.8X. All six reported an expansion of
their visual field area while viewing through the amorphic
lens. Instruction in the use of the amorphic Fresnel trioptical
system was provided by orientation and mobility specialists in
both indoor and outdoor environments. The system was then
loaned for a two week trial period. Five patients reported
increased confidence while traveling after this time. There
were no complaints of dizziness or disorientation. All five
were dispensed the amorphic Fresnel trioptical system. Dur-
ing a one year follow up interview it was found that four of
the five individuals continued to find the system helpful and
used it on a regular basis.
All five individuals who received the amorphic Fresnel

trioptical system had expressed concerns in addition to mobil-
ity for which the cylindrical minifier was helpful. These
included a mother of two children who needed to supervise
their activities in the playground, an accountant who needed
to scan his account sheets, a businessman who wanted to see
people at meetings as well as locate objects on his desk, a
mechanic who needed to see a larger area while working on
an engine, and a minister who wanted to view his hearing
impaired congregants signing during conversations.
214
4. Conclusion
Modification of the full field amorphic lens system into a

superior mount bioptic design allows for the placement of
Fresnel prisms onto the carrier lenses. This amorphic Fresnel
trioptical system enhanced the mobility performance of 5 out
of 6 individuals with field constriction secondary to retinitis
pigmentosa. Four of the 6 reported overall satisfaction with
the system after one year's use. It is acknowledged that due
to the small sample size, a more extensive study would be
conducted to verify our findings. Further studies should also
include investigation of the system's efficacy for other field
constricting pathologies such as end stage glaucoma.
It should be emphasized that proper instruction in the use

of the amorphic Fresnel trioptical system is essential to
ensure successful adaptation. Orientation and mobility spe-
cialists should be involved whenever possible in order to
include general orientation and mobility instruction that is so
important to individuals with severe field constriction.
5. References
1. N.J. Weiss: Treatment of the retinitis pigmentosa

patient: Initial report. Optom. Wkly. 68, 1067 (1977)
2. N. Drasdo: Visual field expanders. Am. J. Optom.

Physiol. Opt. 53, 464 (1976)
3. W.W. Hoeft, W. Feinbloom, R. Brilliant, R. Gordon, C.

Hollander, J. Newman, E. Novak, B. Rosenthal, E.
Voss: Amorphic lenses: A mobility aid for patients
with retinitis pigmentosa. Am. J. Optom. Physiol.
Opt. 62, 142 (1985)
4. R.T. Jose, A.J. Smith: Increasing peripheral field

awareness with Fresnel prisms. Opt. J. Rev. Optom.
113 (12), 33 (976)
5. P.D. Gadbaw, W.A. Finn, M.T. Dolan, W.R. De l'Aune:

Parameters of success in the use of Fresnel prisms.
Opt. J. Rev. Optom. 113 (12),41 (1976)
215
6. N.J. Weiss: An application of cemented prisms with

severe field loss. Am. J. Optom. Arch. Am. Acad.
Optom. 49, 261 (1972)
Sensorimotor Adaptation to Telescopic Spectacles
Joseph L. Demer, Jefim Goldberg, Franklin I. Porter, Herman

A. Jenkins
1. Introduction
The National Society to Prevent Blindness estimates that 1.4

million persons in the U.S. are severely visually impaired
although more than 90% of those with irreversible impair-
ment have a degree of residual vision [1]. A great majority
could theoretically benefit from spectacle magnification aids
[2]. However, many low vision patients cannot either tolerate
spectacle magnifiers or use them effectively. This may be
because useful vision requires not only a sufficiently large
image on the retina, but also stability of the image. Inade-
quate image stability is likely a reason for the failure of
many patients to benefit from spectacle magnifiers.
Images on the retina must be kept stable despite ubiqui-

tous head movements resulting from tremor, heartbeat, pos-
tural instabilities, and ambulation. Such head movements
may seriously impair vision, because retinal image motion of
more than a few degrees per second degrades acuity [3]. Ret-
inal image stability is achieved by the vestibulo-ocular reflex
(VOR) in conjunction with visual tracking mechanisms.
Although normal VOR operation is often taken for granted,
its loss produces such a catastrophic effect that even reading
in bed is impossible without steadying the head against a sta-
ble object [4].
The VOR uses head velocity information sensed by the

inner ear to reflexively move the eyes to reduce retinal
image slip. VOR "gain" is a quantitative measure of the
effectiveness of the reflex and can be defined as eye velocity
divided by head velocity. When supplemented by visual track-
ing mechanisms, gain normally equals 1.0, indicating that
compensatory eye velocity is equal and opposite to the per-
turbing head velocity. The method of calculating VOR gain is
shown in Fig. 1, where the amplitude of the slow phase sinu-
soidal envelope ER is divided by the amplitude of the head
velocity sinusoid Hp to find gain. In alert subjects in total
217
darkness, VOR gain is about 0.7 - 0.95. The increase in VOR

gain produced by vision is the result of visual-vestibular
interaction (VVI). The combined compensatory response of
the vestibular and visual systems in light is called the visual-
vestibulo-ocular reflex (WOR). WOR gain is defined to be
eye velocity divided by head velocity, as measured in the
light.
E0
10j
30
E0
6' ,, , , , ,, 'I'd' ,, , , " '20

Time(sec)
Figure 1. A record of the vestibulo-ocular reflex (VOR).

Lower tracing represents head velocity, H.
Upper tracing E shows horizontal eye position,
consisting of compensatory slow tracking phases
and resetting quick phases. Rightward eye
movements are shown as positive. Eye velocity
E is obtained by differentiation of eye position;
slow phases have a sinusoidal envelope while
quick phases are truncated due to their high
velocities.
When spectacles are worn, the WOR gain required for

eliminating retinal image slip is not 1.0; it is equal to the
218
magnification factor of the spectacles [6]. Most of the bur-

den for this adjustment to spectacles falls to a long-term
adaptive mechanism that effectively recalibrates VOR gain
because visually mediated tracking is too slow to compensate
for rapid head movements [7,8]. This adaptive process is a
form of motor learning called VOR gain plasticity. Adaptive
changes in VOR gain have been shown to result from the mag-
nifying effects of ordinary refractive spectacle lenses [6].
Animal experiments have shown that VOR gain, as measured
in darkness, undergoes gradual plastic adaptation to ultimate-
ly compensate for 2X telescopic spectacle magnification
[9,10]. Plastic VOR gain reduction has been demonstrated in
several normal humans [11-14]. Substantial adaptive increases
in VOR gain have been demonstrated in a limited number of
normal humans after several days of wearing 2X telescopic
spectacles [15,16].
The importance of VOR and WOR adaptation to tele-

scopic spectacles is underscored by the observation that reti-
nal image motion of only a few degrees/second significantly
degrades visual acuity [3] and motion of 15 to 25 degrees/
second reduces acuity almost five-fold [17]. If not appropri-
ately compensated by eye movements, the retinal image slip
produced by an ordinary 90 degree head turn would exceed
100 degrees/second and reduce visual acuity of a normally
sighted person to less than 6/30 (20/100). Such would be the
case with unmagnified vision if WOR gain were zero, or with
2X telescopic spectacles if WOR gain were 1.0. Passive
head movement has been demonstrated to reduce the visual
acuity of normal subjects wearing 2.2X telescopic spectacles
for the first time although the deficit usually resolves after a
period of adaptation [18]. Further, when VOR and WOR
adaptation to telescopic spectacles are incomplete, subjects
uniformly report headache, nausea, oscillopsia, and blurred
vision during head movements [15-18].
Low vision patients may be expected to have particular

difficulty achieving adaptation to telescopic spectacle magni-
fication. Many low vision patients are elderly, and at least
one study has demonstrated a decline in VOR gain with age
[19], although results on this question are mixed [14]. Most
low vision patients are fitted with 4 to 6X telescopic devices,
producing an immediate requirement for a VOR gain increase
called by COLLEWIJN et aZ. "highly, even extremely,
demanding •.• "[7]. Further, reading is a major task for which
low vision patients employ spectacle magnifiers. Reading
219
requires the material to be held close to the head. Because

the head and eyes do not have the same axis of rotation,
VVOR gain must be higher at near in order to be compensato-
ry for head movement. Normal persons under conditions of
unmagnified vision achieve a gain of over 1.5 for a 10 cm tar-
get distance [20]. This means that a low vision patient wear-
ing an 8X spectacle magnifier might require a VVOR gain of
12 to read clearly! Such extreme adaptation may be impossi-
ble, or at least may require considerable time and effort to
acquire. Even VVOR gain adaptation to 2X or 4X spectacles
may be impossible for certain neurologically impaired
patients since lesions of parieto-occipital cortex [21], the
cerebellar flocculus [22], and inferior olivary nucleus [23]
abolish VOR gain plasticity. VOR gain plasticity has been
noted to vary from person to person in a normal population
[18].
In order to evaluate VOR and VVOR adaptation to tele-

scopic spectacle low vision aids, a series of tests has been
devised in our laboratory that employ a relatively brief period
of exposure to magnified vision using telescopic ::Ipectacles.
Testing of short-term adaptation has been chosen for feasibil-
ity as a clinical test which, with appropriate clinical correla-
tions, may prove prognostically useful. Results of our prelim-
inary experience are reported here.
Fifteen normally sighted subjects, ranging in age from 21 to

56 years, and six low vision subjects, ranging in age from 10
to 59 years, gave written informed consent and participated
in this study. (Preliminary data on seven of these subjects has
been previously published [18].) Most normal subjects had
best corrected visual acuity of 6/6 (20/20) or better in both
eyes, but one subject had unilateral visual loss from amblyo-
pia and another from macular degeneration. The low vision
subjects had best corrected visual acuities ranging from 6/21
(20/70) to 6/180 (20/600) in the better eye, and all but one
were monocular. Two low vision subjects had visual loss from
toxoplasmosis, two from retinopathy of prematurity, one
from diabetic vitreo-retinopathy, and one from optic atrophy
secondary to hydrocephalus.
220
Eye movement data was obtained using DC-coupled

electro-oculography (EOG) during rotation about a vertical
axis. Rotation followed a sinusoidal or a sum of five sinu-
soids. The eye position signal was digitally sampled at 200 Hz
and eye velocities were determined by digital differentiation
and smoothing. Slow phase eye velocity was obtained by
excluding saccades and quick phases using interactive digital
filtering. Best fit sinusoidal curves were automatically gen-
erated for each eye velocity cycle and frequency. Gain was
taken to be the ratio of the amplitude of slow phase eye
velocity to the amplitude of head velocity and is reported as
the mean and standard deviation over three to ten cycles of
sinusoidal rotation. When a head velocity stimulus consisting
of the sum of sinusoids was employed, the amplitude of each
component frequency of response was taken from the best fit
sinusoid.
During eye movement recording, subjects were seated in

a rotatory chair with their heads strapped to a headrest. A
period of several minutes adaptation to ambient illumination
was allowed for stabilization of the corneo-retinal potential.
Calibration of the EOG signal was performed with subjects
wearing only their corrective spectacles for distance and was
achieved by asking subjects to fixate lighted targets at 15
degrees to the left and right of a center target. The calibra-
tion sequence was repeated until reproducible EOG potential
changes were measured; VOR gains were computed on the
basis of the means of one or more calibration trials immedi-
ately before and after each 1-3 minute period of data collec-
tion. This technique has been shown to result in highly repro-
ducible measurements of VOR and VVOR gain [18]. During
VOR and VVOR gain measurements, subjects performed men-
tal exercises aloud and were instructed to look straight
ahead. These activities controlled mental set which can
influence measured VOR gain [24]. Subjects were surrounded
by a vertically striped curtain located approximately 70 cm
from the eyes.
Dynamic visual acuity (DVA) is defined to be visual acui-

ty achieved using a spectacle magnification device during
head movement. DVA was measured using a standard Snellen
letter chart at a distance of 4 m during sinusoidal head and
body rotation, usually at 1.0 Hz, amplitude 30 deg/sec. DVA
was compared with static acuity, similarly measured but with
the head motionless. Acuities were recorded binocularly.
221
For each subject, VOR gain was initially measured in

darkness at an amplitude of 60 deg/sec. VVOR gain was
measured in light either without or with telescopic specta-
cles. The velocity amplitude for head rotations in light was
30 or 60 deg/sec. Telescopic spectacles were Galilean bino-
cular telescopes having a visual field confined to the central
13 degrees either nominally 2.2X fixed focus or 2.0X variable
focus.
Following initial measurements, subjects wore telescopic

spectacles for a 15 minute period while undergoing sinusoidal
rotation at 0.1 to 0.4 Hz, amplitude 20 or 30 deg/sec. A tele-
vision display 4 m away served as a visual target during the
adaptation period. At the end of the adaptation period,
measurements were again made of VOR gain, VVOR gain with
magnified vision, and, in some cases, DVA.
3. Results
3.1 Normally Sighted Subjects
Initial VOR gain was recorded in darkness at a frequency of

head rotation of 0.1 Hz, amplitude 60 deg/sec. As seen in
Fig. 2, VOR gain for the normal subjects was 0.74 +/- 0.10
(mean +/- standard deviation, SD, n=14). With unmagnified
vision, this VVOR gain was 1.08 +/- 0.07 (mean +/- SD, n=14).
When subjects wore 2X or 2.2X telescopic spectacles for the
first time, mean VVOR gain increased to 1.50 +/- 0.39 (mean
+/- SD, n=lS) although a few subjects exhibited erratic per-
formance and experienced VVOR gain decreases. Some sub-
jects noted oscillopsia, palmar diaphoresis, and a vague vis-
ceral discomfort during head motion when telescopic
spectacles were initially worn. Records of VVOR eye veloci-
ty with and without telescopic spectacles are seen in Fig. 3.
Dynamic visual acuity (DV A) was studied in seven nor-

mally sighted subjects. DVA was initially measured in each
test session by asking subjects to read a Snellen acuity chart
while wearing 2.2X telescopic spectacles. Acuity was first
measured with the subject's head stationary, and again with
the head and body undergoing sinusoidal rotation in the rotat-
ing chair. Most measurements of DVA were made at 1.0 Hz
with an amplitude of head rotation of 30 deg/sec. Static vis-
222
2.0
..
1.8
1.6
~ 1.4
1.S0±O.39 _ .-
.
~
c;j
!C 1.2
z
:c -!.:-
I, .. 1.01±O.07
~ 1.0
C
i= .:
! 0.8
0.6
..··
~O.74±O.10
0.4
0.2 '----'-_ _ _....l...-_ _- - - ' ' - - - _

DARK LIGHT LIGHT
UNMAGNIFIED MAGNIFIED
Figure 2. VOR gain in darkness and VVOR gains with nor-

mal and 2X-2.2X magnified vision in 15 normal
subjects. Measured at head rotation frequency
of 0.1 Hz; amplitude 60 deg/sec in darkness, 30
deg/sec in light. Values are means +/- standard
deviation.
ual acuities varied from subject to subject. DVA was one to

three lines Snellen poorer than static acuity in four subjects;
in the remainder, the smallest line on the chart was read
under both static and dynamic conditions.
After initial gain and DVA measurements, subjects wore

telescopic spectacles for a IS-minute period of adaptation to
continuous sinusoidal rotation in the light while viewing a dis-
tant video screen. Rotation was at frequencies between 0.1
and 0.4 Hz, amplitude 20-30 deg/sec.
After 15 minutes of adaptation to telescopic spectacles,

measurements of gain and DVA were repeated. By this time
subjects reported little or no oscillopsia. Mean VOR gain for
the group increased from 0.74 +/- 0.10 (mean +/- SD, n=14)
223
1. ••
TIME (SECONDS)
Figure 3. VVOR eye velocity recorded for typical normal

subject with unmagnified (A) and 2X magnified
(B) vision. Rotation was at 0.1 Hz, 60 deg/sec.
Gain increased from 1.03 with unmagnified
vision in A to 1.78 with magnified vision in B.
Dots represent digitally sampled eye velocities;
solid curves represent best fit sinusoids. Right-
ward eye velocities are shown as positive.
initially to 0.83 +/- 0.12 (p<0.05) after adaptation. Nine of

the 14 subjects showed statistically significant VOR gain
increases of 7-31% (Fig. 4). Figure 5 shows VOR eye veloci-
ties before and after telescope adaptation; an obvious
increase in the amplitude of the eye velocity sinusoid occurs
after adaptation.
The effect of adaptation on visual-vestibular interaction

is seen in Fig. 6. VVOR gain with magnified vision was
unchanged for the group, being 1.50 +/- 0.39 (mean +/- SD,
n=15) before adaptation and 1.49 +/- 0.27 (n=13) after adapta-
tion. Two subjects were not studied after adaptation for
technical reasons. Inter-individual variability was somewhat
reduced after adaptation since the population standard devia-
tion decreased from 0.39 to 0.27. Some subjects with initially
high VVOR gains experienced small decreases after adapta-
tion (Fig. 6).
DVA was again measured after adaptation to telescopic

spectacles. In all four subjects in whom DVA had been worse
than static visual acuity initially, there was an improvement
in DVA after adaptation. This represented an improvement
of 30-100% in visual angle resolvable.
224
1.00
0.90
z
~a: 0.80
o
>
0.70
0.80
INITIAL ADAPTED
Figure 4. Plastic VOR gain increases in normal subjects

after 15 minutes adaptation to 2X or 2.2X tele-
scopic spectacles. Measured for sinusoidal head
rotation at 0.1 Hz, amplitude 60 deg/sec.
3.2 Low Vision Subjects
The clinical characteristics of the low VISIon subjects are

summarized in Table 1. Most of the low vision subjects were
functionally monocular, so monocular electrode placement
was employed around the seeing eye. EOG voltage reductions
resulted from partial retinal detachments or scars. Reduced
EOG voltage was compensated by increasing preamplifier
gain.
All low vision subjects were able to fixate the calibration

lights against a dark background. One subject with 6/180
(20/600) vision (L.N. in Table 1) could not locate the targets
under normal room illumination, precluding VVOR measure-
ments in this individual. The accuracy of calibration sac-
cades was generally poorer in low vision than in normal sub-
jects. In addition, low vision subjects employing eccentric
fixation occasionally vary their retinal fixation point. Eccen-
225
Figure 5. VOR eye velocity during sinusoidal head rota-

tion at 0.1 Hz, 60 deg/sec before (A) and after
(B) adaptation to 2.2X telescopic spectacles.
VOR gain was 0.71 initially and 0.93 after adap-
tation. Note increase in amplitude of the eye
velocity response, shown as solid sinusoidal
curves. Dots represent digitally sampled eye
velocities.
trically fixating subjects must be instructed to look at the

target in exactly the same way every time.
One low vision subject (D.G., age 33, Table 1) was found
on eye movement testing to have an intense spontaneous hori-
zontal nystagmus, precluding gain measurements. Several
other subjects, including F.M. (Table I) had low intensity
spontaneous nystagmus in darkness that was suppressed by
vision and, therefore, was clinically not apparent.
Subjects F .M. and B.S. exhibited asymmetric VOR gains,

higher for slow phases to the right than to the left. This
effect, seen in Fig. 7, can readily be quantified by computer
methods used in this study.
VOR gain plasticity was evaluated in five low vision sub-

jects (excluding D.G., age 33, Table I). In three of five sub-
jects, 15 minutes adaptation to 2.2X telescopic spectacles
produced VOR gain increases of 24-43%. In two subjects the
VOR gain increases were significant at the 0.005 level. It
may be seen from Table 1 that VOR gain increases were
observed in three subjects who regularly made extensive use
of telescopic spectacles in employment, walking, driving and
226
2.0
1.8
...
..
1.8
1.50±O.3I-·- --i- 1.41±O.27
1.4
z
:c
..
1.2
c:I
-:-:. 1.08±O.07
0
1.0 ·1·
~
0.8
0.8
0.4
0.2 L-_-'-_ _ _....J.._ _ _ _L - _

INITIAL INITIAL FINAL
UNMAGNIFIED MAGNIFIED MAGNIFIED
Figure 6. VVOR gains for 15 normal subjects with

unmagnified vision, and with telescopically
magnified VlSIon initially and again after 15
minutes adaptation to telescopic spectacles.
Values are means +/- standard deviations.
Table 1
Clinical data of low vision subjects.

Best COHee ted Telescope Plasticity
Ocular Vision Type VOR Gain I Postadoptotion
Potient Age OcclIpation Condition (Without Tele9:opes) And Uscge Preadaptotion (Percent Change)
D.G. 45 Secretory Toxoplasmosis Right NLP2 3.5X 0.46.!.0.13 0.66.!. 0.06

Left 6/90 6-8 hours/day (43%)
reading, typing
F.M. 32 Student Toxoplasmosis Right 6/36 4X 0.69.:. 0.03 0.92,.!. 0,08

Left 6/120 TV Viewing (33%)-
Walking, Dl-iving
L.N. 4 10 Student Optic Atrophy Right 6/180 2X 0.61.!. 0.03 0.76.!. 0.07
Hydrocepha Ius Left Lp3 Schoolwork, (24%)-
TV Viewing
5.J. 59 Retired Diabetic Right 6/120 3X 0.84.±. 0.08 0.81.!. 0.04

Vitreo- Left NLP2 TV Viewing
Retinopathy
B.S. 32 Librarian Retinopathy of Right 6/21 8X 0.54.!.0.15 0.48.±. 0.08

Prermturity Left Lp3 Occasiono [ Use
D.G. 33 Clerical Retinopathy of Right 6/24 6X (Nystogm.Js) (Nystagmus)

Prematurity Left Lp3 Reading Only
I. Coin reported as mean + standord deviation.
2. NLP--no light perception.
3. LP--light perception only.
4. Non-English speaker.
p < 0.005
227
~ A B
~
f\iJvvvt ·
120 .-------;-------,,-----,
UJU ••
>UJ
UJ II) •
~ia ...
,: ~ . - '.
~ Q -. .
~ .12. t----.-~~~~ _ _~ L_ ~~~.----<

~ J. •• 30 60
TIME (SECONDS)
Figure 7. VOR eye velocity in low VISIon subject F .M.

before (A) and after (B) adaptation to telescopic
spectacles. Note that slow phase eye velocities
to the right (positive) are greater than to the
left «negative) reflecting asymmetric VOR
gain. VOR gain averaged over left-hand and
right-hand half cycles increased from 0.69
before adaptation to 0.92 after adaptation. Gain
increased both to the right and left.
schoolwork. The two subjects who did not exhibit VOR gain
increases used their telescopic spectacles for limited purpos-
es such as television viewing.
Visual-vestibular interaction (VVI) was evaluated in three

low vision subjects (excluding D.G., age 33, and L.N.). In
three subjects, VVOR gain measured while wearing 2.2X tele-
scopes was low and performance was erratic. Only one sub-
ject, S.J. (Table 1), was able to use magnified vision to
achieve a VVOR gain exceeding 1.0. SUbject S.J. had a gain
of about 1.40 under these conditions, not significantly
increased after adaptation to telescopic spectacles.
Dynamic visual acuity was tested with 2.2X telescopic

spectacles in all six low vision subjects. Unlike normal sub-
jects, low vision subjects did not lose measurable visual acui-
ty during head motion.
228
4. Comment
These studies demonstrate that a significant adaptive VOR

gain increase can be induced and measured in normally sight-
ed and low vision subjects in response to the wearing of a tel-
escopic spectacle low vision aid. A measurable VOR gain
change occurs after only 15 minutes of passive head rotation
when such an aid is worn. The amount of plastic VOR gain
change varied from subject to subject, and was even absent in
a few. These studies did not attempt to'define optimum con-
ditions for inducing VOR gain plasticity, nor did they define
the limits, such as 6X or. 8X, of plasticity. Such determina-
tion will require further research. Since the low vision sub-
jects having the greatest plasticity also were able to use tele-
scopic spectacles most extensively in their visual
rehabilitation, it is tempting to speculate that a subject's
innate capability for VOR plasticity may determine that per-
son's capacity for visual rehabilitation with telescopic specta-
cles. Such a speCUlation would be premature, however, since
greater plasticity may be the result of, rather than the cause
of, extensive everyday experience with telescopically magnif-
ied vision.
These studies demonstrate that visual-vestibular interac-

tion (VVI) occurs in normally sighted and at least some low
vision subjects when 2X or 2.2X telescopic spectacles are
worn. VVI acts to increase gain of compensatory eye move-
ments during magnified vision to stabilize images on the reti-
na. VVI does not appear to be improved after adaptation to
2X or 2.2X telescopes under the conditions tested although
adaptation may be helpful under the more demanding condi-
tions imposed by more powerful telescopes. Poor VVI may be
a factor limiting usefulness of telescopic spectacle aids for
some low vision patients.
These studies demonstrate that visual acuity during head

motion (DVA) is significantly improved in some normally
sighted subjects after adaptation to telescopic spectacles.
This adaptation is the result of both VOR gain plasticity and
VVI. Both increase gain to reduce retinal image slip and
.improve visual acuity. The failure of DVA to improve after
telescopic adaptation in low vision subjects is likely to be an
artifact of the limited range of acuities available at the low
vision end of printed Snellen charts. Testing with charts hav-
ing more acuity lines in the low vision range may be expected
to show that DVA also improves in low vision subjects after
229
adaptation to telescopes. Adaptation may also have a more

important influence on DVA when more powerful telescopes
are employed. These preliminary data support the hypothesis
that insufficient ocular motor adaptation can cause failure to
achieve functional vision with head-mounted magnification
devices. Incompletely adapted subjects experience blurred
vision and incoordination, effects that would disable low
vision patients. Further study is needed to determine if eye
movement testing has clinical usefulness in the visual rehabil-
itation of low vision patients.
5. Acknowledgements
This research was supported by grant EY 06394 from the

National Eye Institute and by the Clayton Foundation for
Research. Kim Schmidt serves as clinical coordinator of
these studies. Rebecca Tyler prepared the manuscript.
6. References
1. Vision Problems in the U.S. (National Society to Pre-

vent Blindness, New York 1980)
2. R. Browning, R. Jose: The low vision triangle. Rehab.

Optom. 2, 12 (1984)
3. G. Westheimer, S.P. McKee: Visual acuity in the pres-

ence of retinal image motion. J. Opt. Soc. Am. 65,
847 (1975)
4. J .C.: Living without a balancing mechanism. N. Engl.

J. Med. 246, 458 (1952)
5. D.A. Robinson: Control of eye movements. In Hand-

book of Physiology, Sect. 1, The Nervous System, Vol.
1, Motor Control, Pt. 2, ed. by J.M. Brookhart, V.B.
Mountcastle (American Physiological Society, Bethes-
da 1981) pp. 1275-1320.
230
6. s.c. Cannon, R.J. Leigh, D.S. Zee, L.A. Abel: The

effect of rotational magnification of corrective spec-
tacles on the quantitative evolution of the VOR. Acta
Otolaryngol. (Stockh.) 100, 81 (1985)
7. H. Collewijn: Latency and gain of the rabbit's optoki-

netic responses to small movements. Brain Res. 36, 59
(1972)
8. J. Lanman, E. Bizzi, J. Allum: The coordination of eye

and head movement during smooth pursuit. Brain Res.
153, 39 (1978)
9. F .A. Miles, B.B. Eighmy: Long-term adaptive changes

in primate vestibulo-ocular reflex. 1. Behavioral
observations. J. Neurophysiol. 43, 1406 (1980)
10. J.L. Demer: The variable gain element of the

vestibulo-ocular reflex is common to the optokinetic
system of the cat. Brain Res. 229, 1 (1981)
11. A. Gonshor, G. Melvill Jones: Short-term adaptive

changes in the human vestibulo-ocular reflex arc. J.
Physiol. (Lond.) 256, 361 (1976)
12. A. Gonshor, G. Melvill Jones: Extreme vestibulo-

ocular adaptation induced by prolonged optical rever-
sal of vision. J. Physiol. (Lond.) 256, 381 (1976)
13. M. Shimizu, T. Yagi, S. Sekine, S. Baba, Y. Kobayashi,

T. Kamio: Test of vestibulo-ocular reflex using hori-
zontal vision-reversal prisms and visual suppression
test. Pract. Otol. (Kyoto) 74, 1691 (1981)
14. S. Sekine: Age changing effect of the vestibulo-ocular

reflex in humans. Pract. Otol. (Kyoto) 76, 1471 (1983)
15. G.M. Gauthier, D.A. Robinson: Adaptation of the

human vestibulo-ocular reflex to magnifying lenses.
Brain Res. 92, 331 (1975)
16. Y. Istl-Lenz, D. Hyden, D.W.F. Schwarz: Response of

the human vestibulo-ocular reflex following long-term
2X magnified visual input. Exp. Brain Res. 57, 448
(1985)
231
17. F.E. Guedry, J.M. Lentz, R.M. Jell: Visual-vestibular

interactions. 1. Influence of peripheral vision on sup-
pression of the vestibulo-ocular reflex and visual acui-
ty. Aviat. Space Environ. Med. 50, 205 (1979)
18. J.L. Demer, J. Goldberg, H.A. Jenkins, F.!. Porter:

Vestibulo-ocular reflex during magnified vision: Adap-
tation to reduce visual-vestibular conflict. Aviat.
Space Environ. Med., in press (1986)
19. C. Wall, F.O. Black, A.E. Hunt: Effects of age, sex,

and stimulus parameters upon vestibulo-ocular respon-
ses to sinusoidal rotation. Acta Otolaryngol. (Stockh.)
98, 270 (1984)
20. J. Goldberg, E. Icaza: Effects of imagined target dis-

tance on vestibulo-ocular reflexes. Soc. N eurosci.
Abstr. 11,322 (1985)
21. J.L. Demer, R.J. Tusa, S.J. Herdman: Unilateral cere-

bral cortical ablations impair vestibulo-ocular reflex
(VOR) gain and plasticity in the cat. Soc. N eurosci.
Abstr. 9, 868 (1983)
22. D.A. Robinson: Adaptive gain control of the vestibulo-

ocular reflex by the cerebellum. J. Neurophysiol. 39,
954 (1976)
23. G.M. Haddad, J .L. Demer, D.A. Robinson: The effect

of lesions of the dorsal cap of the inferior olive on the
vestibulo-ocular and optokinetic systems of the cat.
Brain Res. 185, 265 (1980)
24. C.C. Barr, L.W. Schultheis, D.A. Robinson: Voluntary,

non-visual control of the human vestibulo-ocular
reflex. Acta Otolaryngol. (Stock.) 81, 365 (1976)
Most Useful Visual Aids for the Partially Sighted
Gerald Fonda
1.
An accurate refraction can result in significant improvement

in distant vision in approximately 15 per cent of patients.
For distance, approach magnification (non-optical) and hand-
held telescopes are the most useful visual aids.
II
1.
l'
14
II:::
11
1.
i ·, Visual acuity
chart or
obIed of ,..ard
4
. . ..
2
......
... 0 •
F..t or inch.s
Figure 1. Approach magnification (non-optical)
Approach magnification is produced by reducing the dis-

tance between the observer and the object of regard. By
advancing from 6 m (20 feet) to 3 m (10 feet) the magnifica-
233
tion is 2X, by advancing from 6 m (20 feet) to 1.5 m (5 feet)

the magnification is 4X, and by advancing from 6 m (20 feet)
to 0.6 m (2 feet) the magnification is lOX. Approach magni-
fication is simple, natural, variable (the closer the greater),
and always available. It does not constrict the field of vision.
The application is great; for example, a magnification of 4X
can be produced in the classroom by moving from the back to
the front row and 20X by walking up to one foot. from the
blackboard. The greatest application is for television. A
patient with 1/60 (3/200) vision can enjoy baseball and foot-
ball by sitting 0.6 m (2 feet) from the screen. The television
is a friend to the partially sighted who is also deaf. However,
one needs to reassure the patient that color television is not
harmful even when he sits .3 m (one foot) from the monitor
for unlimited hours.
The main advantage of the 8X monocular focusable tele-

scope is great magnification and the range of focus from
infinity to 25 cm. This is not possible by use of binoculars or
telescopic spectacles. Vision of 6/60 (20/200) can be
improved to 6/7.5 (20/25); and vision of 6/120 (20/400) can be
improved to 6/15 (20/50) by the use of the 8X monocular tele-
scope. Visual improvement by a telescope is determined by
dividing the denominator of the Snellen fraction by the power
of the telescope.
The need for a telescope, which is infrequent and inter-

mittent, is best filled by a hand-held telescope. The occa-
sional need for a telescope occurs when the patient cannot
move close, such as in an auditorium or theatre, and to read
street signs and numbers. Many visually-impaired patients do
not experience binocular vision, so it is unnecessary to wear
binocular telescopic spectacles.
Near vision is aided by an accurate refraction, approach

magnifica tion (non-optical) and prescribing of optical aids,
such as half-eye spectacle magnifiers, hand magnifiers and
paperweight-type magnifiers. The approach magnification
has greater application for near than distance. Decreasing
the reading distance from 50 cm (20 inches) to 25 cm (10
inches) produces 2X magnification; to 12.5 cm (5 inches), 4X
magnification, and bringing the paper up to 5 cm (2 inches)
from the eye produces lOX magnification.
Many young patients with accommodative reserve, those

who are myopic, produce magnification by the approach
234
method. The approach method is flexible and variable,

because the patient can adjust the distance of the object
from the eye; hence, the magnification is produced according
to the size of the type to be read. The smaller the type, the
greater the magnification required; therefore, the closer he
must bring the paper to his eyes. The highly myopic patient
has a built-in magnifier, because he only needs to remove his
glasses and bring the paper to his far point. For example, a
20-diopter myope with 6/120 (20/400) vision can read the
newspaper (8-point type) at 5 cm from his eyes. Here the
examiner can tell the patient with pathological myopia that
he is lucky, because he has a built-in magnifier by taking his
glasses off. Reading at such a close distance is not harmful.
Those who read at close range because accommodation can be
exerted have a greater range of sharp vision, because their
far point is not restricted by the far point of the myopic eye
or the focal point of a plus lens, such as 10 cm, the focal
point of a +10.00 D lens.
Spectacle magnifiers can be prescribed in the form of

bifocals, trifocals and half-eye spectacle magnifiers. When
the distant correction makes significant improvement in dis-
tant vision, bifocals or trifocals are indicated. They are also
advantageous for a child or young adult who would be con-
spicuous wearing half-eye spectacles. The great disadvantage
of high-addition spectacle magnifiers is the close reading dis-
tance, because it is very difficult to read or work at distances
less than 25 cm (10 inches) when a person has been used to
reading at 40 cm (16 inches) for over 40 years. The closer the
reading distance, the greater the resistance for accepting a
spectacle magnifier.
Figure 2 shows a +6.00 D binocular reading addition with

segments decentered in 6 mm each. The diameter of the
right segment 4 mm below the top is 18 mm in comparison to
a diameter of 28 mm on the left segment. An individual
reads at a point 4 mm below the top of the segment. This
demonstrates the advantage of flat-top segments for high-
additions which increase the field of vision. However, in
practice, this does not prove to be the disadvantage as the
measurements suggest. An individual must read or work at
the short distance of 17 cm (7 inches) with the +6.00 D addi-
tion which is awkward.
Figure 3 shows the flat-top trifocal segment most com-

monly used, decentered in 6 mm each. For binocular vision,
235
Figure 2. High-addition bifocal. R: Add +6.00 D; 25 mm

round top segment. L: Add +6.00 D; 28 mm
flat-top segment.
bifocal or trifocal segments should be decentered in one mil-

limeter for each diopter of reading addition for each eye. If
segments are not adequately decentered, the patient will
experience double vision or discomfort. The advantage of the
trifocal is that it doubles the reading distance, permitting the
wearer to read larger type at a greater and more comfortable
distance.
For smaller type the paper is brought up to a distance of

17 cm (7 inches). Bringing the reading material from 34 cm
to 17 cm produces continuous vision and magnification of two
times.
The maximum binocular bifocal additions range from

+6.00 D to +12.00 D. The maximum binocular addition is
+8.00 D for trifocals. The maximum binocular reading addi-
tion for high plus such as for an aphakic correction is +6.00 D
because the strong convex distant correction produces a base-
out effect when converging to read at a close distance.
236
Figure 3. Trifocal: Reading addition +6.00 D. Interme-

diate addition +3.00 D. Segments dec entered in
6 mm each.
Stronger reading additions can be prescribed for monocu-

lar vision. The bifocal segment does not need to be decen-
tered except to align it with the pupil.
The strongest successful monocular trifocal that I have

prescribed is a +12.00 D, intermediate addition +6.00 D; focal
distance of +12.00 D reading addition is 8.3 cm (3.3 inches),
focal distance of intermediate addition +6.00 D is 16.6 cm
(6.6 inches). This increases the reading distance and range of
magnification 8.3 cm (3.3 inches).
Half-eye binocular spectacle magnifiers are most useful,

because the strongest binocular magnification can be pre-
scribed; stronger than a bifocal, because a frame with a 53
mm PD can be used and 221:1 prism diopters can be incorporat-
ed into the lens. However, I do not recommend half-eye
spectacle magnifiers stronger than +14.00 D with 181:1 base-in,
because the strong prism may blur the vision.
No weaker cylinder than 2.50 D is needed, because the

patient is reading by virtue of an enlarged image. The same
237
Table 1
Most useful half-eye binocular spectacle magnifiers
Magnifica- Approximate
Range of Powers tion Visual Indication
O.U. + 6.00 = 8~ base-in 1\ 6/18 to 6/30

20/60 20/100
O.U. + 8.00 10~ base-in 2 6/30 to 6/36
20/100 20/120
O.U. +10.00 12 6 base-in 2\ 6/36 to 6/60
20/120 20/200
O.U. +12.00 = 14 6 base-in 3 6/60 to 6/90
20/200 to 20/300
correction is prescribed for both eyes even when there is a

difference in the distant correction. The patient who has
monocular vision does not complain about a binocular correc-
tion.
The most useful non-spectacle magnifiers are the paper-

weight type and the hand magnifiers.
Figure 4 shows a +12.00 D hand magnifier held at a dis-

tance from the eye. The magnification is variable depending
upon the distance of the object from the focal point of the
lens. When the lens is adjacent to the object there is no mag-
nification; when the object is midway between the lens and
its focal point the magnification is 2X; and the maximum
magnification is obtained when the object is at the focal
point of the lens. However, the field of vision is small and
distortion is so severe that the maximum magnification is not
useful. For practical purposes the average magnification is
not more than 3X. Two times is the magnification generally
used with a +12.00 D hand magnifier. The great advantage of
hand magnifiers is that objects can be seen at a normal read-
ing distance.
Figure 5 shows a +12.00 D hand magnifier held at short

vertex distance same as a spectacle lens. Holding the lens
close to the eye increases the field of vision about three
times and produces constant magnification of 3X. This type
238
Figure 4. Hand magnifier used at distance from eye.
of magnifier is the most useful and versatile, because it can

be used at a distance from the eye, close to the eye and con-
veniently carried.
Figure 6 shows a range of the most useful powers to be

held at a distance from the eye which will generally fill the
need when a stronger magnifier is required. A +20.00 D hand
magnifier is the maximum strength to be used at a distance
from the eye, because the range of focus is 5 cm and the field
of vision is severely constricted.
The paperweight-type magnifier is most useful, because

it produces a constant magnification of 1.aX, diffusely
increases illumination, provides binocular vision and almost
doubles the magnification when used in combination with a
spectacle magnifier. Print can be observed by looking at an
angle of 45 degrees from perpendicular which cannot be
observed through a stand magnifier. The magnification of the
paperweight type can be increased to 2X by lifting it up two
to four millimeters from the paper.
239
Figure 5. Hand magnifier held close to eye.
Figure 7 shows the best spectacle and non-spectacle com-

bination. The paperweight type almost doubles the magnifi-
cation of the spectacle magnifier, increases the reading dis-
tance and produces binocular vision.
240
Figure 6. Range of most useful hand magnifiers to be

held at a distance from the eye: +6.00 D,
+12.00 D, +20.00 D
241
Figure 7. Paperweight-type magnifier used in combina-

tion with spectacle magnifier.
2. Summary
Most useful visual aids for distance are the approach (non-
optical) method and monocular 8X focusable telescope.
Most useful spectacle magnifiers for near vision are high

addition bifocals, high addition trifocals and half eye specta-
cle magnifiers. The advantage of high addition trifocals is
that they double the reading distance permitting an individual
to read larger type at a more comfortable distance. Trifocals
provide continuous vision and produce two times magnifica-
tion by bringing the paper from the focus of the intermediate
addition up to the focus of the reading addition.
Depending upon the distant correction, bifocals with

reading additions from +6.00 D to + 12.00 D produce binocular
vision and trifocals up to +8.00 D. Bifocal and trifocal seg-
242
ments must be dec entered in one millimeter for each diopter

of reading addition for each eye. Stronger reading additions
can be prescribed for monocular use. The segments must be
aligned with the pupil.
The most useful non-spectacle magnifiers are the

paperweight-type and hand-held magnifiers. The most useful
powers are +6.00 D, +12.00 D and +20.00 D. The great
advantage of hand magnifiers is that an individual can read at
a greater distance from the eye. Magnification varies with
the distance of the object from the focal point of the magnif-
iers.
The paperweight-type magnifier produces a constant 1.8X

magnification. It enables a person to read at a more natural
reading distance. The magnification is almost doubled when
used in conjunction with a spectacle magnifier.
Paradoxical Cases of Visual Improvement Offered by
Above Average Lighting Levels in Cases of
Albinism and Retinitis Pigmentosa
Sudhir Patel
1. Introduction
In cases of albinism, the patient is generally photophobic.

The eye possesses little/insufficient amounts of selectively
light absorbing pigment. Light absorption is limited by the
ocular media and entering flux limited by the pupil. In retini-
tis pigmentosa, however, the patient may still be photopho-
bic, even though in the early stages of the condition the
symptoms are those of poor visual performance in poor light
conditions [1]. Secondary to retinitis pigmentosa the patient
may have ocular complications such as cataract and glaucoma
[2]. The cataract is usually posterior sub-capsular and this in
itself can greatly affect visual performance [3], because the
cataract is close to the eye's nodal point. In advanced retini-
tis pigmentosa the sufferer is photophobic due to such medial
opacities. In both albinism and retinitis pigmentosa we can
assume that tinted optical appliances are of great help and
this is found to be true [4]. However, in routine low vision
care, we have found cases where visual performance is fur-
ther improved by the use of refractive error correction and/
or low visual aids in conjunction with the careful use of light
at above average illuminance levels.
2. Materials/Methods
Of albinism and retinitis pigmentosa patients attending a rou-

tine low vision clinic, 10 were selected whose problems were
concerned with near vision, regardless of the specific type of
albinism and retinitis pigmentosa suffered. The 10 were indi-
vidually assessed and corrected in the most appropriate and
comfortable manner. For all 10 patients ambient lighting was
controlled using a 60 Watt angle-poise lamp or a lamp built
244
into an illuminated low visual aid. The preferred light level

and direction was determined:
i) by the clinician, manually, by adjusting the position of

the angle-poise lamp in relation to the patient and print of
regard, and
ii) in the illuminated aid by manually determining the

preferred direction of illumination (e.g. from right, left, top,
bottom, or intermediate) and adjusting the rheostat control-
ling the lumen output.
Later the illuminance produced by each subjectively pre-

ferred lighting system was measured usin&r~ light meter
(Spectrophotometer "301" by Spectraphysics ). Each sub-
ject was individually tested in a common room using the same
reading material in order to minimize effects of varying
reflectance values and room/background lighting levels.
3. Results
All results are set out in Tables 1 and 2.

245
Table 1
A: Albinism; RP: Retinitis Pigmentosa; N: Nystagmus; C:

Cataract; .... : Significant Refractive Correction
SUBJECT LOW VISION REFRACTIVE CORRECTED. I LIGHTING/AID.

CAUSE ERROR NEAR SYSTEM USED.
VISION
S McG A & 'N ,I Astigmatism J6 60 \~att lamp

over right
shoulder
A McG A &N ,I Astigmatism J6 60 Watt lamp

over right
shoulder and
near addition
JA A &N ,I Astigmatism J6 60 Watt lamp

over either
shoulder and
near addition
B McW A&N Nil J6 Telescopic aid

& 60 Watt lamp
TG RP & C Nil J2 XIO Eschenbach
RP RP & C Nil J12 60 Watt lamp

from left side
AG RP & C Nil J12 X5 Rayli te
SAW RP & Aphakia ,I J6 X7 Rayli te
AD RP ,I J4 60 Watt lamp
from left side
& near addition
DH RP & C Nil J12 XIO Eschenbach

246
Table 2
llluminance Characteristics of Lighting Systems Used. Pho-

topic filter used during illuminance measure
SYSTEM/AID ILLUMINANCE (LIDO
X10 Eschenbach 425
X5 Raylite 235 (min) 478 (max)
X7 Raylite 195 (min) 540 (max)
60 Watt frosted lamp 375 (min) 600 (max)
Background 295
4. Discussion
The results of Table I show that in this short study all sub-
jects preferred a lighting system for near vision use, even
though almost all subjects complained of photophobia to some
degree (e.g. during a bright day or even on a dull day with
overcast sky). Why should this be? In all cases the working
distance necessary in order to perform the near vision task is
low. The hand-held illuminated devices are provided with a
guard which ensures correct positioning of the reading mat-
ter. The guard shields against all external light and obviously
must be compensated for. In all cases where such a device
was preferred, the subject actually preferred the lamp to be
on "full". The subjects who used only their refractive error
correcting spectacles tended to hold print very closely. As
Taylor [4] notes, the albino will hold print up to 4 inches from
the face to compensate for the lack of macula pigment. In
such cases the shadow of the head coupled with the nearness
of the task will reduce overall luminance. In conclusion, an
247
increase in lighting is required because of the relative light

reduction produced by the nature of the task and how the task
is performed. Nevertheless, above local levels were required.
The 1977 IES value for general offices is 500 lux [5]. Such a
level is rarely achieved in practice but it would appear that
this level is 'more preferred' by this particular group of sub-
jects. Each subject preferred a direction for illumination. It
appears that this direction ensures the minimal amount of
intra-ocular light scatter (and therefore comfort) in relation
to the subject's individual ocular condition and state of visual
disability. Once the preferred direction is found, the clini-
cian should determine the ideal lighting level. One must
remember that this is a time-consuming procedure, but it
should be performed in order to achieve the best result.
5. Conclusion
For near vision, in cases of albinism and retinitis pigmentosa,

consideration of the direction and levels of lighting ought to
be made. The careful application of light may subjectively
improve visual status and performance, even when the patient
complains of "photophobia".
248
6. References
1. H.F. Falls: Retinitis pigmentosa-like syndromes. In

Retinal Diseases, Symposium on Differential Diagnos-
tic Problems Of Posterior Uveitis, University of Cali-
fornia, San Fransisco, 1965, ed. by S.J. Kimura, W.M.
Caygill (Lea & Febiger, Philadelphia 1966) pp. 262-6.
2. J. Cuthbert, R.M. Clayton: Cataract in association

with retinitis pigmentosa: Analysis of the crystallin
subunit composition. In Problems of Normal and
Genetically Abnormal Retinas, ed. by R.M. Clayton, J.
Haywood, H.W. Reading, A. Wright. (Academic, Lon-
don 1982) pp. 369-75.
3. R. Hess, G. Woo: Vision through cataracts. Invest.

4. W.O.G. Taylor: Aiding vision in albinism. Trans.

Ophthalmol. Soc. U.K. 104, 309 (1985)
5. llluminating Engineering Society, London: IES Code

General Schedule: Offices and Shops (IES, London
1977)
Magnification Efficiency in the Low Vision Patient
Franklin I. Porter, Joseph L. Demer
1. Introduction
Visual impairment, or the loss of visual acuity and/or visual

field due to some pathology of the eye or brain, afflicts an
estimated 6 million people in the United States [1]. In the
State of Texas, more than 392,000 people are said to have a
best corrected visual acuity of less than 6/21 (20/70) [2]. One
method of maximizing visual functioning in such persons is
through the use of the unique services available at low vision
rehabilitation clinics. Employing specialized refracting tech-
niques and using optical devices such as telescopes and micro-
scopes, the low vision clinician strives to maximize remaining
residual vision. Improvement of visual acuity can be attained
by enlarging the object size, as by magnification; accenting
definition, as by emphasizing functionally significant spatial
frequencies; or varying the illumination, as by altering the
contrast [3]. Telescopes have long been used in visual reha-
bilitation. These devices enlarge distant objects; with modi-
fication telescopes will magnify near objects as well [3]. The
telescopes may be prescribed for academic, vocational, rec-
reational, and mobility uses [4].
It has been estimated that 15 to 20% of the patients seen

at a low vision clinic are prescribed some form of telescopic
device [5]. Often, however, the use of a head mounted mag-
nification device does not provide the anticipated theoretical
visual benefit. The adverse cosmetic appearance, expense,
and long period of training needed to adapt to head mounted
magnification devices are major limitations on their empiri-
cal trial in individual patients [4]. Therefore, the prediction
of the effectiveness of telescopes for individuals undergoing
visual rehabilitation would be invaluable.
The magnification efficiency factor (MEF) is a measure

of the effectiveness of head mounted magnification devices
in improving visual acuity. The factor is expressed by the
formula:
250
Actual Acuity with a Telescope

MEF
Predicted Acuity with a Telescope
where predicted acuity with a telescope equals magnification

of a telescope times unmagnified acuity. The ideal MEF is
equal to UI\ity. For example, if unmagnified acuity for a giv-
en low vision patient is 6/30 (20/100), the predicted acuity
using a 2X telescopic spectacle would be 6/15 (20/50). If
actual acuity with the telescope were also 6/15 (20/50), then,
MEF 20/50 20/50

1
(2X) (20/100) 20/50
If, as is often the case, the actual acuity with the telescope
were only 6/18 (20/60), the MEF would be 0.84, representing
an improvement in visual function of only 84% of that pre-
dicted. In this study, we measured actual MEF's for 4X tele-
scope spectacles in a group of low vision patients using Snel-
len acuity and contrast sensitivity measurements.
Snellen acuity measurements resolve high contrast visual

stimuli of varying visual angles. Snellen testing thus evalu-
ates only one dimension of visual functioning. It has been
suggested that this limitation of Snellen acuity can be over-
come by testing contrast sensitivity, which measures both
spatial frequency and contrast thresholds [6]. Use of a con-
trast sensitivity testing procedure provides a multidimension-
al approach. The contrast sensitivity function (CSF) should
therefore provide a method of measuring visual functioning
more sensitive than Snellen acuity.
As the CSF is a multidimensional measurement of visual

functioning encompassing a broader range than the Snellen
acuity measurements, it was included in our study. Ideally,
spatial frequency threshold, the number of cycles per degree
of visual angle, should be increased by telescopic spectacles
in proportion to the telescopic spectacle magnification fac-
tor.
251
Contrast threshold is defined as the level of contrast that

is barely perceptible to the subject and its reciprocal is the
contrast sensitivity. When viewing an object through a tele-
scope or other compound optical system, though the general
illumination of a field is changed, the apparent color and
brightness of objects do not change proportionately but tends
to remain constant, approximating the appearance under
unmagnified illumination [7]. We theorize that there should
be little change in relative brightness using telescopic specta-
cles. If this is the case, the improvement in the contrast sen-
sitivity function produced by telescopic spectacles should be
predictable from the telescopic magnification factor shown in
Fig. 1. Figure 2 is the graphical representation of the CSF of
a typical low vision patient. The spatial frequency threshold
is at 4 cycles per degree (c/deg), roughly correlating with
6/36 (20/120) Snellen. Figure 3 demonstrates the theoretical
effect of telescopes on the CSF; note the one octave shift for
the spatial frequency for a 2X telescope and the two octave
shift for the 4X telescope. In this study, actual contrast sen-
sitivity functions of low vision patients were measured both
unmagnified and using telescopic spectacles. This enables
comparison of the actual benefit of telescopic spectacles
with the theoretical benefit.
252
RETINAL IMAGE OF
MAGNIFIED/UNMAGNIFIED GRATINGS
On Cha" On Cha"
Unmagnlllod Unmagnified
On Rollnl 2. On Retina
Unmognlfled
On Aetlna 4x
On Rlllnl
Unmagnltlod
Figure 1. Sinusoidal contrast sensitivity gratings as they

would appear on the retina when viewing the
gratings under unmagnified conditions and
through 2X or 4X telescopes.
253
EXAMPLE LOW VISION

CONTRAST SENSITIVITY FUNCTION
1000 0.001
__ 0
~ --
:;
...J
0
300 0.003
:l:
~ III
iii --w
zw a:
0.01 :l:
III ~
~ -- III
~
III
c( _ 20/120 c(
a: 30 4 0.03 a:
CPO Snellen __ Z
/
~ ~
Z
0 -- 0
U U
10 0.1
12 18
SPATIAL FREQUENCY
CYCLES PER DEGREE
Figure 2. Hypothetical CSF for a low V1S1on patient

wearing only a spectacle refractive correction.
The curve is extrapolated to intercept the spa-
tial frequency axis; the intercept of 4 cycles
per degree correlates roughly with a Snellen
visual acuity of 6/36 (20/120). Contrast thresh-
old is the inverse of contrast sensitivity.
254
EXAMPLE LOW VISION

CONTRAST SENSITIVITY FUNCTION
WITH TELESCOPES
1000 0.001
__ 0
> --
I- ..I
:; 300 0.003 0
j: :I:
I/)
iii - - -- w
Z II:
W 100 0.01 :I:
I/) I-
1-
I/)
-- - - Ii;
'"
II:
I-
30
2X TELESCOPE
0.03
'"
II:
I-
Z
~ -- --0
0 0
10 0.1
1.5 12 18
SPATIAL FREQUENCY
CYCLES PER DEGREE
Figure 3. Hypothetical CSF for the example patient in

Fig. 2, as ideally improved by 2X or 4X tele-
scopic spectacles. The 2X telescope increases
the spatial frequency intercept by one octave
from 4 to 8 c/deg (CPD) while the 4X telescope
increases the intercept by 2 octaves, to 16
c/deg (CPD). Maximum contrast sensitivity is
hypothesized to be unaffected by the magnifi-
cation.
2. Materials, Methods and Procedures
Thirty four randomly selected patients of the Baylor College

of Medicine/Cullen Eye Institute and the Veterans Adminis-
tration Medical Center Houston Low Vision Clinics partici-
pated in this study. Patients ranged in age from 13 to 89 with
an average age of 56.2 years. There were 19 females and 15
males; a total of 51 eyes were tested. Best refractive cor-
rection was obtained using standard low vision techniques [4].
Subjects accepted in the study met the following criteria:
visual acuity less than 6/21 (20/70), no prominent nystagmus,
and a central visual field greater than 20 degrees. Test tar-
gets, both Snellen and Vistech VCTS 4#=6500 contrast sensitivi-
255
ty charts, were viewed at 3 m (10 feet). Luminance at thi

surface of the contrast sensitivity chart was 103-240 cd/m
or 30-70 ft.-Iamberts.
Snellen visual acuity was determined using either a stan-

dard Snellen chart or Mentor BVAT video display. Acuity was
recorded for each eye both with and without a 4X close focus
telescope commercially available from Designs for Vision Inc.
Determinations were then repeated using the CSF chart. In
both cases, best refractive correction was placed in the trial
frame and the telescope was mounted to a trial frame ring.
Testing procedures for the use of the CSF chart were fol-
lowed using standard techniques [8]. Data were recorded and
placed in patients' charts and a duplicate was filed for evalu-
ation and assessment. Graphical analysis was performed on
each subject and the curves were extrapolated to find the
intercept on the spatial frequency axis.
Legal requirements regarding informed consent were

met.
3. Results
The MEF was calculated for 51 eyes of 34 subjects, and is dis-

played in the form of a histogram in Fig. 4. For all subjects,
mean MEF was 0.83, with a standard deviation of 0.28. By
Student's t test, mean MEF was thus significantly less than
unity at the p = 0.0005 level. The mode of the MEF distribu-
tion in Fig. 4 lies between 0.7 and 0.8, indicating that the
most frequently achieved MEF was also less than unity. Nev-
ertheless, there was a large individual variation in MEF: the
range was from 0.42 to 1.67. MEF exceeded 1.2 in 15 eyes.
CSF was also measured in 51 eyes, both under conditions

of unmagnified vision and using 4X telescopic spectacles.
The use of telescopes enhanced the spatial frequency thresh-
old in most subjects. A typical example of an improved spa-
tial frequency threshold is seen in Fig. 5. With telescopic
vision, the spatial frequency threshold was improved by one
octave or less in 19 eyes, by two octaves in 22 eyes, and by
more than two octaves in 10 eyes.
256
MAGNIFICATION EFFICIENCY - 4X TELESCOPES

10
I/)
W 7
>-
~ 6
o
a:
w
ID 4
::I!
i 3
MAGNIFICATION EFFICIENCY FACTOR
Figure 4. Histogram of magnification efficiency factors

achieved by low vision patients using 4X tele-
scopic spectacles.
A majority of subjects experiencing spatial frequency

threshold improvements from telescopic vision also exhibited
increases in contrast sensitivity. The improvement in both
the spatial frequency and contrast threshold is seen in Fig. 6,
where CSF's are plotted for a long-term telescopic spectacle
user with and without the use of 4X magnification. Forty-
three percent of the eyes that experienced spatial frequency
threshold improvements from telescopic vision did not exhibit
improvements in contrast threshold. There was no correla-
tion between an individual subject's MEF and his improvement
in spatial frequency threshold.
257
1000 .001
o
~ 300
::; .003 5
l:
i= t/)
iii W
a:
zw l:
I-
~ 100 .01 l-
t/) t/)
C C
a: a:
I- I-
Z z
o o
() 30 .03 ()
10
1.5 3 6 18
SPATIAL FREQUENCY
(CYCLES PER DEGREE)
NAME: BW
Figure 5. Actual eSF's measured with Vistech VeTS

4#=6500 chart of a low vision patient with and
without the use of 4X telescopic spectacles.
Note the increase in spatial frequency threshold
without change in contrast sensitivity.
258
1000 .001
o
~ 300
:; .003 5
j: J:
I/)
iii W
a:
z
w J:
I-
~ 100 .01 l-
,,....
I/) I/)
c( c(
a: a:
I-
Z
o ," " ~---.,
,,
-.,-----~
I-
z
o
U 30 ,, .03 U
,,
10
~
1.5 3 6 12
\ 18
.1
SPATIAL FREQUENCY
(CYCLES PER DEGREE)
NAME: HS
Figure 6. Actual eSF measured with Vistech VeTS #6500

chart of a low vision patient with and without
the use of 4X telescopic spectacles. Both spa-
tial frequency threshold and contrast sensitivity
improve with telescopic spectacles.
4. Comments
This study has investigated the effectiveness of 4X telescopic

spectacles in improving visual function. A measure of the
effectiveness of telescopic spectacles in improving Snellen
visual acuity has been devised; this is the magnification effi-
ciency factor (MEF). The data in Fig. 4 clearly indicate that
while MEF varies substantially from low vision subject to sub-
ject, the expected value of MEF for a low vision population is
less than the ideal value of unity.
259
The failure of average MEF to reach unity indicates that

some factor or fac.tors limit the efficiency of 4X telescopic
spectacles. Possible optical factors that may reduce effi-
ciency include geometric and chromatic distortion, and
reduced light gathering by the telescopic lens system. Senso-
rimotor factors that may reduce magnification efficiency
include inappropriate ocular motor reflex compensation for
involuntary head movements. Incomplete adaptation of ocu-
lar stabilization reflexes to telescopic spectacles produces
instability of images on the retina resulting in reduced visual
acuity [9]. Regardless of the optical or sensorimotor causes
of reduced MEF, the anticipated MEF for any individual sub-
ject may depend on the power of the telescope chosen, as has
been our clinical impression.
Although average MEF for the low VISIon population

appears to be deficient in that it is below unity, a substantial
number of eyes using 4X telescopes achieve MEFs exceeding
1.2. This implies that such eyes achieve a functional visual
benefit greater than that provided by 4X magnification alone.
The causes of this effect are unknown, but may include
reduction in glare effects from reduced light gathering by
telescopes, and reduced confusion by masking of the peripher-
al visual field by the telescopes.
If the MEF is employed in the clinical evaluation of low

vision patients, it may be prognostically useful. Patients hav-
ing low MEF's may be expected to derive little benefit from
head mounted magnification devices, while those having
MEF's of unity or greater might obtain functional benefit.
Further clinical testing of this hypothesis is required.
This study also investigated the effect of 4X telescopic

spectacles on the contrast sensitivity functions of low vision
patients. The actual changes produced by telescopic specta-
cles were compared with the expected changes predicted on
the basis of a magnification effect alone (see Figs. 2 & 3).
Based on this prediction of a two octave improvement in spa-
tial frequency threshold resulting from 4X magnification,
37% of eyes achieved less than the predicted improvement,
43% achieved the predicted improvement, and 20% achieved
better than the predicted improvement. The improvement in
spatial frequency threshold for individual subjects did not
appear to be correlated wi.th MEF. This suggests that Snellen
acuity and spatial frequency are distinctly different visual
functions in the low vision patient. Each function may have
independent prognostic importance to the clinician.
260
Just as some low vision patients exhibited improvements

in Snellen acuity greater than would be predicted from tele-
scopic magnification alone, many also exhibited large
improvements in contrast threshold not accountable by simple
magnification alone. The sources of this improvement are
likewise obscure, but might include reduced retinal light level
due to light gathering characteristics of the telescopes.
Further investigation will be required to define the clini-

cal usefulness of measurements of the MEF and CSF in reha-
bilitation of low vision patients using telescopic spectacles.
Studies employing telescopic spectacles of varying powers are
underway to evaluate MEF and CSF to predict functional vis-
ual rehabilitation.
5. Acknowledgements
This research was supported by grant EY 06394 from the

National Eye Institute. Kim Schmidt serves as Project Clini-
cal Co-ordinator of these studies. Cathy Nuckolls, Low
Vision Clinic Co-ordinator, assisted in data collection. M.
Robin Porter provided editorial assistance.
6. References
1. J.A. Kraut: The low-vision examination. In Manual of

Refraction, 3rd ed., ed. by A.E. Sloane, G.E. Garcia
(Little, Brown, Boston 1979) pp. 173-184.
2. Vision Problems in the U.S. (National Society to Pre-

vent Blindness, New York 1980)
3. J.D. Newman: Telescopic systems. In A Guide to the

Care of Low Vision Patients ed. by J.D. Newman, F.A.
Brazelton, S.C. Miller, G.M. Milkie (American Opto-
metric Society, St. Louis 1974) pp. 102-112.
4. E.E. Faye: Clinical Low Vision (Little, Brown, Boston

1976)
261
5. G.O. Hellinger: A study of the degree of persistance

of clinical gains received with patients fitted with low
vision aids. Optom. Wkly.57(15), 29 (1966) 57(16), 23
(1966)

7. D.D. Michaels: Visual Optics and Refraction: A Clini-

cal Approach (Mosby, St. Louis 1975)
8. A.P. Ginsburg, D.W. Evans, M.W. Cannon, C. Owsley,

P. Mulvanny: Large sample norms for contrast sensi-
tivity. Am. J. Optom. Physiol. Opt. 61, 80 (1984)
9. J.L. Demer, J. Goldberg, H.A. Jenkins, F.!. Porter:

Vestibulo-ocular reflex during magnified vision: Adap-
tation to reduce visual-vestibular conflict. Aviat.
Space Environ. Med. (1986, in press)
An Overview on the Use of a
Low Magnification Telescope in Low Vision
George C. Woo
1. Introduction
A Galilean telescope in its simplest form is a two element

system consisting of a positive lens as an objective and a neg-
ative lens as an eyepiece. The system is restricted to lower
magnifications and smaller fields of view in comparison with
a Keplerian telescope. The image through the system, how-
ever, is always erect permitting its use for distance viewing
for partially sighted patients. Other optical factors besides
magnification and field of view that need to be considered
include exit pupil size, focus adjustability, vertex distance,
and image quality in terms of color and brightness. Such non
optical factors as weight, portability, ease of use, appearance
and cost are also influencing variables [1]. In this overview
on the clinical use of low power telescopes in the examination
room, only a few properties will be examined. The use of a
low power full-field telescope in sUbjective and objective
refractions will be discussed. Magnification through a tele-
scope will also be elaborated upon.
1.1 Low Power Telescopes
BIER [2] states there are two basic types of Galilean tele-
scopes used in low vision. They are available either in fixed
focus form for insertion in ordinary spectacle frames or in
variable focus form available commercially. These full field
telescopes are relatively inexpensive and are commonly pre-
scribed as distance aids for low vision patients. Although no
prescription may be incorporated into the eyepiece of these
2.5X or 2.8X low power telescopes, the refractive error of
the patient can be compensated, in the form of equivalent
spheres, by altering the telescope from an afocal to a focal
system. This is achieved by altering the distance between the
eyepiece and the objective. According to EMSLEY [3] this
method was first proposed by von Graefe in 1863. When these
263
telescopes are used by emmetropes, the amount of accommo-

dation required through the telescope can be obtained by
increasing the lens separation. Uncorrected ametropes can
also use these devices by increasing or decreasing the dis-
tance between the eyepiece and the objective lenses depend-
ing upon the error of refraction. There are, however, some
limitations. The separation between the objective and the
eye piece is limited in length and the field of view is general-
ly restricted to 15 degrees or less. The proximity of the
object as viewed by the patient will alter the image vergence
at the ocular to require a large amount of accommodation [4].
On the other hand, instrument myopia or instrument accom-
modation through the telescope may induce accommodation
which partially counteracts the demand for accommodation
through a telescope.
2. Magnification of Low Power Focal Telescopes
Afocal telescopes are used focally by emmetropic users look-

ing at objects other than at infinity or ametropic users with-
out any correction. The question is whether or not the actual
magnification of the telescope will be changed significantly
from the nominal magnification.
If the position of the object is relatively fixed in relation-

ship to the focusable low power telescope, an appropriate
definition of magnification would be the relative size of the
retinal image after the telescope is introduced to the eye
without the telescope (Figs. 1 and Z). The value of that ratio
depends on whether or not the user changes the relative posi-
tion of his eye and the object when he uses the telescope.
There are two extreme cases. In the first the user merely
places the telescope into the space between himself and the
object of regard, adjusting the telescope length until the
object is in focus. In this case the distance from the eye to
the object is the same with and without the telescope. Spec-
tacle magnification in this case will be designated Ml [5].
In the second case the user places the objective lens of

the telescope in his spectacle plane and focuses while holding
the position of the objective lens constant, moving his head
back and forth as the position of the eyepiece is adjusted.
Magnification will be designated MZ. For practical interest
Ml > MZ·
264
< • _____ ~) ~ k'., ~
Figure 1. Geometry of retinal image formation by an

unaided eye (from LONG and WOO [5]).
Figure 2. Geometry of retinal image formation by an

eye-telescope system (from LONG and WOO
[5]).
In practice a user may move both ocular and objective

while focusing a telescope so that his actual retinal magnifi-
cation will lie between M1 and MZ. These values should be
interpreted as the limiting value of retinal magnification. In
265
most cases of interest, the numerical value of the two kinds

of magnification do not differ very much from each other.
LONG and WOO [5] derived a general expression for spec-

tacle magnification. It is used to determine the magnifica-
tion of focal telescopes for correcting ametropia and/or for
viewing objects at finite distances. The results of careful
calculations of retinal image magnification show that changes
in telescope magnification large enough to alter acuity by a
line on a LogMAR chart occur only when the viewing distance
is less than 60 cm or when a Galilean telescope used at dis-
tance is adjusted to compensate for a very large ametropia.
In practice, the expression reveals a clinically negligible dif-
ference between spectacle and nominal magnification [5].
M
s
3. Refraction
3.1 Subjective Refraction with the Use of a Low Power

Telescope
In a low vision assessment, the use of a low power telescope

will often enable the practitioner to evaluate the refractive
status of a patient. Refraction in equivalent spheres can be
determined by asking the patient to move the knurled knob
(Fig. 3) slowly either clockwise or counter clockwise until the
visual acuity chart is in best focus at 6 m. The practitioner
can of course choose to turn the knob instead. Provided the
patient does not accommodate at a distance considerably
closer than the location of the visual acuity chart, the indica-
tion is that the patient is hyperopic when the knob is turned
clockwise. Turning the knurled knob clockwise is equivalent
to lengthening the distance between the eyepiece and the
objective. When the knob is turned counterclockwise thus
shortening the separation between the eyepiece and the
objective, the patient is shown to be myopic.
266
Figure 3. Use of a calibrated telescope in subjective

refraction.
The amount of ametropia can be quantified by noting the

setting relative to the zero position. On most of these
devices, there is a zero position marked in red. On either
side of this red mark are inscribed white marks at equal
intervals. Table 1 gives measurements of back vertex powers
of a typical 2.5X monocular achromatic telescope taken from
a Sportscope when it is placed in the lensometer and adjusted
to different positions away from the zero position. The minus
sign indicates shortening of the separation· and that the ver-
gence leaving the eyepiece is divergent: the plus sign indi-
cates lengthening the separation and that the vergence leav-
ing the eyepiece is convergent. Table 1 shows that although
positions of white marks are separated linearly at equal inter-
vals, they are not linearly related dioptrically.
Repeated measurements on a number of 2.5X monocular

achromatic telescopes of the same type yield similar findings.
Occasionally, there is a difference of +/- 0.25 D which can be
attributed to a misalignment of the individual white mark
from the zero position when the reading is taken through the
lensometer.
267
Table 1
Calibrated back vertex powers of a 2.5X telescope at specific

intervals.
Counter-
clockwise BVPln Clockwise BVPin
marks dlopters marks diopters
----
-1 -0.75 +1 +0.75
-2 -1.50 +2 +1.25
-3 -2.00 +3 +1.75
-4 -2.50 -+4 +2.50
-5 -3.00 +5 +3.00
-6 -3.50 +6 +3.50
-7 -3.75 +7 +4.25
-8 -4.25 +8 +5.25
-9 -4.50 +9 +6.25
-10 -5.00 +10 "7.25
Another consideration is that the demand on accommoda-

tion through the telescope is substantially greater than indi-
cated by the object distance. The demand is due to the image
vergence at the ocular when an object is imaged through the
telescope. AnI appr~ximate formula used to calculate the
vergence is L2 = M Ll[4], where M is the magnification of
the telescope Ll is the object vergence at the plane of the
objective, and L2 is the image vergence at the plane of the
eyepiece.
For low magnification telescopes of 1. 7X and 2.5X, the

respective discrepancy is 0.50 D and 1.00 D when the distance
is 6 m. A 4X telescope on the other hand requires an accom-
modation of approximately 2.75 D for the same object dis-
tance. Thus a compensating lens should be incorporated in
the calibration of a 4X telescope if it is to be used for subjec-
tive refraction.
The use of a calibrated telescope in determining the

refractive error of low vision patients clinically has been
reported by WOO [6]. The technique, however, is far from
being precise as pointed out by BAILEY [1]. It does not, for
example, provide accurate astigmatic corrections. Although
268
instrument accommodation would generally counter the

effect of image vergence through a telescope, the net result
is determined by the accommodative state of the patient,
thus contributing to the inaccuracy of sUbjective refraction
through a telescopic device. Information on the amplitude of
accommodation of low vision patients is essential to obtaining
a more accurate refraction.
In order to obtain a more accurate reading, the examiner

could "fog" the patient by lengthening the distance between
the objective and the eyepiece of the telescope after the sub-
jective refraction has been established either by the patient
or the examiner. This procedure is similar to the fogging
technique employed by some automated refraction systems
and by conventional sUbjective refraction.
Occasionally, however, it is impossible to determine

refractive errors of some low vision patients using conven-
tional refractive techniques including radical retinoscopy. It
is in these cases that the technique becomes invaluable in
estimating the refractive error subjectively.
3.2 Objective Refraction with the Use of a Low Power

Telescope
Retinoscopy is an objective technique used to measure

refractive errors. In retinoscopy there are two systems in
operation [7]. The illumination system begins with the light
source of the instrument and ends in the patient's retina and
the observation system begins on the patient's fundus as a
light patch and ends in the examiner's eye. The retinoscopic
finding determines the amount and type of refractive error by
bringing the far point of the patient's eye coincident with the
plane of the retinoscope. When this occurs, the examiner has
reached the "neutral" point or the "flashing" point in retinos-
copy. With the use of a low power telescope, the same prin-
ciple holds. The required emerging vergence from the objec-
tive can be obtained by having the illuminated retinal patch
focused at the examiner's entrance pupil. A schematic dia-
gram illustrates this principle in Fig. 4. As an example, a
typical 2.5X telescope would have the following values:
269
F1 +15.00 D (objective)
FZ -37.50 D (eyepiece)
separation distance of F1 and FZ = 4 cm

working distance = 60 cm
par;""t io-t..+t.I-opo,-4.~'-------ki"3 cI;.""~... ---~ ~

...... ~
dis~
schematic diagram
Figure 4. Schematic representation of telescopic retinos-

copy.
Given the above values the required emerging vergence

from the objective would be +1.67 D in order to have the illu-
minated retinal patch focused at the observer's entrance pupil
when arriving at the neutral point in retinoscopy. The calcu-
lated yeutr:2lizing lens using the simplified vergence formula
[4] LZ = M L1 is equal to +10.44 D for an emmetrope. Thus
by calculation it is predicted a neutralizing lens of approxi-
mately +10.50 D is required for an emmetrope. The tech-
nique is similar to loose lens retinoscopy whereby lenses are
inserted behind the eyepiece at the spectacle plane at regular
intervals until the "with" movement of the reflex is neutral-
ized. Insertion and removal of lenses in this manner are cum-
bersome and the vertex distance is not always maintained.
Alternative methods such as increasing the separation
between the objective and eyepiece by a specific amount and/
or placing low power reading caps in front of the telescope
can reduce the amount of positive lenspower required at the
spectacle plane immediately adjacent to the eyepiece.
270
Thirty normal subjects were sUbjected to telescopic reti-

noscopy. Preliminary data indicate correspondence between
the experimental results and the predicted values. Discrep-
ancies however do occur. These can be attributed to sources
of error usually attributed to retinoscopy including 1. inexact
working distance 2. scoping off the patient's visual axis 3.
failure to obtain a reversal 4. failure to locate the principal
meridians 5. failure to recognise scissors motion and 6. fail-
ure of the patient to fixate the distance target [8]. It is our
observation that there appears to be a zone of neutrality once
the "neutral" point is reached contributing perhaps to much of
the discrepancy in most cases. To verify this, we are plan-
ning to refract two groups of subjects with and without cyclo-
plegia. These results would then tell us whether accommoda-
tion could be another influencing factor in telescopic
retinoscopy. In addition to conventional sources of error in
retinoscopy, there are other factors to be considered in tele-
scopic refraction. These include vertex distance, the tilt of
the telescope, aberrations of the telescope, Modulation
Transfer Function (MTF) of the telescope, alignment of the
exit pupil of the telescope with the pupil of the eye and the
brightness of the reflex. Clinically there is little or no
advantage obtained by refracting low vision patients with a
telescope. The employment of such a technique will not pro-
vide additional information in refraction. Observation of
media opacities in some low vision patients could occasionally
be made easier because of the larger reflex seen by the
examiner.
4. Summary
For practical purposes actual magnification of a lower power

focal telescope and its nominal magnification may be viewed
as being identical. The use of a low power telscope in refrac-
tion has been described. This technique for subjective refrac-
tion can be useful in estimating the refractive error of some
low vision patients. The use of the same device in objective
refraction provides little clinical advantage.
271
5. Acknowledgements
I thank W.F. Long, C. Machan, D. Otto and L. Hanna for their

assistance. The project was supported in part by an E. Burton
low vision research fund.
6. References
1. I.L. Bailey: Telescopes - their use in low vision.

Optom. Mthly. 69, 634 (l978)
2. N. Bier: Correction of Subnormal Vision, 2nd ed. (But-

terworths, London 1970) p. 32.
3. H.H. Emsley: Visual Optics, Vol. I (Hatton Press, Lon-

don 1955) pp. 100-101.
4. A.N. Freid: Telescopes, light vergence and accommo-

dation. Am. J. Optom. Physiol. Opt. 54, 365 (l977)
5. W.F. Long, G.C. Woo: The spectacle magnification of

focal telescopes. Ophthalmic Physiol. Opt. 6, 101
(l986)
6. G. Woo: Use of low magnification telescopes as

optometers in low vision. Optom. Mthly. 69, 529
(l978)
7. D.D. Michaels: Visual Optics and Refraction - A Clini-

cal Approach, 3rd ed. (C.V. Mosby, St. Louis 1985) p.
297.
8. T.P. Grosvenor: Primary Care Optometry - A Clinical

Manual (Professional Press, Chicago 1982) pp. 157 -
158.
Observations from the Psychology of Reading
Relevant to Low Vision Research
John Baldasare and Gale R. Watson
1. Introduction
The psychology of reading encompasses an extensive body of

literature dating back to the early 1900s [1]. This research
has attempted to outline and describe the mental operations
that readers engage in to comprehend printed language and
has, in addition, sought to determine the factors responsible
for individual differences in reading ability. Although the
subjects for these studies typically have been either elemen-
tary school children or college aged readers without anyocu-
lar pathologies, the methodologies employed in these studies
or the conclusions drawn from them may be of value to
researchers investigating the reading abilities of various low
vision populations.
Psychologists investigating reading have long assumed

that reading is a complex task that entails a sequence of
component processes. HUEY [1], for example, observed that
any complete theory of reading would necessitate descrip-
tions of "very many of the most intricate workings of the
human mind". Since reading begins with the input of visual
pa tterns and ends with an understanding of the meanings rep-
resented by those patterns, psychologists have assumed that
some sequence of mental operations must intervene to enable
the reader to derive meaning from printed symbols. As illus-
trated in Fig. 1, the models of reading adhere to this concep-
tua,lization of the reading process. For example, Fig. 1 illus-
trates the model proposed by JUST and CARPENTER [2].
Although other models are available that differ in terms of
the specifics of the descriptions, common to most models is
the conceptualization that the global task of reading is sup-
ported by the operation of a series of component sub-
processes. In the JUST and CARPENTER model [2], for
example, reading requires eye movement processes represent-
ed by the stage "get next input"; word identification process-
es, represented by the stages labeled "extraction of visual
273
features" and "word encoding"; as well as comprehension pro-

cesses represented by the stages of "integration" and the
component referred to as "sentence wrap-up".
Extract Physical
Features
Encode Word and

Access Lexicon
Ass1gn Case
Roles
Integrate with
Representat i on
of Previous Test
lves
Figure 1. A schematic diagram of the major components
involved in reading ability. From "A Theory of
Reading: From Eye Fixations to Comprehen-
sion" by M.A. Just and P.A. Carpenter, 1980,
Psychological Review 87, 331.
This framework for understanding reading has some

immediate implications for research efforts involving low
vision readers. These models suggest that various compo-
nents contribute to the overall skill at reading. Thus, reading
274
I 2 3 4 5 6 7 8 9 I 2
1566 267 400 83 267 617 767 450 450 400 616
Flywheels are one of the oldest mechanical devices known to man. Every intemal-
3 5 4 6 7 8 9 10 II 12 13
517 684 250 317 617 1116 367 467 483 450 383
combustion engine contains a small flywheel that converts the jerky motion of the pistons into the
14 15 16 17 18 19 20 21
2i4 383 317 283 533 50 366 566
smooth flow of energy that powers the drive shaft.
Figure 2. Eye fixations of a college student reading a

scientific passage. Gazes within each sentence
are sequentially numbered above the fixated
words with the durations (in msec) indicated
below the sequence number. From "A Theory of
Reading: From Eye Fixations to Comprehen-
sion" by M.A. Just and P.A. Carpenter, 1980,
Psychological Review, 87, 330.
difficulties may be a consequence of a disruption at any of a

number of stages in the reading process. Thus, to understand
low vision reading requires knowing which components of
reading are disrupted. Various investigators [3,4] have, for
example, demonstrated that those with central field loss read
at rates considerably slower than those with other kinds of
visual difficulties and have examined how stimulus variables
such as contrast, pixel density, and print size affect perform-
ance. In conjunction with research of this type, the reading
models suggest exploring the processing abilities of low vision
readers. If those with central loss read at slow rates,
researchers might investigate the extent to which difficulties
in eye movement control or word recognition abilities con-
tribute to slow performance. In addition, since the various
components of the reading process are interrelated, research-
ers might investigate the extent to which difficulties in these
visual components of the process limit higher level compre-
hension processes. Although the reading difficulties encoun-
tered by the low vision adult are clearly related to sensory
limitations, consideration of the additional components of the
reading process is suggested in order to identify the major
obstacles encountered and thus be in a position to more
effectively remediate those problematic skills.
275
This approach to understanding reading has, to some

extent, influenced the methodological approach psychologists
have used to investigate reading. Investigators have typically
focused their experimental efforts on some specific compo-
nent of the reading process to identify how that particular
process is accomplished in isolation. With this information as
a baseline, of sorts, variations that result when components
interact (presumably the situation in natural reading) can be
more easily observed and quantified. Basically, investigations
into eye movement control, word recognition abilities, and
comprehension strategies have been conducted. The focus of
this brief review will be on those studies investigating eye
movement and word recognition abilities. Since there is a
large visual component to each of these processes, various
ocular pathologies may have different effects on these vari-
ous components. For example, a significant reduction in acu-
ity may impact word recognition abilities but have only mini-
mal effects on saccade performance. On the other hand,
small asymmetrical central scotoma extending into the right
field may have a large impact on saccade control and yet not
significantly disrupt the word recognition process.
2. Studies on Eye Movement Control
The studies on eye movement control typically involve col-

lege aged subjects silently reading meaningful text displayed
on a CRT monitor. Subjects are required to answer a set of
true-false questions after reading each passage in an effort to
force subjects to read the passages for comprehension. As
subjects are reading the text, eye-movement recording
devices provide continuous on-line recordings of each sub-
ject's eye movements.
The data derived from these studies [5,6,2] indicate that

there exists a substantial amount of within-subject variability
in both fixation durations and the length of forward saccades.
For example, RAYNER and McCONKIE [6] report that the
fixation durations of a single reader range anywhere from 50
msec to over 1500 msec. The extent of forward saccades can
range from as little as two character spaces to eighteen char-
acter spaces or more. In addition, the data indicate that each
of these aspects of eye movement behavior are separately
controlled. For example, RAYNER and McCONKIE [6] ana-
276
lyzed the eye movement data of six subjects to determine if

any relationship existed between the duration of a fixation
pause and the extent of the following forward saccade. The
correlations among these subjects ranged from -0.041 to
0.108 with an overall average correlation across all subjects
of -0.006. Apparently then, eye movement control consists of
two independent decision mechanisms: one that decides when
to initiate the next saccade and a second that determines
saccade extent.
One variable responsible for controlling saccade extent is

the length of the word to be fixated [7,5]. Readers will move
their eyes approximately five character spaces rightward if
the next word is six characters in length but will move right-
ward 8.3 characters if the next word is ten letters in length.
Thus, readers typically make a longer saccade as the word to
the right of fixation increases in length. In addition, skilled
readers are more likely to program a saccade that results in
fixating the next word in the text as the next word increases
in length. For example, if a reader is fixating a four letter
word, the probability is 0.45 that the following four letter
word will be fixated and 0.92 if the following word is ten let-
ters in length. Essentially, decisions regarding whether to
make a saccade that results in "skipping over" the next word
depend, in part, upon the length of the word to the right of
fixation.
The question of how far to the right of fixation skilled

readers are able to acquire word length information for sac-
cade control was investigated by McCONKIE and RAYNER
[8]. They demonstrated that readers make use of word length
information as far as twelve character spaces, or about 3
degrees, from the center of fixation. When word length
information was unavailable to subjects beginning at 12 char-
acters from the center of fixation, the saccades of subjects
were significantly shorter than in a condition where word
length was continuously available in the periphery. Similar
results have been obtained by IKEDA and SAIDA [9].
Two implications follow from the studies briefly summa-

rized thus far. First, the studies suggest that during a fix-
ation pause, readers are engaged in two processes: identifying
the word available in foveal vision and acquiring length infor-
mation from words further into the periphery in order to plan
the subsequent saccade. In fact, the studies suggest that the
perceptual span for acquiring word length information may be
at least as large as 3 degrees from the center of fixation.
277
Secondly, the studies suggest that saccade planning is a

process that takes place fixation to fixation. Given the sig-
nificant variability in the extent of forward saccades and the
fact that local text characteristics such as the word length
have a significant impact on saccade planning, some
researchers have argued for the momentary control of eye
movements.
A recent experiment by RAYNER and POLLATSEK [10]

provides data that support this conclusion. The study
involved having subjects read text through a window that
moved in synchrony with their eye movements. In one condi-
tion, the size of the window was held constant and in a second
condition window size varied from fixation to fixation. The
results indicated that saccades increased in length as window
size increased but no differences in saccade length were
found between the constant and variable window conditions.
This result demonstrates that readers do program saccades
given the information they have available during a particular
fixation and engage in this planning on a fixation to fixation
basis.
These studies also raise some questions regarding low

vision reading. For example, readers with macular degenera-
tion are typically encouraged to make use of their remaining
vision by viewing targets eccentrically. If peripheral retinal
locations are being used to identify symbols, can these read-
ers also acquire the visual information that is available some-
what further into the periphery for planning saccadic eye
movements? Can the process of guiding eye movements
become as automatic for low vision readers as it is for read-
ers with intact central vision so that word identification and
eye movement processes can take place concurrently?
Assuming that eye movement control is a skill that, like any
other skill, develops with practice, how much practice in eye
movement control is necessary to make this component of
reading automatic? Is there any relation between the amount
of practice necessary and size of the central scotoma? And
finally, to what extent does the enlargement of text compli-
cate saccade performance? For example, HELLER and
HEINISCH [11] have demonstrated that the saccades of nor-
mally sighted subjects reading enlarged text are significantly
smaller than those that result when smaller sized text is pre-
sented. Consequently, a potential trade-off may result in low
vision reading whereby word recognition processes are facili-
tated by magnification but saccade control is either compli-
278
cated or becomes increasingly more significant as a rate-

limiting factor in reading.
3. Studies on Word Recognition Processes
The process of word recognition is another component of

reading that depends, in part, on how well the reader can uti-
lize the visual information from the printed page. In a
variety of studies, psychologists have investigated the degree
to which overall reading skill is related to this more basic
component and how this particular aspect of reading is
accomplished. These studies have obvious implications in low
vision reading research.
In the psychology of reading literature, the word recogni-

tion process is referred to as lexical access. It is assumed
that word identification occurs when the visual information
from the printed page activates representations of known
words held in memory [12]. Figure 2 presents the fixation
durations (in msec) of a college aged subject reading mean-
ingful text displayed on a CRT monitor at a rate of about 200
words per minute [2]. This subject fixated the word "fly-
wheel" for about 1500 msec and the word "are" for about 200
msec reflecting the extent of variability typically observed.
As others have pointed out [13], one observation regard-

ing this protocol that is particularly noteworthy is the fact
that most of the content words in the text are fixated. In
fact, JUST and CARPENTER [2] have estimated that approxi-
mately 80% of the context words of text are fixated by
skilled readers. This relatively high rate of fixation frequen-
cy is observed when subjects are reading relatively difficult
material such as the passage presented in Fig. 2 and also
when subjects are reading much less difficult and more pre-
dictable material such as Reader's Digest stories. McCONK-
IE and ZOLA [14] demonstrated that while fixation durations
vary as a function of predictability, the probability of fixat-
ing a word remained at about 0.90 regardless of the extent to
which the word was specified by contextual information.
One reason for the high sampling rate of skilled readers is

related to the perceptual span for word identification. Sever-
al researchers [3,4] estimate that the perceptual span for
word recognition processes is relatively small. Thus, although
279
readers can acquire the less detailed information specifying

word length over an area spanning approximately twelve
characters to the right of fixation, the area from which they
can acquire the more detailed information necessary for word
identification is estimated to be approximately four charac-
ters to the right of fixation. Skilled readers sample a good
deal of the text partly because the perceptual span for word
identification is relatively narrow [13].
In order to examine how the word recognition component

of reading is accomplished, psychologists have relied on very
simple tasks that require that subjects only recognize words
and do not also involve any of the other reading components
such as eye movement or comprehension processes. One of
the tasks used is referred to as a naming task in which sub-
jects are presented a series of unrelated words one at a time
in the center of a video monitor. SUbjects are simply
required to name the word aloud as quickly and accurately as
possible. The latency to initiate the vocalization is used as
the dependent measure.
Using this task, PERFETTI and HOGABOAM [15] com-

pared third and fifth grade readers who were either high or
low in reading comprehension. Their results demonstrated
that high ability readers were faster at naming words than
were low ability readers. In addition, the magnitude of the
differences in naming speed increased as word difficulty
increased. Low ability readers were approximately 150 msec
slower at naming high frequency words and about 400 msec
slower at naming low frequency words than high ability read-
ers. Other investigators have also found a strong relationship
between the ability to recognize words in isolation and over-
all reading ability. For example, SHANKWEILER and
LIBERMAN [16] reported that between 25% and 64% of the
variability in paragraph fluency can be accounted for by dif-
ferences in naming speed.
Researchers have taken this evidence to indicate that

overall reading ability differences result from differences in
some very basic, rate-limiting, components of the reading
process [13,17]. As the models of reading indicate, readers
are engaged in several processes more or less simultaneously.
The comprehension processes of reading (e.g., integrating the
individual word meanings to derive the meaning of the whole
sentence, making inferences, etc.) obviously require some
amount of cognitive effort on the part of the reader. How-
280
ever, the amount of attention readers can devote to the com-

prehension processes will depend upon how much of their
attention is already required for such basic activities as eye
movement control and word identification processes. If iden-
tifying words in print is relatively difficult for a reader, it
can be assumed that processing resources that would other-
wise be available for comprehension are being directed
toward word identification components. As word identifica-
tion and eye movement processes become effortless and auto-
matic [18], the reader can essentially concentrate on the
comprehension components of reading. The skilled driver is
one who is able to execute the basic components of gear
shifting, braking, and clutching with a minimum of conscious
attention. Thus, one of the factors that differentiates skilled
and less skilled readers is a difference in a basic ability at
recognizing words presented in isolation.
4. Use of Context in Word Recognition
Psychologists have typically assumed that two sources of

information can be used to recognize words. Readers have
the option of relying on the visual information from the print-
ed page or the contextual information derived from compre-
hension processes to identify words. Presumably both sources
of information contribute to word recognition performance in
the natural reading situation. This assumption was explicitly
investigated by PERFETTI and ROTH [19]. Presenting chil-
dren with words that were visually degraded by randomly
deleting dots from computer-printed words, PERFETTI and
ROTH [19] demonstrated that deficient visual input could be
compensated for by contextual information. In fact, as is
shown in Fig. 3, the study demonstrated that a reduction of
visual input by as much as 42% had no effect on recognition
speed if the words were presented in a context that made
them highly predictable.
Although reliance on context may compensate for

degraded visual input, suggesting a rehabilitation strategy for
low vision readers, some additional studies indicate that cau-
tion is required before recommending this option. The
PERFETTI and ROTH [19] study also investigated differences
between good and poor readers in terms of their reliance on
context. Comparing latencies to name the last word in a sen-
281
2000
Isolation
1900
1800
1700
1600
Low
1500
1400
Medium
> 1300
.
u
c
2l 1200
--1
1100
..
c
.g 1000
u
;;: 900
"g 800
'"
~
700
600
500
400
300
200
100
0
o 21 42
Percent Degraded
Figure 3. Word identification latencies as a function of

contextual constraint and degree of degrada-
tion. From "Some of the Interactive Processes
in Reading and Their Role in Reading Skill" by
C.A. Perfetti and S. Roth, 1981, Interactive
Processes in Reading.
tence when that item is either highly predictable, unpredicta-

ble, or anomalous, PERFETTI and ROTH [19] demonstrated
that poor readers are more affected by manipulations of con-
text than are good readers. These data are reproduced in Fig.
4. Unpredictable words take approximately 90 msec longer
than predictable words for good readers to identify but take
almost twice as long (170 msec longer) for poor readers to
identify. The differential effects of context on the perform-
ance of good and poor readers are even more evident when
the context is anomalous. Similar results have been obtained
by STANOVICH and WEST [20]. Perhaps, less skilled readers
are more dependent upon contextual cues for word recogni-
tion because they are not as efficient as skilled readers at
using the visual information available. Reliance on context is
282
1000
VI
-c
c:
.,
0
0
900
~
~
.5i
>-
0
c: 800 LESS SKILLED
~
CJ
...J
c:
:2CJ e--- -- -- _____ _
-~~
700
:Ec
.,
~
600
Anomalous (0%) Unpredictable (3%) Predictable (80%)

Context Type
Figure 4. Word identification latencies for skilled and

less skilled fourth grade readers as a function of
word predictability. From "Some of the Inter-
active Processes in Reading and Their Role in
Reading Skill l1 by C.A. Perfetti and S. Roth,
1981, Interactive Processes in Reading.
a strategy used by less skilled readers to compensate for their

less than efficient use of the available visual information.
283
5. Assessment of Reading Ability Among Low Vision

Readers
One of the obvious conclusions to draw from the research

briefly summarized is that a complete assessment procedure
would require an entire battery of sub-tests which would be
administered to low vision subjects while, in addition, sophis-
ticated eye tracking equipment recorded fixation durations,
forward sac cades, and regressive eye movements. Equally
obvious is the observation that any practical assessment tool
would necessarily involve some compromises.
Of the assessment procedures presently available, some

require that subjects read meaningful text (e.g. Sloan Contin-
uous Text Reading Cards, the scrolling text procedure
employed by LEGGE et al. [3]) whereas some require only
that subjects identify unrelated letters and words (e.g.,
Bailey-Lovie Word Reading Chart [21]; Pepper Visual Skills
for Reading Test, see BALDASARE et al. [22]. The choice
between these two options necessarily results in some trade-
offs. While the procedures employing meaningful text more
closely approximate natural reading, the introduction of con-
textual cues that can compensate for visual deficiencies [19]
may lead to an underestimation of the visual requirements for
reading. On the other hand, the use of unrelated words, while
providing a more accurate assessment of visual recognition
skills, is a task one step removed from the natural reading
situation.
On a different dimension, the Pepper VSRT does rather

closely sample the different visual component processes
required in natural reading. As an assessment of performance
with stationary text, the VSRT requires that subjects engage
in both saccadic and return sweep eye movements. In fact,
the test has been designed to make these movements increas-
ingly more difficult as successive lines are encountered.
Thus, a subject's performance on the VSRT necessarily
reflects the contribution of eye movement processes. In
addition, the test requires the identification of unrelated
words, a task that previous research [17] indicates is strongly
predictive of overall reading ability. Other assessment
instruments either have not been designed to systematically
manipulate eye movements processes (e.g. Sloan Reading
Text Cards, Bailey-Lovie Word Reading Cards) or may intro-
duce eye movements that are atypical of the natural reading
situation (e.g., scrolling text procedure).
284
In terms of experimental control, the ability to externally

vary reading rate in the scrolling text procedure enables an
elegant and precise determination of maximum achievable
reading rates. Inspection of the data derived with this proce-
dure [3] makes it clear that an assessment of performance
limits is possible when rate is externally varied. The rate of
reading the VSRT is, however, controlled by each subject.
Although subjects are instructed to read as quickly and accu-
rately as possible, no determination of performance limits
can be made with this procedure.
Finally, the scrolling text procedure is particularly well

suited for investigating how various visual characteristics of
a display affect reading performance, whereas the Pepper
VSRT is better suited to assess ability differences in some of
the processing components of reading (e.g., eye movement
control and word recognition).
In summary, the research on the reading abilities of the

normally sighted has provided information specifying how
various components of the reading process are accomplished.
This information can provide researchers in low vision with
some additional insights regarding reading that may contrib-
ute to their research endeavors.
6. Acknowledgements
The authors gratefully acknowledge support from the Nation-

al Eye Institute (ROl 6420) and the National Institute of
Handicapped Research (G008435033). In addition, we would
like to thank Stephen Whittaker for his valuable comments in
the preparation of this manuscript.
285
7. References
1. E.B. Huey: The Psychology and Pedagogy of Reading

(MIT Press, Cambridge 1968)
2. M.A. Just, P.A. Carpenter: A theory of reading: From

eye fixations to comprehension. Psychol. Rev. 87, 329
(1980)

Psychophysics of reading n. Low vision. Vision Res.
25, 253 (1985)
4. K. Rayner, J.H. Bertera: Reading without a fovea.

Science 206, 468 (1979)
5. K. Rayner: Eye movements in reading and information

processing. Psychol. Bull. 85, 618 (1978)
6. K. Rayner, G.W. McConkie: What guides a reader's

eye movements? Vision Res. 16, 829 (1976)
7. J .K. O'Regan: Structural and Contextual Constraints

on Eye Movements in Reading. (Ph.D. University of
Cambridge, 1976)
8. G.W. McConkie, K. Rayner: The span of the effective

stimulus during a fixation in reading. Percept. Psycho-
phys. 17, 578 (1975)
9. M. Ikeda, S. Saida: Span of recognition in reading.

Vision Res. 18, 83 (1978)
10. K. Rayner, A. Pollatsek: Eye movement control during

reading: Evidence for direct control. Q. J. Exp. Psy-
chol. 33A, 351 (1981)
11. D. Heller, A. Heinisch: Eye movement parameters in

reading: Effects of letter size and letter spacing. In
Eye Movements and Human Information Processing,
ed. by R. Groner, G. McConkie, C. Menz (Elsevier,
New York 1985)
12. M. Coltheart, E. Davelaar, J.T. Jonasson, D. Besner:

Access to the internal lexicon. In Attention and Per-
286
romance VI, Proceedings of the Sixth International

Symposium on Attention and Performance, Stockholm,
1975, ed. by S. Dornic (Erlbaum, Hillsdale 1977) pp.
535-555
13. C.A. Perfetti: Reading Ability (Oxford University

Press, New York 1985)
14. G.W. McConkie, D. Zola: Language contraints and the

functional stimulus in reading. In Interactive Process-
es in Reading, ed. by A.M. Lesgold, C.A. Perfetti
(Erlbaum, Hillsdale 1981)
15. C.A. Perfetti, T. Hogaboam: Relationship between

single word decoding and reading comprehension skill.
J. Educ. Psychol. 67, 461 (1975)
16. D. Shankweiler, I.Y. Liberman: Misreading: A search

for causes. In Language by Ear and by Eye, ed. by J.F.
Kavanagh, I.G. Mattingly (MIT Press, Cambridge 1972)
pp.293-317
17. K.E. Stanovich: Individual differences in the cognitive

processes of reading. I. Word decoding. J. Learn.
Disabil.15, 485 (1982)
18. D. LaBerge, S.J. Samuels: Toward a theory of auto-

matic information processing in reading. Cog. psy-
chol. 6, 293 (1974)
19. C.A. Perfetti, S.F. Roth: Some of the interactive pro-

cesses in reading and their role in reading skill. In
Interactive Processes in Reading, ed. by A.M. Lesgold,
C.A. Perfetti (Erlbaum, Hillsdale 1981)
20. K.E. Stanovich, R.F. West: The effect of sentence

context on ongoing word recognition: Tests of a two-
process theory. J. Exp. Psychol. (Hum. Percept.) 7,
658 (1981)
near-vision chart. Am. J. Optom. Physiol. Opt. 57, 378
(1980)
22. J. Baldasare, G.R. Watson, S.G. Whittaker, H. Miller-

Shaffer: The development and evaluation of a reading
287
test for low vision macular loss patients. J. Visual

Contrast Polarity Effects in Low Vision Reading
Gordon E. Legge, Gary S. Rubin, Mary M. Schleske
1. Introduction
Nearly all books, magazines and newspapers are printed with

black letters on a white background. As Fig. 1 illustrates, it
is possible to display text as white letters on a black back-
ground. We refer to these alternatives - black-on-white or
white-on-black - as the contrast polarity of the text.
Although the choice of contrast polarity has little effect on
the reading performance of normally sighted people, it is
important for some individuals with low vision.
man y myths
tall tales
Figure 1. Text may be printed in either of two contrast
polarities, black on white or white on black.
We have examined the significance of contrast polarity in

the course of a series of psychophysical studies of low vision
reading [1-5]. In this paper, we will begin by describing our
general method for studying reading. Then we will show that
289
our understanding of low VISIon reading is enhanced if we

classify subjects according to the integrity of their central
fields and the clarity of their ocular media. Next, we will
present evidence that subjects with cloudy ocular media
exhibit a contrast polarity effect; they read white-on-black
text faster than black-on-white text. Finally, we will consid-
er abnormal light scatter within the eye as an explanation of
this effect.
2. Method
In our experiments, subjects read aloud text that appears on

the screen of a TV monitor. The text is stored in computer
memory and is displayed upon command with the aid of a vid-
eo frame buffer. In a reading trial, a single line of text drifts
smoothly from right to left across the screen, like the moving
overlays sometimes used on television to give weather infor-
mation. At any given instant, ten characters are visible on
the screen. Drift speed is controlled by the computer at the
direction of the experimenter. The subject reads the line of
text aloud as it drifts by. If no errors are made, the experi-
menter increases the speed on the next trial. This procedure
continues until a speed is reached at which the subject begins
to make errors. At this point, we compute reading rate in
words/minute, having corrected for the proportion of errors.
We have shown that this measure of reading rate is robust and
provides a good estimate of a subject's maximum reading
rate.
3. Effects of Character Size and Contrast Polarity
How does reading speed depend on character size? This ques-

tion has received surprisingly little attention. In connection
with low vision, character size is critical because it deter-
mines appropriate magnification. We measured reading rate
across a wide range of angular character sizes for both nor-
mal and low vision observers. We varied angular character
size by changing the subject's viewing distance, taking care to
refract where necessary. Figure 2 shows reading data for
normal subjects. Reading rate in words/minute is plotted as a
290
function of angular character size, both on logarithmic

scales. Character size covers a 400 to 1 range from 0.06
degrees to 24 degrees. Letters of 0.06 degrees (or 4 minutes
of arc) are smaller than the 6/6 Snellen letters. Twenty four
degree letters are enormous, spanning about 6 inches at a
reading distance of 40 cm (16 inches). Data of three subjects
are represented by symbols of different shape. The individual
differences are very slight. Data for the two contrast polari-
ties are shown by open and closed symbols. There is no sys-
tematic effect of this variable for normal subjects. This
finding was confirmed in a separate study using black letters
on colored backgrounds and colored letters on black back-
grounds.
,....
-Q)
:::J
c:
E
....
en
300
"0
....
0
~
w 100
l-
e:(
a:
(!) C. OK
z 30 O. KS
o o OGP
e:(
w 10-charllctar window
a:
10
.03 .1 .3 3 10 30
CHARACTER SIZE (degrees)
Figure 2. Reading rate is plotted as a function of.angular
character size for three subjects with normal
vision. Open and filled symbols show data for
the two contrast polarities.
291
Notice that the curve in Fig. 2 has a broad peak at about

250 words/minute, extending from about 0.4 degrees to 2
degrees. This is the range of character sizes for which nor-
mal reading is fastest. To the left of the peak, there is a rap-
id decline as character size approaches the acuity limit. To
the right of the peak, there is a more gradual decline for
large letters. However, reading rate is still about 70 words/
minute for 24 degree letters. This finding is encouraging for
low vision because it indicates that it is possible to read high-
ly magnified text quite rapidly.
We conducted similar measurements on a sample of low

vision subjects. Their results can be best understood if we
keep in mind two major distinctions - residual central vision
vs. central-field loss and clear vs. cloudy ocular media.
Figure 3 shows data for a subject with optic nerve atro-

phy. Once again, reading rate is plotted against character
size. The upper solid curve represents average results from
normal subjects (Fig. 2). This low-vision subject has clear
ocular media. Typical of such subjects, there is little or no
effect of contrast polarity on reading as indicated by the near
conjunction of open and closed symbols. Typical of subjects
with residual central vision, the curve has a peak, that is,
there is a limited range of character sizes for which reading
is optimal. Here, it is between 3 degrees and 6 degrees.
Since this is larger than character sizes encountered in every-
day reading, this subject certainly requires magnification.
But notice that excessive magnification will result in reduced
reading speed. Also note that this subject's peak reading
speed is about 100 words/minute, potentially of great benefit
to him.
Figure 4 shows a different pattern of results, but one typ-

ical of subjects with central field loss. This subject suffers
from ocular histoplasmosis. His media are clear. Once again,
there is no systematic difference for the two contrast polari-
ties. Characteristic of subjects with central loss, the data do
not show a well-defined peak. They rise steadily to the larg-
est character size. His best reading speed is only 50 words/
minute, a factor of two slower than the subject in Fig. 3.
Figure 5 shows data for a subject with severe corneal

vascularization. His fields are intact, but unlike the previous
subjects, his media are very cloudy. Typical of such subjects,
he. shows a contrast-polarity effect, that is, he reads better
292
r...
OJ 300 Normal Vision
+J
:J
.....C
E 100
-......
(f)
lJ
L
o 30
~
'-"
W
I-- 10
<
0:::
1.J
Z 3 Observer 0
....... • Black-on-Whitg
o o Whitg-on-Black
<
W
0:::
. I .3 3 10 30

Figure 3. Reading rate is plotted as a function of charac-
ter size for a low vision subject with residual
central vision and clear ocular media.
for white letters on a black background (open symbols) than

for conventional black-on-white (filled symbols).
In the clinic, it is impractical to measure entire reading-

rate curves like those shown in Figs. 2-5. However, we have
found that commonly available clinical data can already pro-
vide some of the salient information. For example, it would
be useful to know the peak reading rate that any given low
vision subject might expect to attain. Figure 6 shows pre-
dictions of peak reading rate from our data, based on a
multiple-regression analysis using just two clinically available
variables - presence/absence of central vision and clear/
cloudy ocular media. The numbers in the cells give the pre-
dicted peak reading rate for the four types of subjects. It is
293
,-...
OJ 300 Normal Vision
+J
:J
C
......
E 100
"-(J)
-0
L
0 30
3:
'-J
W
~ 10
c:(
0:::
~
Z 3 Observer 0
I-i
• Black-on-Whit&
0 o White-on-Black
c:(
W
0:::
. I .3 3 10 30

ter size for a low vision subject with central-
field loss and clear ocular media.
evident that the presence or absence of central vision is criti-

cal while the state of the ocular media plays an important but
lesser role. This simple four-way prediction amounts for 64%
of the variance in our measurements of peak reading rate.
In the clinic, it is important to determine which low

vision subjects read better with white-on-black text since
only relatively expensive electronic aids are capable of con-
trast reversal. We have discovered that the presence of clou-
dy media is a very good predictor. We computed the ratios of
peak reading rates for white-on-black and black-on-white
texts for each member of a sample of low vision subjects.
Figure 7 shows separate histograms for subjects with clear
294
""'OJ 300 Normal VisiOn

+J
:J
C
.....
E 100
""-(f) I
I
I
LJ I
L I
o 30
I
¢
~
'-./ I
I
I
W
I- 10 ~
I
< c1
0:::
L')
Z 3 Observer J
~
• Black-an-White
o o White-an-Black
<
W
0:::
•1 .3 3 10 30

ter size for a low vision subject with residual
central vision but cloudy ocular media.
and cloudy media. A ratio of 1.0 along the horizontal axis

means that the peak reading rates were the same for the two
contrast polarities. Open bars show results for subjects with
clear media. There was a maximum deviation of 11% from a
ratio of 1.0, and most of these subjects were within 5%. This
means that subjects with clear media exhibit little or no con-
trast polarity effect. The shaded bars show results for sub-
jects with cloudy media. All of these subjects had ratios
greater than 1.0 indicating faster reading for white-on-black
text. Improvements ranged from 10% to 52%.
We found a second way of predicting contrast polarity

effects. We measured reading acuity in both polarities, once
with regular Sloan M cards and once with contrast reversed
295
PREDICTIONS OF PEAK READING RATES

(words/minute)
OCULAR MEDIA
Clear Cloudy
Intact 131 95
CENTRAL
FIELD
Loss 39 29
Figure 6. Estimates of peak reading speed are shown for

four categories of low vision subjects.
(white-on-black) M cards. The difference in the two acuities

was predictive of a contrast polarity effect in reading. In
statistical terms, the pair of M acuities and the designation
of a subject's media as clear or cloudy together accounted for
70% of the variance in the ratio of reading speeds.
296
10.-,----.---,----.----.---.----.-.
en 8 D CLEAR OCULAR MEDIA

a:
W _ CLOUDY OCULAR MEDIA
>
a:
w 6
en
al
0
U.
0
4
a:
w
al
::E
:::::>
z 2
0'---'----
.95 1.05 1. 15 1.25 1.35 1.45 1.55
RATIO OF PEAK READING RATES
Figure 7. These histograms give the number of low vision

subjects with the indicated ratios for white-on-
black to black-on-white reading speed. A ratio
of 1.0 means that the peak reading rates were
the same for the two contrast polarities. Sepa-
rate histograms are shown for subjects with
clear and cloudy ocular media.
297
4. Explaining Contrast Polarity Effects
Why do subjects with cloudy ocular media manifest contrast-

polarity effects in reading?
An eye with cloudy media is subject to abnormal light

scatter. The simplest model of this process has two ele-
ments. First, we assume that a fixed fraction of the incident
light is diverted from image formation into scattered light.
Second, we assume that the scattered light from a point
source is not spread uniformly across the retina, but is depos-
ited as a uniform veiling luminance within a restricted retinal
area. A more refined model would provide an optical point-
spread function representing details of the spatial distribution
of scatter. The veiling luminance will act as a whitewash to
reduce stimulus contrast.
Why should light scatter cause a contrast polarity effect

in reading? Why is there a difference between white letters
on a black background and black letters on a white back-
ground? The answer is that there is an asymmetry in the
amount of light scattered. We used a photometric technique
to estimate that when a white page contains single-spaced
lines of black print, only about 16% of the page is actually
dark. This is illustrated by the black square on the white
page in Fig. 8; it covers 16% of the area. The remaining 84%
is high reflectance white and a source of scattered light.
When contrast is reversed, only the 16% of the page occupied
by the print is white. There is subtantially less light available
for scatter, as indicated by the small white square on the
black page in Fig. 8.
We now consider some quantitative implications of our

light-scattering model. The model's performance is compared
with extensive psychophysical measurements on the right eye
of subject GEL, one of the authors. He has cloudy media due
to vascularization but no known retinal or other neural
involvement. His acuity is 6/300 (20/1000). Such cases of
pure but severe ocular clouding are rare.
We estimated the proportion of light diverted from image

formation into scatter in GEL's right eye. This was done by
comparing his contrast thresholds for the recognition of 6
degree letters with corresponding thresholds for normal
observers. Contrast threshold was elevated by factors of 16.9
for black letters on a white background and 12.1 for white
298
Figure 8. The black square covers about 16% of the

white page and represents the proportion of a
page with single-spaced lines of black print that
is actually dark. The white square on the black
page illustrates the proportions when contrast is
reversed. Clearly, there is more light available
for scatter in the black-on-white case.
letters on a black background. Once again, this difference

can be traced to differences in the white and black content of
the images. Using our simple model, these ratios translate
into a scattering factor of 92.4%, that is, over 90% of the
incident light is diverted into veilng luminance that attenu-
ates retinal-image contrast.
In Fig. 9, "fortune" is printed with approximately 90%

contrast, "working" with 30%, "tearful" with 10%, and "visi-
ble" with 3% contrast. Imagine reading texts printed with
these different contrasts. SUbject GEL's cornea attenuates
contrast by more than a factor of 10 for 6 degree letters.
Therefore, even when stimulus contrast is 100%, his retinal
299
image contrast lies somewhere between "tearful" and "visi-

ble". Curiously, his subjective experience is not that of a low
contrast world. To him, 6 degree characters appear black and
of high contrast even though we can be sure that his retinal
images are very washed out. This is really not so strange.
Letters on or near the 6/6 (20120) line of a Snellen chart
appear black to the normal eye despite the fact that retinal
image contrast is reduced by diffraction and optical aberra-
tions. Apparently, our perceptual systems convey informa-
tion about stimuli as they exist in the world, not their corre-
sponding retinal images.
Figure 9. Examples are shown of words printed at con-

trast levels of approximately 90%, 30%, 10%
and 3%.
300
What effect should this contrast reduction have on GEL's

reading performance? If his retinal function is intact, his
performance should match that of normal subjects when their
text is reduced in contrast by a suitable amount. Therefore,
we need to know how normal reading speed depends on con-
trast. Figure 10 shows reading rates as a function of contrast
for three normal subjects. Reading rate rises with contrast,
but the curve flattens out at the high end. This means that
subjects with normal vision are relatively tolerant to reduc-
tions from maximum contrast. In Fig. 11, the upper curve
shows average results for normal subjects (from Fig. 10). The
diamond symbols located at the right of the graph show how
subject GEL's reading rate depends on contrast. Notice the
pronounced contrast polarity effect; open symbols represent-
ing white-on-black reading consistently lie above filled sym-
bols at all contrast levels. Notice also that GEL's curves are
steeper and lie substantially below the normal curve. The
steepness reveals that any reduction from maximum contrast
causes substantially reduced reading speed. Can we account
for the way GEL's data deviate from normal by using the sim-
ple scattering model? Using the scattered fraction of 92.4%
derived from threshold measurements, we shifted GEL's data
points to lie at the retinal contrasts deduced from the model.
The results are shown by the circle symbols located at the
left in Fig. 11. Data for the two contrast polarities are shift-
ed by different amounts owing to the different amount of
light scattered by print of the two polarities. Following these
differential shifts, the two sets of data overlap. The model
therefore accounts for the contrast polarity effect. Some of
the shifted data points lie on or near the normal curve, but at
the lowest contrasts they drop below it. To the extent that
the data lie on the solid curve, the simple scattering model
also accounts for the effects of contrast on GEL's reading.
Light scatter is often implicated in glare effects. In our

reading measurements, we kept the visual field above and
below the line of text dark to eliminate glare. In everyday
reading, however, glare light may arise from the white page
above and below the line being read. The glare may be treat-
ed as a veiling luminance cast over the line of text by the
glare source. In our laboratory, we have been studying glare
in low vision. Figure 12 illustrates the target we use. The
subject's task is to detect a dark letter on a small, circular
white field. The contrast of the letter is reduced until the
threshold is reached. An annular glare ring surrounds the tar-
get. From the measurement, we can infer the veiling lumi-
301
".... 300
QI
+'
:J
....C 100
E
........
(f)
"U
L
a 30
~
'oJ
W
I- 10
OBSERVERS
<
0::: • KD
-
L:l • 14K
Z 3 • DR
o 6° Letter.
< Block-an-White Text
W
0:::
.03 •1 .3
CONTRAST
Figure 10. Reading rate is plotted as a function of con-
trast for three subjects with normal vision.
nance associated with the glare ring. Figure 13 shows glare

data for GEL's right eye. Veiling luminance is plotted against
the inner diameter of the glare ring. The numerical values on
the Y-axis depend on the intensity and area of the glare ring.
Suffice it to say that the glare effect is substantial and
extends over many degrees, but with diminishing effect away
from the glare source.
We used these data to estimate how glare affects GEL's

reading performance. We measured his reading speed for
lines of black text that drifted through white fields {windows}
of varying height. {See Fig. 14 for samples of three such win-
dows.} The narrowest fields were only 6 degrees high, just
high enough to encompass the letters. As the window height
increased, more and more light was available above and below
302
r"\ 300
...,OJ Normal Vision
:J
....C 100
E
"en
"0
L
o 30
...,~
~ 10
<
0::::
-
OBSERVER GEL OBSERVER GEL
l:I TRANSFORMED DATA 6 0 LETTERS
Z 3
o • Black-an-White • Black-an-White
< o White-an-Black o White-an-Black
W
0::::
.03 •1 .3
CONTRAST
Figure 11. The diamonds show reading rate as a function
of contrast for observer GEL. He has severely
clouded media. The circles show the same
data after they have been shifted according to
the light-scattering model to X-axis values
corresponding to GEL's retinal image contrast.
the text for glare. The results are plotted in Fig. 15. Evi-
dently, GEL's reading speed declines substantially as the win-
dow height increases. The solid curve was derived as a pre-
diction from the glare data in Fig. 13. The fit is good.
303
GLARE
DISTANCE
::::::::::::::::::::::::::::::;: ,:::~~........i~::::'~:::::::::::::::::::::::::::::::
............................ •....••.........••....•.. .....•..•....•.••.••.••. .,,;;,..........................
~
.. ................•....:::::
::::::::~:.:~:~:~:~:~:.:.:.. ~.:.:.:.:.:.:.:.:.:::::::::::::::::::::: ~.~.~................
........ BACKGROUND :.::::::::::::::::::::
:::::::: ........... a.... GLARE SOURCE rI';;- ~:::
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::=::
.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:
:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:...............................................•............•••..••...
Figure 12. In the glare experiment, subjects attempt to
identify a letter on a small circular white
field. A bright, annular glare source surrounds
the target.
304
OBSERVER GEL
80 6° LETTERS
w 60
u
z
-
z<
~
::J
..J
40
l..:)
Z
-
-..J
w
>
20
5 10 15 20 25
GLARE RING INNER DIAMETER (deg)

Figure 13. Inferred values of veiling luminance are plot-
ted as a function of the inner diameter of the
annular glare ring for an experiment involving
subject GEL.
305
Figure 14. Examples are shown in which the window

height is varied. The greater the window
height, the greater the problem posed by glare
for reading.
306
100 "'---'-~--r--..,.....--r-----'r--~-...,..-""T"""---'
""'OJ
+' • OBSERVER GEL
::> ----SCATTERING MODEL
....
C 80
E eo Let.t.er •
.........
(J)
1:)
• Black-on-Whlt.e
L 60
o
~
'-J
~ 40
<
a=
-
t::I
Z
o
<
w
a:::
20
•
WINDOW HEIGHT (deg)

Figure 15. Reading rate is plotted as a function of win-
dow height for subject GEL.
5. Conclusions
1. Contrast polarity does not affect reading speed in nor-

mal vision.
2. In low V1Slon, contrast polarity effects are confined

principally to subjects with cloudy ocular media.
3. A simple model based on light scatter can account for

contrast polarity effects and, in part, for the dependence of
reading speed on contrast.
307
4. Light scatter can also help us to understand the

effects of glare on low vision reading performance.
6. Acknowledgements
Supported by U.S. Public Health Service Grant EY02934.
7. References

25, 239 (1985)

Psychophysics of reading II. Low vision. Vision Res.
25, 253 (1985)

4. G.E. Legge, G.S. Rubin: Psychophysics of reading IV.

Wavelength effects in normal and low vision. J. Opt.
Soc. Am. (A) 3, 40 (1986)
5. G.E. Legge, G.S. Rubin: Psychophysics of reading V.

The role of contrast in normal vision. Vision Res. (in
press)
Effect of Magnification and Field of View
on Reading Speed Using a CCTV
J .E. Lovie-Kitchin, George C. Woo
1. Introduction
Persons with untreatable visual disorders whose vision cannot

be corrected by routine refraction require either optical or
electronic aids to magnify the images of objects of interest.
Spectacles, simple magnifiers, telescopes and Closed Circuit
Television Systems (CCTVs) can all be used as low vision aids
to assist persons with reading. CCTVs generally come in the
form of a television camera connected to a video display
monitor. Optical low vision reading aids are available in dif-
ferent powers, but as the eye-to-Iens distance increases, the
angular field of view and the magnification of simple plus
lenses decreases [1]. Clinicians working in low vision have
often observed that reading speed of low vision patients is
reduced when the high magnification aids are used away from
the spectacle plane, due to the limited number of words seen
at one time. The usual clinical rule of thumb for prescribing
optical aids is, therefore, to prescribe the lowest power which
will enable the patient to read the required print size. It is
hypothesized that this provides the widest possible field of
view through the aid, when the magnifying system is placed
as close as possible to the eye.
While many patients with low vision can read with both
optical aids and CCTVs, the CCTV tends to be used by per-
sons with severe visual impairments who require higher mag-
nification than is readily available with optical aids. CCTVs
also have the added advantage of wider fields of view
(depending on monitor size and the working distance adopted
by the user), and freedom from the viewing distance restric-
tions inherent in high magnification optical systems. MEHR
et al.[2], GOODRICH et al.[3,4], and ZABEL et al.[5] in their
studies of CCTV users have identified additional advantages
that patients are able to use CCTVs for longer durations than
optical aids and CCTVs are flexible enough to compensate for
changes in patients' visual conditions or task requirements.
309
With the facility for interfacing CCTVs with microcomputers,

more persons with low vision are using CCTVs. Different
sized monitors are available for microcomputers and CCTVs,
allowing different field sizes. For use by persons with differ-
ent degrees of vision loss or on tasks demanding different
fields of view, the print size on CCTVs can be varied by
means of the zoom camera lens, or if connected to a comput-
er, by a large print display processor. The magnification used
is essentially left to the choice of the patient, although the
clinician usually advises the patient to maximize the number
of words on the screen by using the minimum magnification
needed to perform the reading task.
Researchers have examined separately the effects of the

number of lines (vertical field) [6], number of characters (hor-
izontal field) [7,8,9], character spacing [7], window size (hori-
zontal field) [6], character size [8,9], defocus and contrast
polarity [8,9] on reading performance with VnTs or CCTVs,
but the effect of overall field size combined with magnifica-
tion has not been previously investigated. In this study, we
conducted experiments to measure the reading speed of 18
persons with normal vision and 10 persons with low vision, at
5 different magnifications and 4 field sizes of a CCTV with a
view to examine the compromise between magnification and
field size and the validity of the usual clinical advice given to
CCTV users.
2. Methods
We measured the reading speed of subjects who read text dis-

played on a CCTV. A VTek Voyager XL Closed Circuit Tele-
vision camera was connected to an Electrohome VI9-143-60
58 cm (23 inch) screen with a P31 (green) phosphor. The larg-
er monitor was used to allow a larger range of magnifications
and field sizes than was possible with the usual Voyager moni-
tor. Text could be displayed in positive or negative contrast
(black print on a green background or green letters on a black
background). Previous studies [2,9] and the clinical experi-
ence of the authors have indicated that when low vision
patients do have a preference, they prefer negative contrast.
Therefore, green print on a black background was used
throughout the experiments. The print/background contrast
was 93% with the characters having an average luminance of
310
410 cd/m2. The luminance of the black matt surround was

less than the screen background. Subjects were allowed to
adopt a normal posture at the test distance of 400 mm for
which all subjects were optically corrected. Periodic checks
were made on the test distance. Six random paragraphs of
text taken from Munro [10] were typed, using a 12 pitch Pres-
tige Elite typeface, into passages containing lines of between
21 and 30 words (lower case letters were 1.65 mm in size).
This was done to enable as many words as possible to be pre-
sented in one sweep of the text across the screen.
Eighteen subjects with normal vision were selected from

staff and students of the School of Optometry, University of
Waterloo. Their ages ranged from 21 years to 70 years with a
mean of 39 years and visual acuities were 6/6 {20/20} or bet-
ter. Ten subjects with low vision, aged between 24 years and
73 years {mean age 46 years} were selected from the School
of Optometry Low Vision Clinic records. The only selection
criteria were that subjects had had a low vision assessment at
the clinic within the previous twelve months and that they
lived within reasonable distance of the University. All the
low vision subjects except one were inexperienced CCTV
users. Distance visual acuities measured with the Bailey-
Lovie distance logMAR chart ranged from 6/15 {20/50} to
6/300 {20/1000}. The subjects with normal vision read binocu-
larly with natural pupils, as did those with low vision. Obvi-
ously, when there were large differences between the acuities
of the two eyes, the low vision subjects were effectively
reading monocularly. Table 1 gives the relevant characteris-
tics of the low vision subjects.
Linear magnification {print size} on the CCTV screen was

varied by means of the zoom lens of the camera, between 5X
and 32X the original print size {i.e. between 8.25 mm and 52.8
mm}. The range of magnification achievable with the Voyag-
er camera and Electrohome monitor was 7X to 54X. It was
decided that a magnification of less than 7X was desirable, as
many low vision patients use near optical aids giving less than
7X. Thus photoreductions of the original typed paragraphs
were made and the print size on the screen adjusted by means
of the zoom lens to a size 5X that of the original typeface.
The highest magnification used {32X} allowed only one line of
print to be visible on the screen with the smallest field size.
The magnifications used covered the range typically adopted
by CCTV users [5,11].
311
Table 1
Selected characteristics of the low vision subjects.
Patient Age Acuity Cause of low vision

No. ( years)
55 6/15 (20/50) Macular degeneration

2 56 6160+ 2 (20/200+ 2 ) Cataract
3 34 6160 (20/200) Optic atrophy (multiple sclerosis)
4 33 6138 (20/125) Strabismic amblyopIa
5 73 6/15 (20/50) Age-related maculopathy
6 57 6/24+ 2 (20/80+ 2 ) Diabetic retinopathy
7 24 6175- 1 (20/250- 1 ) Congenital optic atrophy with
nystagmus
8 42 6/150-2 (20/500- 2 ) Fundus f I av Imacul atus
9 50 6175+ 1 (20/250+ 1 ) Albinism wIth nystagmus
10 37 6/300+ 1 (20/1000+ 1 ) Corneal opacl f Icatlon
Four different field sizes, 40 degrees, 25 degrees, 16

degrees and 10 degrees were used. The largest field size was
determined by the vertical dimension of the monitor and a
progression of field sizes smaller than this was used with the
smallest of 10 degrees considered to be representative of the
small angular field sizes which patients may use with low
vision reading aids. The field sizes and magnifications used
are given in Table 2, together with the horizontal number of
characters on the screen for each condition.
The reading speeds of 18 subjects with normal vision and

10 with low vision were measured for each of the 20 condi-
tions (5 magnifications X 4 field sizes) on the CCTV. The 20
conditions were run in a random block design, in which the
different field sizes were used in random order and all magni-
fications were run (in random order) at each field size before
progressing to the next size. The subjects controlled the
movement of the X-Y platform which held the reading
material. They were required only to move the platform in
one direction (towards them) to move the text across the
screen of the CCTV. SUbjects were requested to read the
text aloud. They were instructed to adopt a reading speed
312
Table 2
Horizontal number of characters on the CCTV screen for

each condition.
.. -
~field 336 X 336 188 X 188 114 X 114 71 X 71

"Size. _ (40 0 ) (25 0 ) (160) (10 0 )
Magn I f Icatl on deg)
No. of Characters
5X 33 18 12
8X 20 11 7
12X 14 5
lOX 9 5 2
32X 2
which ensured comprehension. Comprehension was not tested

as our preliminary trials and the work of others [6,12] had
indicated that comprehension did not change when other vari-
ables were altered, as subjects simply slowed their reading
speeds to maintain comprehension. Similarly, the reading
speeds were not corrected for errors as we found that with
contextual cues within a line, subjects automatically correct-
ed their errors, thus slowing their speed. Subjects were given
practice with moving the X-Y platform and with reading the
text from the screen before commencing the trials.
Each reading trial involved subjects reading one line of

text of 21 words or more. Reading times were measured (to
within 0.1 seconds) by the same investigator throughout. The
trial began with the first word (or part-word for the small
fields) visible on the screen. The subject was told the first
word and, at a given signal, the subject commenced reading
and timing began. The time taken to read to the last full
word seen on the screen was recorded. This was repeated
twice so that the final reading speed (words/minute) is an
average of 3 trials for each condition. Reading times for one
313
line of text ranged from 7.2 seconds (182 words/minute) for a

normal vision subject to 130 seconds (21.7 words/minute) for
a low vision subject. The range of reading speeds found in our
study was similar to that found in studies by GOODRICH et
al.[4], MUTER et al.[12] and LEGGE et al.[8,9]. Within any
one line of text, there was context. Alternating lines of text
were presented in reverse order, i.e. the last line of a para-
graph was presented for the first trial, followed by the third
last line and so on. The purpose of doing this is to avoid con-
text familiarity. The subject's reading was recorded on
casette tape in order to retrieve any times missed.
The experiment took approximately one hour for the nor-

mal vision subjects and between one hour and one and three-
quarter hours for the low vision subjects. Three of the low
vision subjects were unable to read the text at the smaller
magnifications and one subject (with mUltiple sclerosis) tired
quickly so the number of trials at some conditions was
reduced.
For the 18 subjects with normal vision, an optimal cubic

function (at the 5% level) was fitted to the data. Figures IA
and 2A are sections taken through this polynomial surface.
There is a significant difference (F test, p < 0.001) between
the results for the two groups of subjects.
3.1 Field size
The effect on reading speed of varying field size is illustrated

in Fig. 1. As expected, the results for subjects with normal
vision show that reading speed increases with increasing field
size for all magnifications, with some flattening at the larger
field sizes for the lowest magnification. As Figs. IB - IF
indicate, there is a wide variability of results for the individ-
ual low vision subjects at each magnificaiton. Seven of the
ten subjects were able to read the print at the 5X magnifica-
tion (Fig. IB). Three of these subjects (2, 6 and 5) showed rel-
atively good reading speeds and like the normal subjects,
their reading speeds improved with increasing field size. The
other 4 subjects (4, 1, 7, 3) were quite slow reading at this
print size and varying the field size had little effect on read-
314
ing speed. However, as magnification is increased most of

the slow readers improve their reading speeds, while the fast-
er low vision subjects {2, 6 and 5} slow down. The overall
trend for an increase in reading speed with increasing field
size is apparent for the low vision subjects. The inter-subject
variability decreases at the higher magnifications and at 32X
magnification {Fig. IF}, the results are closer to those for the
normal group. Eight of the ten low vision subjects achieved
their fastest reading speeds at the largest field size. It would
appear that, in general, it is important for low vision subjects
to have as wide a field as possible in reading.
3.2 Magnification {print size}
The effect on reading speed of varying magnification is illus-

trated in Fig. 2. For subjects with normal vision {Fig. 2A}, as
expected, reading speed decreases with increasing magnifica-
tion. This agrees with the findings of LEGGE et aZ.[S] over
the range of print sizes used in our study. There is considera-
ble variability in the results of the low vision subjects.
The results of the low vision subjects at each field size

{Figs. 2B - 2E} show the three subjects {2, 6 and 5} who have
similar responses to the normal subjects - that is, decreased
reading speeds with increased magnification. Most of the
other low vision subjects show an initial increase in reading
speed with increases in magnification, to a maximum and
then a flattening or decrease in reading speed with further
increases in magnification. Among this subgroup of low
vision subjects the peak reading speeds are found at between
19X and 24X magnification. This is also in agreement with
the findings of LEGGE et aZ.[9] who found that the low vision
subjects in their study had their fastest reading speeds at
between 3 degrees {13X} and 6 degrees {26X} character sizes.
GOODRICH [4] found that the average magnification chosen
by their 96 CCTV users was 21X.
Subject 10 gives results which differ from the other 9 low

vision subjects. He has had ten years experience using a
CCTV reading aid and while the results in Fig. I {B - F} show
that his reading speed decreases with decreasing field size, at
the highest magnification {Fig. IF} he is obviously faster than
all other low vision subjects. He was in fact faster than many
of the subjects with normal vision at this magnification. Fig-
ure 2 {B - E} indicates that his reading speed increases with
increasing magnification, at first rapidly and then more slow-
315
A
NORMAL VISION
'"
·
"
:.~
:~
\00.0
.E~
..
':!
• 50.0
30.0 30.0
Field Size (degrees) Field Size (degrees)
lOW VISION (SX) LOW Y1S10N(12xl
..
•
.~
:00.0
.. <'
rn .€
!~
• 3 ,
.-~~~,o
".o~
.. ,
3C.O 30.0
Field Size (degrees) Field Size (degrees)
lOW VISiON{2DXl lOW VIS!ON(32Xl
3G.G aD. 0
Field Size (degre-es) Field Size (degrees)
Figure 1. CCTV reading speeds as a function of field size

using 5 levels of magnification. In A, data for
18 subjects with normal vision are represented.
From B to F, individual data for 10 low vision
subjects at each level of magnification used are
represented. The 5 levels of magnification are
5X, 8X, 12X, 20X and 32X. Data of subjects
with normal vision are displayed for easier com-
parison.
ly, with his fastest reading speed being achieved at the higher
316
magnifications (20X and 32X) for each field size. This sub-
ject's distance and near visual acuities measured with the
Bailey-Lovie 10gMAR distance and near charts [13,14] were
6/300 (20/1000) and confirm the findings of GOODRICH et
al.[3] and LAGROW [15] that the reading speed of CCTV users
can improve greatly with training and practice.
Our results indicate that in general, most low vision sub-

jects achieve best reading speeds on the CCTV with a large
field size and high magnification. There was no relationship
between cause of vision loss and reading speed found in this
study. GOODRICH et al.[4] and ZABEL et al.[5] made similar
conclusions but LEGGE et al.[8] have postulated that the type
of field defect - this is, an overall loss of sensitivity as found
with media opacities or a central field loss - is the best pre-
dictor of reading performance. There appeared to be a slight
trend in this study for the subjects with central field defects
(1, 3, 8) to read more slowly than those with media opacities
(2, 9) but the number of subjects was too small to make any
general statement.
No relationship was found between reading speed and dis-

tance or near visual acuities. GOODRICH et al.[16] reported
a correlation coefficient of 0.46 between reading speed and
distance visual acuity for 12 low vision subjects but his subse-
quent studies [3,4] have not confirmed this relationship. A
moderate relationship (r = 0.71) was found between near visu-
al acuity and the magnification giving the fastest reading
speed. That is, subjects with better reading acuities achieved
their fastest reading speeds at lower magnifications than
those with worse reading acuities, but again there was a wide
variation.
A number of studies [7-9,17-19] have investigated the

effect on reading performance of the number of characters in
the field. With different field sizes and print sizes, the num-
ber of characters on the screen varied for each condition (see
Table 2). In Fig. 3, the reading speed data has been re-
analyzed in terms of number of characters. For the 18 nor-
mal subjects, the data can be fitted by an exponential curve
(Fig. 3A), with reading speed increasing as the number of
characters on the screen increases while character size (mag-
nification) has no effect. The results indicate that reading
speed increases up to approximately 15 characters on the
screen, but there is little change in reading speed with fur-
ther increases in the number of characters. LEGGE et al.[8]
317
A
NORMAL VISION
-: 100.0
:.~
"1;'
o •
!I
•
It
50.0
40
25
IS
10
Magnification
LOW VISION(2S"1
LOW VISION (4Q")
-a '0 100• 0
:-
100.0
: -0
0. ~ 0. £
"'1;' ':~
• p
i~ 11
a::: .!
50.0 = 50.0
15.0 20.0
1~. 0 20.0 25.0
"agnllication "agnlflcatlon
LOW vISION(1<rl
LON VISION (16')
" 100.0
.
-: 100.0
!~
,.
0.-
fit .£
"' E O{'
~~ i••~
..
____---"~o:::_---<--2
!j
-----
a:: so.a
a:: So_D
15.0 20.0
15.0 25.0
"agnlflcatlon ".anltlc.Uon
Figure 2. CCTV reading speeds as a function of magnifi-

cation using 4 field sizes. In A, data for 18 sub-
jects with normal vision are represented. From
B to F, individual data for 10 low vision subjects
at each field size used are represented. Data of
subjects with normal vision are displayed for
easier comparison.
found that for normal subjects a window width of 4 charac-

ters was sufficient to allow them to attain their maximum
318
reading speeds. The methods used in these two studies are

not identical. In our study, subjects manually moved the X-Y
platform holding the printed material, whereas in the studies
by LEGGE et al.[8,9] the platform was moved mechanically to
scan the print across the screen. The difference between the
results is substantial and it is therefore unlikely that the
movement of the platform is solely responsible for the dis-
crepancy.
RAYNER et ale [18,19] found that 3 to 4 characters were

needed to the left of fixation and 15 characters to the right
of fixation to achieve best reading performance. Our results
are in agreement with their findings.
For low vision subjects, LEGGE et ale [9] found that 4

characters on the screen allowed maximum reading speed.
Our results vary with individual subjects. Figures 3B-3F give
the results for some of the low vision subjects. Results for
three of the subjects (2, 5, 6) showed the same trends as the
subjects with normal vision. (Results of subjects 2 and 5 are
given in Figs. 3B and 3E). Subject 2 achieves near maximum
reading speed with approximately 10 characters on the
screen, with character size having little effect on reading
speed. He achieves his best reading speed at the minimum
magnification and maximum number of characters. Subject 5
is a slower reader, but reading speed increases with increas-
ing number of characters on the screen, with character size
again having no effect. For these subjects, the number of
characters on the screen appear more important than magni-
fication.
For the other low VISIon subjects, character size did

affect results (Fig. 3D - 3F). At the low magnifications
(small character size) subject 3's visual acuity is probably
limiting reading performance (Fig. 3D). Increasing the num-
ber of characters on the screen therefore gives no change in
reading speed. However, for each increased magnification,
the slopes of the curves increase. For this subject the results
suggest that increased magnification gives more improvement
in reading speed than increasing the number of characters.
However, if the curves can be extrapolated, it appears that if
field size (number of characters) could be increased, while
maintaining high (20X) magnification, reading speed would
improve further. Similarly, the results for subject 8 (Fig. 3E)
who is a very slow reader, indicate that magnification is more
important to him than field size.
319
NORMAL Y~IS::::IO:.:..N_ - - - - lOW VISION (522.- _
..
-: 100.
" loo.a
0-
:a ~'
0-
i~
a •
11
o-
Il:
SO.D
:i~
:.!
II: 50. ~
lO.C 15.0 20.G 25.3

No. of Char act.,. No.of Characte,.
i;"v
lOW VI5ION(S~ ___ ~ LOW VISION (S3l _ _ _ _
,,----
/
/
I
/ " ------
-= ~GG.o
H'(b/___
/
:. ~ I /
.5 "E
l
it /
Q.Oo.'c.,-~~;-;;--,;,;-,.;;-,--;';,,--;;.,---:;,O-;s.Gc--~--,o 10.0 :5.0 20.G 25.0
No. of Char act.,. No. of Charact.,.
-- --- -
LOW VISION{S.~ __ _ LOW V!S!ON(510)
----
/'
/
/
,,'
----
/
Hpi
'V
0-
100.0
/ ,,:00.0 /
i~ /
.
a •
.5 ~ /
11 -•
II: SO.O / !",r~
~
/
10.0 15.0 20.0 25.0

No. of Chancte,. No. ot Chancter.
Figure 3. CCTV reading speed as a function of the num-

ber of characters visible across the screen. In
A, data for 18 subjects with normal vision are
represented. From B to F, individual data for 5
of the low vision subjects using 5 levels of mag-
nification are represented. The 5 levels of
magnification are 5X, 8X, 12X, 20X and 32X.
Data of subjects with normal vision are dis-
played for easier comparison.
The experienced CCTV user (subject 10, Fig. 3F) reached

320
his maximum reading speed at both 5 1/2 characters for 32X

magnification and 9 characters for 20X magnification. Per-
haps with practice, the choice of magnification and field size
becomes less critical for best reading performance.
It appears therefore from these results that some sub-

jects with low vision read more efficiently with wider fields
(maximum number of characters) - these appear to be the
faster readers. The results of most of the low vision subjects
who read more slowly suggest that maximum reading speeds
would be achieved if it was possible to give both high magni-
fication and wide field of view. In the clinical setting, this is
indeed very often the patients' request. Our results substan-
tiate their observation. The results explain the observations
of GOODRICH et al.[3,4] and ZABEL et al.[5] that low vision
patients prefer to use larger CCTV monitors, giving greater
number of characters on the screen, rather than the small,
higher resolution monitors.
4. Conclusion
In summary, reading speed is an individual characteristic of

low vision patients and should be evaluated as part of the
assessment of reading performance [20]. For the low vision
patients with faster reading speeds (perhaps 75 words/minute
and above) it appears to be valid to advise minimum magnifi-
cation for maximum field size on the CCTV. For low vision
patients who read more slowly, reading speed may improve at
higher magnifications, despite reduced field size. For these
patients, reading performance should be assessed at magnifi-
cations higher than predicted from other examination find-
ings, to determine the magnification which will give most
efficient reading. Further work is needed to determine if
these findings also apply to conventional optical low vision
aids.
321
5. Acknowledgements
This work was supported in part by a Ministry of Labour, Gov-

ernment of Ontario Grant. We thank C. Machan, A. Bloesch,
B. Brown and R. Jones for their assistance.
6. References
1. A.W. Johnston: Technical note: The relationship

between magnification and field of view for simple
magnifiers. Aust. J. Optom. 65, 74 (1982)
2. E.B. Mehr, A.B. Frost, L.E. Apple: Experience with

closed circuit television in the blind rehabilitation pro-
gram of the Veteran's Administration. Am. J. Optom.
Arch. Am. Acad. Optom. 50,458 (1973)
3. G.L. Goodrich, E.B. Mehr, R.D. Quillman, H.K. Shaw,

J .K. Wiley: Training and practice effects in perform-
ance with low-vision aids: A preliminary study. Am.
J. Optom. Physiol. Opt. 54, 312 (1977)
4. G.L. Goodrich, E.B. Mehr, N.C. Darling: Parameters in

the use of CCTVs and optical aids. Am. J. Optom.
Physiol. Opt. 57, 881 (1980)
5. L. Zabel, H. Bouma, H.E.M. Melotte: Use of the TV

magnifier in the Netherlands: A survey. J. Visual
6. R.L. Duchnicky, P.A. Kolers: Readability of text

scrolled on visual display terminals as a function of
window size. Hum. Factors 25, 683 (1983)
7. R.S. Kruk, P. Muter: Reading of continuous text on

video screens. Hum. Factors. 26, 339 (1984)

25, 239 (1985)
322

Psychophysics of reading II. Low vision~ Vision Res.
25, 253 (1985)
10. H.H. Munro: The Short Stories of Saki (Modern

Library, New York 1951)
11. B. Brown:. Reading performance in low vision patients:

Relation to contrast and contrast sensitivity. Am. J.
12. P. Muter, S.A. Latremouille, W.C. Treurniet: Extend-

ed reading of continuous text on television screens.
Hum. Factors 24, 501 (1982)
13. I.L. Bailey, J.E. Lovie: New design principles for visu-
al acuity letter charts. Am. J. Optom. Physiol. Opt.
53, 740 (1976)
(1980)
15. S.J. Lagrow: Effects of training on CCTV reading

rates of visually impaired students. J. Visual Impair-
ment & Blind. 75, 368 (1981)
16. G.L. Goodrich, L.E. Apple, A. Frost, A. Wood, R.

Ward, N. Darling: A preliminary report on experienced
closed-circuit television users. Am. J. Optom. Physiol.
Opt. 53, 7 (1976)
17. E.C.Poulton: Peripheral vision, refractoriness and eye

movements in fast oral reading. Br. J. Psychol. 53,
409 (1962)
18. K. Rayner, A.D. Well, A. Pollatsek: Asymmetry of the

effective visual field in reading. Percept. Psychophys.
27, 537 (1980)
19. K. Rayner, A.D. Well, A. Pollatsek, J.H. Bertera: The

availability of useful information to the right of fix-
ation in reading. Percept. Psychophys. 31, 537 (1982)
20. J.E. Lovie-Kitchin, K.J. Bowman: Senile Macular

Degeneration: Management and Rehabilitation (But-
terworths, Boston 1985)
Predicting Reading Performance in Low Vision Observers
With Age Related Maculopathy (ARM)
Gary S. Rubin
1. Introduction
Age-related maculopathy (ARM) is a leading cause of low

vision, especially among the elderly. Loss of central vision
due to ARM can seriously interfere with performance of
everyday visual tasks such as reading and face recognition.
Clinical trials are under way [1] to evaluate the effectiveness
of laser photocoagulation for preventing severe visual impair-
ment due to the growth of new vessel membranes in the
macula. The laser treatment itself produces a permanent,
absolute scotoma. Therefore it is of critical importance to
compare the loss of visual function resulting from the treat-
ment with the loss that will occur if the diseased eye remains
untreated.
Traditionally, clinical trials such as these have relied on

measures of visual acuity as the principal means of assessing
visual function. In our previous work [2] we have found that
acuity is a good predictor of the optimal character size (mag-
nification) for low vision readers. However, acuity is not as
good a predictor of reading rate - especially for observers
with loss of central vision. BROWN [3] reported that contrast
sensitivity measured with Arden plates was better than Snel-
len acuity for predicting low vision reading performance.
However, the Arden plates have been shown [4] to be espe-
cially susceptible to observer bias, particularly when used
with an elderly population. We have measured contrast sensi-
tivity functions (CSFs) of low vision observers using a
criterion-free procedure. I will discuss whether these meas-
ures of contrast sensitivity improve our ability to predict low
vision reading performance.
As a preview of the results, we find that the CSF does

provide some additional information, but that it is unneces-
sary to measure the entire CSF to obtain this information. A
simple measure of letter recognition using large, low contrast
letters is equally informative.
324
2. Methods
In the first part of the study we measured reading rates and

CSFs for 28 low vision observers with diverse pathologies and
degrees of vision loss. Ten of these observers had central
field loss, which we defined as a dense scotoma within the
central 10 degrees (diameter). Thirteen had field loss con-
fined to areas outside the central 10 degrees. The remaining
five had intact fields but cloudy ocular media. Five of the
observers with peripheral loss also had cloudy media.
In the reading experiment, observers read aloud lines of

text that were scanned from right to left across the face of a
TV monitor. The text was generated with a digital image
processing system and displayed wit~a contrast of 96% and a
maximum luminance of 300 cd/m. Reading rates were
measured for letters ranging in size from 24 degrees down to
the observer's acuity limit. For each letter size, the experi-
menter adjusted the scan rate until the observer made one or
two errors in reading a 16 word line of text. A reading rate
was computed as the number of words read correctly per min-
ute.
CSFs were measured with vertical sinewave gratings and

a two-alternative spatial forced-choice procedure. The grat-
ings ~ere displayed on a CRT with a mean luminance of 100
cd/ m and an angular subtense of 12 degrees. The gratings
were turned on gradually and remained on until the observer
responded. Since most of our observers had cutoff frequen-
cies below 10 c/deg, we measured CSFs over an unusually low
range of spatial frequencies - 0.1 to 6 c/deg.
In addition to CSFs, we also measured each observer's

distance acuity using the Good-Lite 3 m (10 foot) chart with
Sloan letters, near acuity with the Sloan M cards, and con-
trast sensitivity for letters using a test to be described below.
325
3. Results
Figure 1 shows reading data for two low VISIon observers.

Character size is plotted on the horizontal axis
300
....
E 100
"(J)
"0
L iii'
a I
I
~ I
'J
30 I
I
I
W I
I- I
< I
0:: d
L.:l
10
Z
• TA - Paripharel Less
o o DE - Cantrel Less
<
W
0::
•1 .3 3 10 30

Figure 1
and reading rate is plotted on the vertical axis, both on loga-

rithmic scales. The solid line at the top is average reading
data for four observers with normal vision. Subject TA,
whose data are plotted with filled symbols, has peripheral
field loss due to optic nerve atrophy, and an equivalent Snel-
len acuity of 6/15 (20/50). His data are typical of low vision
observers with intact central vision in that his reading is fast-
est for an intermediate range of character sizes - about 2
degrees to 3 degrees. Subject DE has a 7 degree central sco-
toma due to macular disease, and an acuity of 6/120 (20/400).
His results, shown as open symbols, are typical of observers
326
with central field loss in that his reading rates are lower
overall and improve or level off with ever increasing letter
sizes.
CSFs for these two observers are shown in Fig. 2.
1000
300
>-
.....
100
>
......
.....
...... 30
Vl
Z
/i>-- ____ -Q
W 10 , , ,,
Vl ,, ,,
..... , ,, ,,
,,
Vl 3 ,, ,,
< CI
a::
..... ~
z • TIl - Peripheral Lass

D
U
o DE - Centra) Loss
.3
.1
.1 .3 3 10
SPATIAL FREQUENCY (c/deg)

Figure 2
The solid curve at the top represents average data for six
observers with normal vision. The curve does not show the
typical high frequency roll-off due to the low spatial fre-
quencies tested.
SUbjects T A and DE are again represented by filled and

open symbols, respectively. Note that DE's contrast sensitiv-
ities are markedly depressed at all spatial frequencies, con-
sistent with KELLY's data [5] for artificial central scotomas.
TA's contrast sensitivities are near normal for the lowest fre-
quencies tested. This would not be expected solely on the
basis of peripheral vision loss. Constricted fields should have
the opposite effect, reducing sensitivity for low spatial fre-
quencies but not high [6]. Similar results were obtained for
327
other subjects with peripheral loss, perhaps indicating some

macular involvement in these eyes as well.
We used a mUltiple regression analysis to investigate how

well measures of visual acuity and contrast sensitivity predict
reading performance. There are many attributes of the con-
trast sensitivity function which might be related to reading
performance: cutoff frequency, peak sensitivity, area under
the curve, or even contrast sensitivity at a particular spatial
frequency. We considered all of these factors, and in our
sample many were highly correlated. Again we found that M
acuity was the best predictor of optimal character size, with
a correlation coefficient of 0.8. None of the contrast sensi-
tivity factors significantly improve the correlation. A scat-
ter diagram for M acuity vs. optimal character size is shown
in Fig. 3. The regression line indicates that the optimal char-
acter size is approximately equal to the M acuity •
./
•
•
10 30
OPTIMAL CHARACTER SIZE (degrees)

Figure 3
When a similar analysis was applied to reading rate, M

acuity did not fare quite so well. It was still the best single
328
predictor but the correlation coefficient dropped to 0.6. The

scatter diagram is shown in Fig. 4. Several contrast sensitivi-
ty factors significantly improved prediction, the best being
either the area under the CSF or the peak sensitivity. With
either of these factors, the correlation coefficient was 0.8.
30r---------r---------~--------.
Correlation
Coefficient = -.63 •
10
>-
l-
S
U
<{
~
3
IL-________ ________ ________

10 30 100 300
MAXIMUM READING RATE (words/minute)
Figure 4
Figure 5 compares M acuity with peak contrast sensitivi-

ty for subsets of our low vision subjects. The shaded bars
represent correlations of reading rate with acuity and the
open bars represent correlations of reading rate with contrast
sensitivity.
Acuity is a better predictor of reading speed than con-

trast sensitivity when the group is considered as a whole, or
when considering just those observers with intact central
vision. But for the subset of subjects with central field loss,
acuity is worse than contrast sensitivity (and neither corre-
lates particularly well). It should be noted that these sub-
samples are small, and additional data are being collected to
determine whether these preliminary conclusions hold up.
329
I--
Z
W m M Acuity
......
U
o Pack Contrcst Sans1t1v1ty
...... .8
LL
LL
W
a .6
u
z
a
......
.4
I--
<
..J .2
W
c:::
c:::
a 0
u
ALL CENTRAL CENTRAL
SUBJECTS INTACT LOSS
Figure 5
CSFs measured in the manner described here require fair-

ly sophisticated equipment and are time consuming. {These
took about one hour.} Recently GINSBURG [7] has introduced
a printed contrast sensitivity testing system that measures
the CSF in a matter of minutes. We have compared the
printed test with an abbreviated CRT-based test (a 20 minute
version) in patients with glaucoma, RP and ARM. Figure 6
compares test-retest reliability with the two methods for one
observer.
This observer, with ARM, gave fairly reliable data with

both tests, although there was a substantial discrepancy at
one spatial frequency with the printed chart.
Results for a second observer with ARM are shown in Fig.

7. Her results were less reliable with the CRT test and
extremely unreliable with the printed chart.
Nevertheless, our results suggest that it is not necessary

to measure entire CSFs in order to predict reading perform-
ance, but only sensitivity near the peak of the function. We
330
0 "' CRT
300
VISTECH CHART
:\
300
__ - -._Q. .TII!:!!
100 0- "
30 ',--.
10
3t
o ~L---'I'=-!5----'.---~-,;O-2~'B-----' 12 16
SPATIAL FREOUENCY lc/degJ SPATIAL FREQUENCY (c/deg)
Figure 6
"
VISTECH CHART
CRT
300
300
100
100
30 .....
30
.-~_d,
,
10
10 b.
'0
o~ , 15 12 18
15 12 18
SPATIAL FREQUENCY (c/deij)
SPATIAL FREOUENCY (c/degl
Figure 7
have examined two methods of measuring peak sensitivity.

One uses large, low contrast letters, and the second uses low
frequency squarewave gratings.
Denis Pelli at Syracuse University has designed a series

of test slides for measuring peak contrast sensitivity. Each
slide contains four Sloan letters subtending 3 degrees and
viewed in a series of descending contrasts. There are two
slides at each contrast. The subject must name the letters,
and threshold is determined by the lowest contrast at which
331
at least 6 of 8 letters are correctly identified. The correla-

tion between peak sensitivity and contrast sensitivity for
these large letters is 0.9.
100
(J)
a: Correlation
ILl
~ Caeff ic ient =.91
•
~
ILl
...J
30
a:
0
lL.
>-
~
:;
i= 10
enz
ILl
(J)
~
(J)
<t 3
a:
I-
Z
0
u
I
3 10 30 100 300
PEAK CONTRAST SENSITIVITY
Figure 8
Other tests using low contrast letters to measure visual

function have been designed by BAILEY and LOVIE [8] and
REGAN and NEIMA [9].
The second test takes advantage of the fact that the con-
trast sensitivity function for squarewave gratings flattens out
below the peak of the sinewave CSF [10]. By measuring sen-
sitivity for a single low frequency grating, one should have a
measure that correlates highly with the sensitivity at the
peak of the CSF. Sinewave CSFs were measured for a sepa-
rate group of 16 low vision observers, all with ARM. Con-
trast sensitivity was also obtained for a 0.2 c/deg squarewave
grating (Fig. 9). Both were measured with a CRT-based
forced-choice procedure. As predicted the square wave sen-
sitivity correlated highly (correlation coefficient = 0.9) with
the peak sensitivity of the CSF.
332
300
>- Correlation
I-
:;
l-
Coefficient =.93 •
(/)
z 100
w
(/)
l-
(/)
<X
0:::
I- 30
Z
0
U
W
.
~
~
I 10
W
0:::
<X
::J
0
(/)
3
3 10 30 100 300
PEAK CONTRAST SENSITIVITY
Figure 9
4. Conclusion
While clinical application of the CSF has greatly increased

over the past decade, it remains to be shown how that infor-
mation can be used to greatest benefit. For the evaluation of
visual function, the CSF provides information which comple-
ments that provided by measures of visual acuity and visual
field. M acuity is useful for predicting the character size
(and hence, magnification) with which a low vision observer
reads best, and contrast sensitivity along with acuity gives
some indication of how well that person should be able to
read.
Our data indicate that careful measurement of one

aspect of the CSF, namely peak sensitivity, is as informative
as characterizing the entire function, at least for patients
with a loss of central vision. A simple letter recognition test
using large, low contrast letters, or a measure of contrast
sensitivity for low frequency squarewave gratings seem to
provide the necessary information.
333
5. References
1. Macular Photocoagulation Study Group: Argon laser

photocoagulation for neovascular maculopathy: three
year results from randomized trials. Archives of
Ophthalmol. (1986) in press.

25, 253 (1985)
3. B. Brown: Reading performance in low vision: Relation

to contrast and contrast sensitivity. Am. J. Optom.
Physiol. Opt. 58, 218 (1981)
4. S. Sokol, A. Domar, A. Moscowitz: Utility of Arden

grating test in glaucoma screening: High false-positive
rate in normals over 50 years of age. Invest.Ophthal-
mol. Vis. Sci. 19, 1529 (1980)
5. D.H. Kelly: Photopic contrast sensitivity without

foveal vision. Opt. Lett. 2, 79 (1978)
6. R. Hilz, C.R. Cavonius: Functional organization of the

peripheral retinal: Sensitivity to periodic stimuli.
Vision Res. 14, 1333 (1974)


(1980)
9. D. Regan, D. Neima: Low-contrast letter charts as a

test of visual function. Ophthalmology (Rochester)
90, 1192 (1983)
10. F.W. Campbell, J.G. Robson: Application of Fourier

analysis to the visibility of gratings. J. Physiol.
(Lond.) 197, 551 (1968)
Visual Impairment and Disability:
Enhancement and Substitution
J.A. Couturier, J. Gresset
1. Introduction
The topic of this paper is a statistical reporting of the servi-

ces and aids provided by the Institut Nazareth et Louis-
Braille to low vision and blind individuals. The purpose of the
reporting is two-fold. We intend to explore the relationship
between the types of aids that are prescribed to the visually-
impaired population following the identification of their level
of impairment. In an attempt to illustrate levels of visual
disability, we also explore the two ways in which it can be
reduced when using vision enhancement and/or vision substi-
tution aids.
The levels of visual impairments are those recommended

by the World Health Organization [1,2]. In this study, the dif-
ferent levels will also refer to the correspondence made by
COLENBRANDER [3] in identifying levels of moderate,
severe, profound, near total and total impairments, corre-
sponding to levels 1, 2, 3, 4 and 5 respectively.
Impairment is associated with disability, another dimen-

sion of vision [3,4]. A visual impairment may give rise to a
disability, that is it can decrease the ability of an individual
to do a specific task like reading. Six different levels of visu-
al disability may be identified: slight visual disability, moder-
ate visual disability, severe visual disability, profound disabil-
ity, near total and finally total visual disability [3]. Each
level refers to a degree at which an individual can or cannot
perform different visual tasks, from gross visual to fine
detailed tasks.
Impairment and disability can be reduced.

COLENBRANDER [3] illustrates two ways to reduce visual
disability, that is using vision enhancement and/or vision sub-
stitution techniques. Examples of vision enhancement aids
used by the author are magnifiers or simple techniques such
335
as finger tracking while reading. Memory, hearing or the use

of braille are examples of vision substitution aids.
2. Method
To introduce this reporting, a profile of low vision and reha-

bilitation services provided by the Institut Nazareth et Louis-
Braille is presented (Fig. 1). Visual impairment and eligibility
are determined by the low vision clinic's eye specialist.
According to the needs expressed by the individuals, visual
disability is measured for specific tasks such as reading, writ-
ing, drawing, computing, traveling, personal management,
household work, and so forth. This is done by low vision
instructors and the rehabilitation personnel involved with
written communication activities, orientation and mobility
(including guide-dogs), activities of daily living, leisure and
vocational activities. A multi-disciplinary training program
is identified and a significant effort is made in giving the aids
to best meet the visual or non-visual needs of the clients,
children or adults. The particularities of this public agency,
as mentioned by GRESSET et aZ.[5], are the grouping of all
the services under the same roof and the provision of services
to all ages through the medicare system.
The aids compiled to illustrate vision enhancement and

substitution are those found under the medicare regulation.
Not all ages, aids and techniques appear in the results.
The subjects are taken from the study initiated by

GRESSET et aZ. The population was seen prior to 1984, that is
from 1977 to 1984, and classification of impairment was done
according to the data of the first low vision examination (in
the first year following the admission). On the other hand,
the aids compiled are those possessed by the subjects at this
present time. This may limit the study, as some subjects may
have had vision deterioration and their level of impairment
may have changed. We have no data to avoid this possible
source of error but the deterioration rate extrapolated from
F AYE's study [6] is limited to 8%.
A total of 274 subjects are in the study, with a maximum

of 60 subjects in each of the six categories of visual impair-
ment. The age range and distribution are shown in Fig 2.
336
Social work
r---- Ophthalmology
Optometry
, Itinerant teacher
LOW VISION CLINIC----1.~ Eligibility ~ Levels of impairment
1
Psychology
Health care
Physical therapy
Occupational therapy
+---========~-----~.~ Disabities
COMMUNICATION SKILLS • reading

•
ORIENTATION AND MOBILITY
•
·
writing
drawing
ACTIVITY
•
OF DAILY LIVING SKILLS
•
•
computing
manual work
LEISURE
•
ACTIVITIES
•
•
traveling
personal management
•
VOCATIONAL SKILLS
•
•
household work
employment
• mainstreaming
Figure 1. Low V1Slon and rehabilitation services profile

of the Institut Nazareth and Louis-Braille reha-
bilitation center.
About 52% of our population was between 21 and 40 years of

age. According to the medicare regulation of 1984, older
individuals could be seen for evaluation and training only.
The data collected to illustrate vision enhancement and

vision substitution aids are a compilation of optical, electron-
ic and mechanical aids prescribed to either our functionally
sighted or functionally blind clients (Fig. 3). Vision enhance-
ment aids include electronic optical aids, absorptive lenses,
reading stand, telescope, microscope, Pupilens, magnifier and
large print calculator. Vision substitution aids include the
337
Number
200 100
90
80
150
70
Number of casas
60
100 50 1m Number of prescriptions
40 .... Frequency of attributions
30
50
20
10
O~~~~~~ 0
11020 21 to 40 41 to 60 61 and older
Figure 2. Age distribution and frequency of attribution in

each age group
Optacon, the APH and conventional taperecorders, the Ver-

saBraille, the talking calculator, arithmetic and geometric
sets, the braille calculator, writing guides, braille writing
aids, the long cane, and guide-dogs. The typewriter has com-
bined visual and non-visual characteristics. The number of
aids shown in the enhancement type, being less than in the
substitution type, implies that vision enhancement aids are
more versatile than vision substitution aids, that is, they can
be used for more than one task. A total of 744 aids have been
prescribed. As the data was collected prior to 1984 and the
medicare law does not include all types of aids for all ages,
computer aids and electronic travel aids are partly compiled.
Adapted aids for household work are also not included, as well
as numerous techniques and skills used in training.
338
7"10
•
120
•
107
100 Enhancemen
89 90
Substilullon
~
80
to 0 Combined
53 52
"
z
40 35 35 36
20
2
0
. .... -g. 8. 8. "a.. ~ ;; ~
"2 (;
..
c: ;;; ;;;
~ ;;
.... (; O!
~;
..
" ~ ~u 'i'8'
. -g. ..... i. " .. . ...... .
c: 0
O!
0 "E "E 0; ;; i ;; ."
":; 0 >.
J!
;; .;
0 0
c:
.
;; !;l c: :;
. .
:; :;
S
0 0
! ! -9on
0
~ 0>
~ !! .9l 0 0> Jl 0
g .!l .0 .0 &~"
Q. -0 i:; Q.
:::! u
Q. Ii
i u .[ (, -' CJ
...
0
>-
?!
8. 8. ;;;
c e- a:
0 c:
;; E
O!
e ..
0
0
O!
.e>
~
Q.
!! !! >
:r Ii
~
Ii
.
'"c: E0 ~ .~
~ ~;;;
Q.
iii a
n: .... Wal
.
D
U « c:
..
.!! !!
w -'
.(
c: i
E
>
c: ~
0
0 <
Figure 3. Aids prescribed for the total population (Popu-

lation N=274, Number of aids N=744)
3. Results
Figures 4 and 5 show the frequency of attribution of aids for

each level of impairment, one aid being reported per case.
When the population under 5 years old and the multi-
handicapped are not included in the reporting, a greater num-
ber of persons utilize at least one aid as the severity of
impairment increases.
Individuals categorized in Level 3 of impairment show the

highest ratio of the number of aids per case (Fig. 6), followed
by those in levels 4 and 5. An average number of 4.2 aids is
utilized by each individual of Level 3.
339
om -Number of cases
~ - Number of al1ributions 100
- - % of allributions 90
80
70
.8E 60
J 50 %
Z
40
30
20
10
0
2 3 4 5 9
Levels of impairment
Figure 4. Distribution of population and frequency of

attribution in each level of impairment
As we compare the types of vision enhancement aids and

vision substitution aids used for each of the 5 levels of
impairment, Fig. 7 illustrates that vision enhancement aids
account for 82% of all prescriptions in Levell. The 13% use
of vision substitution aids stands for the APH taperecorder
(10 users) and the long cane (4 users). This fact is probably
due to, on one hand, students' needs in reading some school
books with the recorder, and on the other hand, the use of the
cane for identification purposes on the street corner. The 2
Perkins braillers mentioned are probably utilized by those
who may have had vision deterioration and would then be
classified in another level.
The same observations may be made for the population of

Level 2 (Fig. 8), as 65% of all prescriptions are vision
enhancement aids and 26% vision substitution aids. Besides
the APH taperecorder and the long cane, the talking calcula-
tor and the Perkins bra iller are used by 5 and 4 persons
respectively.
The population of Level 3 shows a tendency toward vision

substitution aids (Fig. 9), 48% of all prescriptions being noted,
340
. Number 01 cases
~ . Number 01 aUributions
100
90
80
70
A
E
60
%
::J
z 50
40
30
20
10
0
2 3 4 5
Levels 01 impairment
Figure 5. Distribution of population and frequency of

attribution in each level of impairment. Popu-
lation over 5 years old and no associated
impairment.
whereas 44% of all prescriptions are vision enhancement aids.

However, we find 18 CCTV users at this level. A particulari-
ty of this group is the use of both vision substitution aids and
vision enhancement aids by the same person. A main-
streamed student, for example, will use braille for basic
school textbooks like French and Mathematics, use a CCTV
for handouts given in the classroom and use a lOX microscope
to read the large print dictionary. A VersaBraille will be used
for writing schoolworks and a felt-tip pen to write short mes-
sages or a short-answer examination. For this level of
impairment, the typewriter, identified under the combined
type may be considered to be either a vision enhancement aid
or a vision substitution aid.
The population of Level 4 still shows a 15% use of visual

aids (Fig. 10). This observation is interesting if we compare
it to the description of the level of "near total visual disabili-
ty" made by COLENBRANDER [3], by which vision is unrelia-
341
iii - Number of cases

500
FJ.a - Number of aids prescribed 5
- Ratio of aid by case
400 4
CD
~ 300 3
::J
Z
200 2
100
o o
2 3 4 5 9
levels of impairment
Figure 6. Total number of aids prescribed by level of

impairment for population over 5 years old and
no associated impairment
ble and one relies mainly on other senses. Although the

author mentioned no simple one-to-one relationship between
levels of impairment and levels of disability, some links are
observed between the two in the course of this study. The
same observation is made for Level 5 (Fig. 11), as with "total
visual disability", one relies on other senses entirely. In fact,
no vision enhancement is noted at this level. At this present
time, we are, however, inclined to use GENENSKY's "func-
tionally blind" category [7] to characterize the visual disabili-
ty of these two last groups.
The population of these last two levels of impairment

show special characteristics. As we said that vision substitu-
tion aids were less versa tile than vision enhancement aids, we
should be inclined to have more prescriptions at these levels
than at levels 1 and 2. The main reason is that the population
of these levels encompasses more children under 5 years old,
more multi-handicapped and more adults over 40 years of
age. These individuals have not yet received all the attention
342
120
100
80 • Enhallcemenl
~
fa
f
i 60
SubSlilution
Figure 7. Aids prescribed for the population of level 1

(Population N=60, Number of aids N=129)
that is given to the 6 to 20 and the 21 to 40 age groups. The

number of cases studied at these levels is, however, smaller
than in the first three categories.
The sense modality most prevalent through the use of

enhancement and substitution aids is analyzed with our popu-
lation and comparisons are made between the 5 levels of
impairment (Fig. 12). When the typewriter is placed under
the enhancement type for levels 1, 2 and 3, we notice that
vision is the sense that is used by the majority of persons,
with the sense of touch being next. Hearing has probably not
received full justice, in mobility for example. Although the
long cane has been classified under the sense of touch, hear-
ing is still an important sense modality for orientation pur-
poses.
343
120
100
80
~ SubstilUlion
60
[] Combined
40 36 33
18 18 15
20
0
.., ..,c:
". "2.. .,.." .."8. a. 'c 2." .~..~
on on
co
on
8. ~ 0
Ii
0
<:
j! ;,::
c: Q. S
"iii ~ en
u co :>
°a co ;;;
0
a.
co '6 I- U
" ....
::::!!
e-.. co
0 "~
co ~ I-
.2 .~
.,a.
II:
c 0
E .CJ
~ -< ~
Uj co
....I
Figure 8. Aids prescribed for the population of level 2.

344
120
100
80
60
.
33 31
40 25 26
20
0 :2
1 .. "
c
••
• !c ~ 00X. JI! !
I
(; ~
~j '!! 'I? 'iii
"ii JI
'~ • '8~
:1!
0
'iL
""
"'0
'~
•
.......
0
0
I! !
0
0
.
~
I ! >•
Q.
.....
Gl-
i ... l; ~~ Q.
'~
a.
0
~ E
.11 ·c
c 0
Q.
~ 11 :x: OJ
iw < •
~
a:..(
c
9
~ E
t
c
0
0

(Population N=59, Number of aids N=Z44)
345
120
9%
100
•
90
60 77 %
40
20
4 1 2 4
0
.., . .. c
.
.,c -g c .~ (; Oi Oi !!. (;
~ 8.u0 ~
0
'iii ~
;;; ;;
!.l 'E0 'E0 ';;; :;
.c
!!! .!'
., :;
;;; .!! :t '6.
8- D
"~ !"" .,.;, ;;;"
.
~
.!!
1i ., ~
'6
~
~
0
<;
:>
0..
0> u
::.
;;;
,..
~
0..
8. 8. 0; "0>
u
e..,
.~ I-
u
0
co J: .'§
I-
!!! !!! > c
..a.
([
'1" 0 ".i
~ n I ;;; ;;;
~ « 0>
a: 9
c I-
<i. C
iii iii
...J .,
c>
0
0

346
120 3"10
•
•
100 Enhancement
fJ Substitution
80
0 Combined
60
40
15
20 2
.. . . - . - ij f
0 0 0 0 0 0 0 0 0
.. .." ..
.,c
•.
..a .. .
c
8 ..,co "Eco 'ii
Ii 0 (; (;
0 :2 I
'gK K :.=j .~ ~
ii ii .~ ~ " CD
! ~~
CD
Jl
0
~ ~
.
c (; 0 "3 "3 ",oo
<) .~
~ Z
.. ..
c "
•>- .g
<)
! ~ ,;, .II
! • ..
<) ",D
.11 .~
"'.II
_ CD
0
0.. ii "'~
.! -go- :~ ,~ ~ ~
CD U ::i: Q. <)
!
<)
..
<)
0
~ E 0- Ii m E ~~ J! >-
.11 ~c! > c ~
-
0 UlO
c .~ .;; CD
e 11 ::t ii
CD
'"
0..
j ~ •.e' a:
.(
c
.2
0- .~
i
CD
E
w ....I
•c> E
.€
0
0 .<
Figure 11. Aids prescribed for the population of levelS.

(Population N=Z8, Number of aids N=61)
347
100 0/0
• - Vision
90
80 l~wMI - Hearing
70 0- Touch
60
50
40
30
20
10
0
2 3 4 5
Figure 12. Senses modalities through enhancement and/

or sUbstitution aids
4. Conclusions
With all these observations in mind, we want to conclude

about the possible applications one can make with the WHO's
and Colenbrander's models developed to identify levels of
impairment and levels of disability. First, these models are
helpful guides when one wants to describe the characteristics
of the visually impaired person without losing the perspective
of individualized needs. They provide full justice to each
individual, with the exception of level 4 where some function-
ally sighted are gathered with functionally blind. Second,
these models can serve as management tools to better under-
stand and plan the services we want to give to the visually
impaired population. Third, because of the multiplicity of
information gathered under one roof, the models provide
effective means to communicate in the field.
348
In conclusion, we support COLENBRANDER's remarks [4]

on the lack of available accurate scales on visual disability
and the effect of visual and nonvisual aids. A systematic list
of vision enhancement and vision sUbstitution techniques
needs to be drawn up. Documentation of their effects in dai-
ly living tasks is required.
349
5. References
1. World Health Organization: Intemational Classifica-

tion of Impairments, Disabilities, and Handicaps (WHO,
Geneva 1980)
2. Organisation Mondiale de la Sante: Classification

Intemationale des Maladies, (OMS, Geneve 1975)
3. A. Colenbrander: Dimensions of visual performance.

Trans. Am. Acad. Ophthalmol. Otolaryngol. 83, 332
(1977)
4. A. Colenbrander: Low vision: Definition and classifica-

tion. In Clinical Low Vision, ed. by E.E. Faye (Little,
Brown, Boston 1976) pp. 3-6.
5. J. Gresset, P. Simonet: A clinical profile of a young

visually handicapped population. Paper presented at
the International Symposium on Low Vision, University
of Waterloo, 1986.
6. E.E. Faye: The Low Vision Patient, (Grune & Strat-

ton, New York 1970) pp. 185-200.
7. S.M. Genensky: Functional classification system of vis-

ually impaired to replace the legal definition of blind-
ness. Ann. Ophthlamol. 3, 150 (1971)
The Clinical Profile of a Young
Visually Handicapped Population
J. Gresset, P. Simonet
1. Introduction
This paper presents the results of a study done at the low

vision clinic at the Institut Nazareth et Louis-Braille in Lon-
gueuil, Quebec. The Institute offers rehabilitation services
and visual aids to eligible visually impaired persons in western
Quebec and serves both a partially sighted and functionally
blind population, as defined by GENENSKY [1,2]. A low
vision examination is part of the Institute's admission proce-
dure. Information derived from clinical files offers data on a
complete population of visually impaired people, since it
includes all levels of visual impairment.
The study comprises data from 514 cases, approximately

one fourth of the population served by 'the Institute, and
includes more than 150 variables. For this presentation, we
have selected several results which characterize this specific
population. To better understand the particular nature of this
visually impaired population, a brief look at the provincial
program for the visually handicapped is necessary.
The Quebec law stipulates that a person is considered vis-

ually handicapped when acuity in each eye is less than 6/21
(20/70) after appropriate correction, with the exception of
visual aids, or when the visual field in each eye is less than 60
degree in the horizontal or vertical meridians. Every person
who meets this definition is eligible to receive services from
the Ministry of Social Affairs rehabilitation program and the
AMEO program, which sees to the distribution of mechanical,
electronic and optical aids.
The Institut Nazareth et Louis-Braille is one of the five

centers in Quebec which received accreditation for the appli-
cation of these programs. The rehabilitation services are
free, but the aids program is limited by some conditions.
Some types of aids, like the long cane, are universal, while
351
some others depend on the patient's age (under 36) or, since
1984, on the need of a particular aid at work or at school. All
kinds of aids can be loaned. A more detailed description of
these aids is presented by COUTURIER et al.[3].
The introduction of this medicare program was progres-

sive. At first, only those under 19 years of age were eligible.
Then the program was extended for those under 36, and final-
ly to everyone at work or at school. A Quebecer who does
not respond to these conditions can receive free rehabilita-
tion services, but is not eligible for the AMEO program.
However he may buy any aid he wants at a low price. As a
consequence of the medicare regulation, the population seen
at the Institut Nazareth et Louis-Braille is relatively young.
2. Method
In order to reduce errors in our population sample because of

heredo-familial diseases, we have randomly selected a first
file, and then systematically every tenth file in alphabetical
order. The data considered in this study are limited to the
year following admission.
3. Results
Demographic data are presented in Fig.!. The age (average

32 years and 11 months) is the main characteristic of this
population. As shown in the figure, more than half of the
population is under 40 years old. This is noteworthy, since on
the basis of CNIB's statistics [4] more than one half of the
legally blind pouplation in Canada is over 50 years old. We
found a difference between sexes which has already been
reported by F AYE [5].
The classification of pathologies considering the type and

location of the disorder is based on the World Health Organi-
zation's classification [6]. As shown in Fig. 2, the three main
causes of visual impairment are retinal disorders, congenital
anomalies and diseases of the optic nerve. When comparing
these results with other studies [5,7,8], differences are noted
352
IEjo Wo~n I
Sex
Age
Men
250
200
150
100
50
o
o to 5 6 10 20 21 to 40 41 to 60 61 10 80 up 10 81
Figure 1. Demographic data (Age and Sex; Population

N=514)
(Fig. 3). The frequencies of the main retinal disorders of the

present study are compared to those observed by F AYE [5].
The differences in the frequencies of retinitis pigmentosa,
macular degeneration and diabetic retinopathy in this study
compared to that in Faye's study can be explained by the age
difference between the two studies. This is why retinal disor-
ders are seen in 34.24% of cases in the present study com-
pared to 52.12% of cases in Faye's study.
The presence of an associated impairment is found in

7.7% of all cases. Another characteristic of our population is
that one functionally sighted eye is associated with a func-
tionally blind fellow eye in 12.5% of all cases.
The distribution of the ametropia (both spherical and cyl-

indrical) shows a more accentuated dispersion in our popula-
tion than with a normal population [9] (Fig. 4). The effect of
emmetropization seems to be diminished in a visually
impaired population. Another interesting fact concerning
refraction is that visual acuity after refraction is significant-
ly different from the entrance report's visual acuity record-
ing. Mean acuity according to entrance report data is X =
353
Eyeball and mYOPia-::::4~5~rIJlIlmlm

Retina)
• •rIJlIlmlilll.
176
Choroid 13
Glaucoma
Lens
Palhology by alIa Cornea
and Iypa of affection
Orb.t 1
Optic nerve }BJBEBB!IIlflI77
Other 3
Congenital anomalie, 111
Insulhcient diagnOSis.~~~~~l6~3~_+---+_-+-_I---4
o 20 40 60 80 100 120 140 160 180
number of cases
Figure 2. Causes of visual impairment (Population

N=514)
Retinitis pigmentosa
12.6
Albinism
Macular degeneration 24.98
Diabetic retinopathy -INLB

o Faye (1970)
Retrolental fibroplasia
%
0 5 10 15 20 25 30
Figure 3. Main pathologies of the retina and comparison

with the study of Faye (1970). Retinal patholo-
gies are seen in 34.24% of cases in the present
study and in 52.12% of cases in Faye's study.
354
0.1234 = 6/48 (20/160), while mean acuity after refraction in

the low vision clinic is X = 0.1534 = 6/38 (20/125) (T = 4.35
p<0.001 N=299). As mentioned by Faye, the first thing to do
for visually impaired people is refraction.
14%
Spherical ametropia 60%
Astigmatism
12%
50%
10%
40%
8%
30·X
6%
4% 20%
2% 10%
0% 0%
;., ....
'"
0 '"
~
0
0
0
0 0 0
'"0
0
Figure 4. Refractive error distributions
The frequency of visual aids prescription is illustrated in

Fig. 5. Fifty-five percent of all cases received at least one
aid, 39% could not obtain any benefit from them or refused
them, while 6% were functional without any visual aid. The
percentage of visual aids prescription is lower than the one
usually reported [5,10]. This difference can be explained by
the fact that the population in the other study was restricted
mostly to a functionally sighted population.
Aids for distance vision are prescribed in 36.4% of all

cases while near vision aids are prescribed in 46% of all cas-
es. The handheld telescope is the most frequent distance aid
prescribed, while the microscope is the most frequent near
vision aid used (Fig. 6).
The mean magnification for distance aids is 6.30X. The

mean effective magnification, derived from the ratio
between visual acuity with and without aid, is 6.38X. The
paired T-test between these two variables is not significant
355
• Prescribed aids
liD Unnecessary
o No enhancement
Figure 5. Frequency of prescription of visual aids. Total

population (N=514)
Bioptics
9%
lemicroscopes
16%
Microscopes
Hand-held telescopes
91% 58%
Mean ma nification: 6.3 X Mean magnification: 5.9X

Figure 6. Type of aids prescribed.
at the 5% level; the null hypothesis has to be accepted

(T=1.91, p=O.58, N=183). In the same way, the mean magnifi-
cation for near vision aids is 5.98X while theoretical magnifi-
cation is 6.083X. The paired T-test is not significant and the
356
null hypothesis has to be accepted (T=O.13, p=0.9, N=218).

These two computations have been done to verify the validity
of the formula M=required VA/VA [11]. In both cases, the
theoretical calculation of the magnification was useful.
The different levels of impairment based on the WHO's

classification [6,12] were determined (Fig. 7). Two main
fields of impairment are considered: level of visual acuity and
the degree of the visual field. The worst level of impairment,
between acuity and field readings of the better eye, deter-
mine the actual level. Results show that 2.3% of the patients
are not eligible for rehabilitation services, 35.6% are visually
impaired without being legally blind, 27.4% of all cases are at
level 2, 14.2% are at level 3, 10.5% are at level 4, 5% at level
5 and 5% at level 9. In levels 1 to 5, more than 50% of all
people are under 40 years old. At level 9, more than half of
the cases are under 10 years old.
0/0
100
o Frequency by level
90 E1iI % of prescribed aids

80 % of associated Impairment
70
SO
50
40
30
20
10
0
LEVEL 0 LEVEL 1 LEVEL 2 LEVEL 3 LEVEL 4 LEVEL 5 LEVEL 9
(N = 12) (N = 183) (N = 141) (N = 73) (N = 54) (N = 2S) (N = 2S)
Figure 7. Frequency distribution by level of impairment.
The percentage of prescribed aids in each level of impair-

ment is also presented (Fig. 7). The rate of attribution of
aids in categories 1 to 3 varies between 65% and 75%. A 10%
rate of optical and electro-optical aids is still found at level
4. According to Genensky's classification of impairment [1,2]
there is a part of the population at level 4 who are function-
ally sighted. These people are probably the ones with a
severe visual field impairment. Another particularity of our
population is the growing occurrence of associated impair-
ment, along with the visual impairment, as shown by the
357
curve of Fig. 7. More details about the clinical management

of the different categories of impairment are presented by
COUTURIER et al.[3].
Facts about the use of the WHO's classification of

impairment in general eye care practice are discussed. There
seems to exist an excessive number of cases in the legally
blind categories (Fig. 8). Methods and charts generally used
to measure visual acuity are not suited for a person who is
visually impaired. Shortcuts for visual acuity measurements
such as "count fingers", "hand movement" and so on, are too
numerous and influence excessively the number of individuals
in levels 4 and 9.
0/0
50
Classlficalion of Impairment according
42 . 4
10 data of entrance report
40
30
20 17 . 7
14 .2
0
LEVEL 9
(N • 1) (N • 73)
50
40 ClaSSification of .mpalrment followIng

354
teftaClion at the ol w vISIon clinic
30
20
(N • t2) (N • 8
13) (N • 14
1)
Figure 8
358
Some entrance reports do not include any quantified

measure of visual acuity. However, acuity is measured
through a low vision routine (Table 1). The study shows that
although 6 eyes were considered as "no light perception", vis-
ual acuity obtained at the clinic was shown to vary between
6/2000 (20/6666) and 6/105 (20/350). Fourteen eyes classified
as "light perception" were shown to have a visual acuity vary-
ing between 6/240 (20/800) and 6/60 (20/200). Four eyes clas-
sified as "light projection" obtained a measurement between
6/540 (20/1800) and 6/18 (20/60). The most frequent unquan-
tified visual acuities are "count fingers" and "undetermined".
In the first case, the visual acuity measured is between 6/240
and 6/30 (20/800 and 20/100), which excludes the value gener-
ally associated with "count finger", 6/360 (20/600). In the
second case, "undetermined" readings were noted through a
visual acuity range of 6/120 to 6/12 (20/400 and 20/40).
Table 1
Comparison of visual acuity readings between entrance report

unquantified data (left) and low vision clinic evaluation data
(right).
No light perception 6 0.003 ?V.A.? 0.057
6/2000? V.A.>6/10S
Light perception 14 0.041 ?V.A?0.1

6/240?V.A?6/60
Light projection 4 0.0417?V.A.?O.OS

6/240?V. A?6/1 20
12 0.0111 ?V.A.?0.333
Hand movement
6/S40?V.A.?6/18
Count fingers 74 0.0417?V.A.?0.2

6/240?V.A.?6/30
Undetermined 74 0.04 76?V .A.?O.S

6/120?V.A.?6/12
In conclusion, the data shown here are varied and numer-

ous. The purpose is to present some epidemiological facts
359
about a young visually impaired population, functionally

sighted and functionally blind, and a view of the services
received by this population in a low vision clinic. The impor-
tance of computation of theoretical magnification in a low
vision routine is stressed. Some facts about the classification
of visual impairment from reports of general eye care practi-
tioners raise the question that too many individuals may be
considered as legally blind. The need for standardized meth-
ods to measure visual acuity in general eye care practice is
brought out. These methods would impact on the quality of
refraction and the appropriateness of the level of impairment
in which the individual is classified.
4. Acknowledgement
The authors are indebted to the Institut Nazareth et Louis-

Braille for permitting access to clinical files.
360
5. References
1. S.M. Genensky: Functional classification system of vis-

ually impaired to replace the legal definition of blind-
ness. Ann. Ophthalmol. 3, 150 (1971)
2. S.M. Genensky: Data concerning the partially sighted

and the functionally blind. J. Visual Impairment &
Blind. 72, 177 (1978)
3. J.A. Couturier, J. Gresset: Visual impairment and dis-

ability: Enhancement and substitution. Paper present-
ed at the International Symposium on Low Vision, Uni-
versity of Waterloo, 1986.
4. Institut National Canadien pour les Aveugles: Rapport

statistique sur la population aveugle au Canada
inscrite a l'Institut National Canadien pour les Aveu-
gles, (Toronto 1984)
5. E.E. Faye: The Low Vision Patient, (Grune & Strat-

ton, New York 1970) pp. 183-200.
6. Organisation Mondiale de la Sante: Classification

Intemationale des Maladies, (OMS, Geneve 1975)
7. A. Sorsby: The Incidence and Causes of Blindness in

England and Wales 1948-1962, Reports on Public
Health and Medical Subjects, No. 114 (HMSO, London
1966)

1972)
9. A. Sorsby, M. Sheridan, G.A. Leary, B. Benjamin:

Vision, visual acuity, and ocular refraction of young
men. Br. Med. J. 1, 1394 (1960)
10. R.E. Schwartz, G. Burnie: Special report: Low vision

centre, Maryland Workshop for the Blind. Optom.
Wkly. 51, ;2685 (1960)
361
1l. E.B. Mehr, A.N. Fried: Low Vision Care. (Professional

Press, Chicago 1975) pp. 81-103
12. World Health Organization. International Classifica-

tion of Impairments, Disabilities and Handicaps, (WHO,
Geneva 1980)
Pathology Characteristics and Optical Correction
of 900 Low Vision Patients
A.R. Hill, A. Cameron
1. Background
When low vision clinics were being established 25 years ago in

the United Kingdom, there was a shortage of trained person-
nel and severe financial limitations. Preliminary screening of
potential candidates for low vision services were introduced
to assess their probable need for this form of low vision care
[1-3].
In this paper we describe the results from 900 partially

sighted patients who were seen for assessment by the low
vision practitioner at Edinburgh Royal Infirmary between
1970 and 1980 because it illustrates how simple screening
procedures may be used to influence the cost effectiveness of
providing a specialist service heavily constrained by limited
manpower and running-cost resources. Although the primary
provision of low vision services in the UK is still funded by
the National Health Service the principles of screening are
equally applicable to private practice, but more particularly
have practical relevance for the provision of such services to
geographically remote or disadvantaged communities.
2. Screening for Low Vision Management
In the provision of health services, screening is usually direct-

ed towards the detection of occult disease or defect by the
rapid application of special tests, examinations or even ques-
tionnaires. The aim is to sort out apparently well persons
who may have a disease from those who do not have a disease
on the assumption that some effective intervention may be
taken to benefit the affected person. In this respect, the
function of screening is to distinguish those individuals who
need no further examination [5]. Where the subsequent
advice or action which can be taken for an individual is limit-
363
ed by the availability of professional expertise or restricted

clinical services, then the screening referral criteria may be
adjusted accordingly in order to maximize the efficiency of
interventive programs. The principal aim of screening in
situations where resources are limited is to identify those
individuals who are most likely to benefit from the subse-
quent action taken. If the screening is to be worthwhile, a
referral criterion has to be determined which optimizes the
costs and benefits of all the factors associated with the out-
come (ASPINALL and HILL [5]).
The adoption of screening principles to identify those vis-

ually disabled individuals most likely to benefit from detailed
low vision assessment will help to maximise the net benefit of
providing the service. This was the approach taken by the
Edinburgh Royal Infirmary when it was realised that the
resources available would only permit a detailed low vision
assessment of less than 30% of those partially sighted
patients attending the hospital each year. Patients were
placed in one of two categories: those patients who would
derive benefit from optical appliances including spectacle
mounted aids, and those patients who would not benefit from
spectacle mounted aids. Only the former patients would
receive a full low vision assessment. The question was, how
to distinguish between the groups.
Since only low vision aids for reading were available, the
two patient groups may be represented graphically by two
frequency distributions as a function of the initial presenting
best near visual acuity with conventional spectacles (Fig. 1).
It was estimated that there was considerable overlap of these
two popUlation groups making selection on the basis of initial
near visual acuity. This difficulty of distinguishing between
two overlapping popUlation distributions is common to all
screening and may be simplified as a 2 X 2 contingency or
decision matrix (Fig. 2). There are four possible outcomes
from the screening test:
1. true positive referrals (i.e. those patients referred

for full assessment who derived benefit from special
aids),
2. true negative referrals (i.e. those patients not

referred for further low vision assessment and who
would not have derived benefit from the service),
3. false positive referrals and

364
4. false negative referrals.
Would benefit
from spectacle
mounted LV A
Would not benefit

from spectacle
Frequency mounted LV A
N8
C
Decreasing near visual acuity
Figure 1. Theoretical representation of the principle of

screening in low vision assessment. A near visu-
al acuity cut-off value C produces misclassifi-
cations of those patients who are likely to ben-
efit and those who are not likely to benefit
from the use of spectacle mounted low vision
aids.
The latter two are misclassifications and represent the two

types of decision error. The designed aim of screening is to
minimize false positive referrals since these are unnecessary
and costly. A cut-off screening criterion such as at point C in
Fig. 1 which gave less than 5% false positive referrals was
therefore used in the Edinburgh low vision service. This, of
course, meant that there were many false negative referrals,
being visually disabled persons who would have derived ben-
efit from spectacle mounted low vision aids but were not
referred for further assessment in the low vision clinic. In a
monophasic screening problem such as this [6] there is a
direct trade-off between false positive and false negative
errors. Hence, there were almost certainly many false neg-
ative referrals in the Edinburgh study. In screening, there is
no right or wrong cut-off criterion. Its choice will depend
upon the relative weights or utilities which a decision maker
places on the two misclassification errors.
365
SCREENING TEST
+VE -VE
Would benefit from FALSE

TRUE
spectacle mounted NEGATIVE
POSITIVE
low vision aids
Would not benefit TRUE

FALSE
from spectacle
POSITIVE NEGATIVE
mounted aids
Figure 2. Contingency matrix showing the four outcomes

when using a single cut-off screening criterion
for distinguishing between visually disabled
patients likely to benefit from spectacle mount-
ed low vision aids and those not likely to ben-
efit.
In the Edinburgh study, no statistics were kept on those

patients not referred for full low vision assessment, so the
number of true negative and false negative referrals are not
known. When combined they accounted for approximately
two-thirds of those patients recognised as being partially
sighted who attended the Eye Hospital annually. Since more
recent studies of the prevalence and incidence of visual dis-
ability show that, in the UK, there may be one-third more
blind persons than are registered as blind and double the num-
ber of partially sighted registered persons [7], then the low
vision service provided at Edinburgh Royal Infirmary for the
South Eastern region of Scotland was a gross underprovision.
By adopting specific screening principles it was possible to
maximise the efficiency of the service with approximately
95% correct referrals. A private practitioner could use a
similar approach for rapidly identifying those patients likely
to be helped by the limited range of special low vision appli-
ances he has at his disposal.
366
3. The Screening Criterion
A few recent studies have been directed towards determining

what simple clinical measures of visual function would best
predict a patient's reported visual disability. For example,
ROSS [8] has shown that contrast sensitivity performance was
well correlated with perceived visual disability as assessed by
questionnaire amongst patients with cataract or moderate
glaucoma.
In the absence of properly evaluated screening criteria, a

practical approach was adopted. Only patients who were
classified as being "partially sighted" were considered eligible
for screening, (i.e. patients whose best spectacle corrected
distance vision was either
1. 6/120 (20/400) - 6/60 (20/200) with full field of vision

or
2. 6/24 (20/80) or less, with moderate contraction of the

visual field, opacities in the ocular media or aphakia
or
3. 6/18 (20/60) or better, with gross visual field defect.)
The near vision capabilities of these patients was then

assessed using a set of Keeler Aspheric lenses available in the
following magnifications: 2X, 3X, 4X, 5X, 6X and 8X. The
lenses could either be placed in a trial frame (or Halberg clip)
or hand held by the patient as a "quizzer". If the near vision
could be improved with one of these lenses to at least N12
(using Times Roman near vision test type), then that patient
was referred for a full low vision assessment.
4. Low Vision Assessment Procedure
For reasons of economy, only one manufacturer's low vision

appliances were available. These were the full Keeler set of
microscopic, telescopic, telemicroscopic and aspheric specta-
cle magnifiers. While we appreciate the limitations of using
appliances from one manufacturer, we believe that the inclu-
sion of additional spectacle mounted aids would not have
367
made any substantial difference to the results presented here.

The range of magnifications available for each of the low
vision aids used are indicated in Table 1. (During the mid
1970s the range of magnification was changed for some of
these aids.) The initial presenting near visual acuities were
determined with a near vision test chart (Times Roman N
type) viewed at a distance of 38 cm with patients wearing
their appropriate near vision spectacle correction. Illumina-
tion was provided by an adjustable tungsten lamp. All
patients were seen within two weeks of referral from the
screening and the following information was recorded: name,
sex, age, pathology, near visual acuity (i.e. initial visual acui-
ty), near visual acuity with best low vision aid (i.e. final visu-
al acuity), type of aid supplied, magnification of low vision
aid supplied, whether monocular or binocular low vision aid,
number of times attending the low vision clinic during the ten
year period.
Table 1
The range of magnifications available for each of the Keeler

vision aid types used in the study. (Note that the available
magnifications changed during the period of the study. X
indicates aids available from mid 1960 to mid 1970, 0 indi-
cates the range of aids available after mid 1970.)
AID TYPE MAGNIFICATION
1.6 2 3 4 5 6 8 10 12 15 20
Spectacle microscope o iii iii iii iii
Aspheric magnifier iii iii iii iii iii iii X

Bar-type telemicroscope iii iii iii iii X X X
Full-field telemicroscope iii iii iii iii iii iii iii
Bioptic unit iii iii 0 0 0

Distance telescope with caps iii
X mid 1')60' s
o mid 1970's
All patients were discharged from the low vision clinic as

soon as a satisfactory solution had been found for them.
368
However, no one was refused a follow up appointment at the

clinic if they experienced difficulties with their aid. A sys-
tem of sending out annual letters to patients was used to
identify those who were no longer using their aid. All
patients attended the clinic on a minimum of two occasions;
once for the initial assessment and once for fitting of the aid.
Patients were given instruction and tuition in the use of the
aid and the final near visual acuity was determined. If a fur-
ther attendance at the low vision clinic was considered desir-
able, this was usually arranged for three months after issue of
the aid.
5. Patient Characteristics
The age distribution of the 900 patients attending the clinic

over the ten year period is shown in Fig. 3. The distribution
is bimodal and resembles that found at other low vision clin-
ics [9]. Few people in their teenage years and early twenties
attend low vision clinics. It is almost certainly a function of
two principal factors. One is that young people with visual
disability tend to have pathological conditions which are fair-
ly stable. The other reason is that these patients find most
low vision aids cosmetically unacceptable. What is not clear
is which of these two factors is the more important. A closer
exam ina tion of the special problems of this age group of visu-
ally impaired is necessary to determine whether any improve-
ment in aid design or special low vision counselling would be
helpful. The difference in the age distributions for males and
females almost exactly matches the population sex differenc-
es in Scotland for the middle of the study period [10]. The
exception is for patients over the age of 85 years, where
there are three times more females than males in the popula-
tion. When patient attendances at the clinic are grouped
according to whether a person is likely to be in full-time edu-
cation (6-20 years), a member of the working population
(21-65 years) or retired (>65 years), it becomes evident that
the major task in a low vision clinic is the management of
visual disability amongst the elderly (Table 2).
The prevalence of the major diseases for three different

age groups of the sample population is summarized in Table 3.
These figures indicate the primary single reason for presenta-
tion at the clinic and do not include multiple pathologies.
369
Table 2
Proportion of patients attending the low vision clinic as a

function of three population age groups.
6-20 years 3.3 3.2 6.5

21-65 years 17.6 19.2 36.8
66-95 years 21.7 35.0 56.7
TotaLs 42.6 57.4 100.0
From a total of 33 diseases seen in the clinic, by far the most

prevalent condition amongst this group of patients is age
related maculopathy. It accounts for 39.3% of our patients
and closely matches the value of 42.0% for patients attending
the Moorfields Low Vision Clinic in London [11]. The impor-
tance of this largely untreatable condition as a cause of low
vision can be appreciated when it is realised that it accounts
for 63.7% of patients seen over the age of 65 years. Unless
considerable strides are made during the next decade in the
treatment of this condition, it is likely to place an even
greater demand on low vision services as life expectancy in
the population increases. The second major ocular condition
amongst patients over the age of 65 presenting at the clinic
was myopic degeneration with the unusually high prevalence
of <60%. This is almost certainly a consequence of the cho-
sen screening cut-off criterion and is therefore unrepresenta-
tive of the importance of this pathology as a cause of visual
handicap in an unselected popUlation of the same age. The
third most prevalent condition seen at the clinic was glauco-
ma, but this was seen with only one tenth the frequency of
age related maculopathy.
For the majority of the adult popUlation between the ages

21 and 65 years, the major cause of visual disability in the
western world is now diabetes. It has become the major
cause of blindness in the United States [12]. For this same
age group of the working population, it was also one of the
370
100
90 FEMALES
MALES
80
70
Number
60
r-'I
I
01 I I
cases I I
I I
I I
50 I L_.,
I I
. . _J I
I
40 I
I
I
.. _oJ I
r-~ I
30 _.J I
L_
20
I
I
L_
10
- ;; '"- '"'" -... '"... - .'" -... .'"

0
~
co
~ '"
co
~ '"01 ~
'" ...'"0 ... '"
0 Ol 01 Ol
I I I I I I I I I I I I I I I I I I I
'" ;; '" " co co co

'" Ol
'"01 OJ CD
'" '" '" '" '" '"

0 0 Q Q 01 Q Q Q 01
Age groups (yeara)
Figure 3. Age distribution as a function of sex for the

sample of 900 low vision patients.
two major causes of visual handicap seen in the low vision

clinic during the period of study.
Among the younger age group (6-20 years), congenital

cataract, optic atrophy and albinism were the major diseases.
For congenital cataract recent developments in its manage-
ment by early surgery and contact lenses are beginning to
reduce the significance of this condition as a major reason for
visual disability in the young [13].
The major causes of visual disability in the western world

are now primarily the untreatable diseases. In particular, age
related maculopathy, diabetic retinopathy and glaucoma. The
371
Table 3
Prevalence (%) of the major diseases of screened patients

attending the low vision clinic for three age groups and as a
function of all presenting cases.
OcuLar condition Age in years
6-20 21-65 66-95 6-95
n=59 n=3:51 n=510 n=900
Age reLateci macuLopathy 9.7 63.7 39.3
r'lyopi c degeneration 1.7 14.':; 6u.8 9.0
Diaoetic retinopathy 13.9 6.1 8.6
CorneaL conditions 3.4 8.8 4.1 5.8
GLaucoma 1.7 4.5 6.7 5.6
Disseminated choroiuitis 5.1 6.9 3.5 4.9
HacuLar dystrophy 8.5 12.4 1.2 4.7
Cataract 22.0 1.0 3.3 4.6
Optic atropny 15.3 7.S 4.0
Vascular concitions 4.2 3.9 3.0
Hereait'r) !';acuLar dystrophies 6.7 4.2 2.1
ALbinism 13.5 1.S
Chronic U\I('; tis 0.1 1.3
RetinaL detacn"ent 1.7 0.1 1 .3 1•2
17 othe r conaitions each < 1.0 of totuL
prominence of myopic degeneration and corneal conditions in

the list of conditions in the present sample (Table 3) is a con-
sequence of the special screening criterion adopted for this
study.
6. Near Visual Acuity and Pathology
The distribution of near visual acuity with best spectacle cor-

rection at the time of initial examination is shown in Fig. 4.
When the data are adjusted for the different proportions of
males to females in the sample population (i.e. 1:1.35), the
distribution of initial near visual acuities is identical for each
sex. The absence of very low initial near visual acuities is a
direct consequence of the screening criteria adopted for
referring patients to the low vision ciinic.
The partitioning of these initial near visual acuities as a

function of their disease is shown in Table 4 for the ten most
prevalent conditions. By combining category N8 with NIO
372
FEMALES
MALES
120
100
80
Number .... -
of
cases
-, ,---I
I I
60 I I
L_.J
-,
40
6 8 10 12 14 18 24 36 48
Initial near visual acuity

(Times Roman point size)
Figure 4. Distribution of initial near visual acuities with

best spectacle corrections at a reading distance
of 25 cm. (All 900 patients had been pre-
screened and were able to achieve at least N12
with an aspheric hyperocular not exceeding
32.00 D.)
and N36 with N48 it is possible to conduct a chi square analy-

sis on these data. This analysis showed there to be a signifi-
cant difference in the proportion of initial n~ar visual acui-
ties across the different disease conditions, (X = 91.99 for df
= 45, p< 0.0005). However, by removal of the "myopic degen-
eration", group the chi square analysis shows there is no sig-
nificant difference in the initial near visual acuities across
the remaining groups. This effect is almost certainly a con-
sequence of the preliminary screening. While the visual acui-
ty loss in myopic degeneration is undoubtedly less, on aver-
373
age, than tha t produced by the other identified ocular

conditions, the more severe cases of very low visual acuity
have been excluded from the sample. The effect has been to
truncate the sample distribution slightly (Fig. 4) with the
result that differences between those conditions responsible
for very poor vision have been masked. Although the use of
preliminary screening for referring patients to the low vision
clinic has produced an effect in which the near visual acuity
is independent of their disease (with the exception of myopic
degeneration), it cannot be assumed that the quality of vision
experienced by patients is also independent of the diseases.
This effect illustrates the inadequacy of standard reading
type for character visual function. Studies on the dependence
of visual acuity on illumination, contrast and distance have
shown large inter and intra-individual differences amongst
partially sighted persons which masked systematic differenc-
es due to pathology [14]. Furthermore, in a study on rehabili-
tation of patients with age related maculopathy, LOSHIN and
WHITE [15] have shown that the preferred eye for reading
appeared to be related to the peak of the contrast sensitivity
function rather than to Snellen or resolution acuity. Low
vision practitioners, therefore, should become acquainted
with the limitations of standard high contrast reading type if
they are to appreciate the wide differences in quality of
vision experienced by their partially sighted patients.
Table 4
Distribution of initial near visual acuities with best spectacle

correction as a function of their disease. (Near acuity in
Times Roman point size at a reading distance of 25 cm)
PathoLogy InitiaL near visuaL acuity

N8 N10 N12 N14 N18 N24 N36 N48
Age reLated macuLopathy 4 31 53 51 70 85 53 10

['Iyopic degeneration 10 11 28 11 13 4 3 1
Diabetic retinopathy 0 5 13 10 21 19 5 4
CorneaL cond i t ions 3 4 15 9 8 10 3 0
GLaucoma 2 6 11 8 7 11 3 2
Disseminated choroiditis 1 5 4 5 3 13 8 0
hacuLar oystrophies 2 3 8 1 10 12 6 U
Cataract 3 6 13 13 22 19 2 0
Optic atrophy U 4 4 7 7 8 4 2
Vascular conditions 1 0 6 2 7 11 5 2
374
7. Low Vision Performance of Screened Patients
Less than 5% of the patients referred to the low vision clinic

following screening could be helped by a spectacle mounted
low vision aid from the limited range of aids available in the
clinic (Table 1). Only the results of the 900 patients supplied
with low vision aids are reported here and the distribution of
near visual acuities obtained with the aid prescribed is shown
in Fig. S. All except 18 patients (i.e. 2%) achieved a near vis-
ual acuity of N8 or better. When these results are examined
according to the disease condition, there was little practical
difference in the achieved performance, although a greater
proportion of the poorer final acuities is found among
patients with macular disturbance and consequent central vis-
ual field loss (Table 5). These results illustrate the high
validity of the chosen screening criterion of N12 or better
obtained with the simple series of aspheric lenses. Further-
more, these results confirm that the limited range of aids
available in the clinic has placed no practical constraint on
the final visual acuity.
375
240
FEMALES
220 MALES
200
Number
of
cases 180 --,
160
1--
L_
40
20
o +--,--r--r--f
5 6 8 10
Final near visual acuity
(Times Roman point size)
Figure 5. Distribution of near visual acuity with pre-

scribed spectacle mounted low vision aids for
males and females (note collapsed ordinate
scale).
376
Table 5
Distribution of final near visual acuity obtained with pre-

scribed low vision aid as a function of disease condition.
PathoLogy Near visuaL acuity with LVA
rJ5 N6 N8 N10
Age reLated macuLopathy 149 155 46 6
I'lyopi c degeneration 35 37 8 1
Diabetic retinopathy 41 27 8 1
CorneaL conditions 27 2:5 !'. 1
GLaucorfla 25 19 6 (I
Disseminated choroiditis 17 24 3 0
MacuLar dystrophies 19 20 1 2
Cataract 26 12 2 1
Optic atrophy 13 16 7 0
VascuLar conditions 15 14 4 1
8. Low Vision Management
The financial constraints meant it was only possible to issue

one spectacle mounted aid to each patient. The distribution
of magnifications used for the 900 patients is shown in Fig. 6
and how these magnifications were shared amongst the dif-
ferent spectacle mounted aid types is given in Table 6. It will
be seen that the most frequently used aid was a bar-type tel-
emicroscope of 4X magnification. Of particular interest is
the observation that 89% of patients were satisfactorily
helped from a range of only four magnifications (i.e. 3X, 4X,
5X and 6X) in each of three types of aid. Two of these aid
types were telemicroscopes (i.e. the bar-type having a fixed
focus distance of 16 cm and the full-field type), both of which
may be used as a binocular unit for magnifications of 4X or
less and 3X or less respectively. A total of 269 (i.e. 29.9%)
binocular units were issued to patients, the proportion of
binocular to monocular aids being greater for the better final
visual acuities.
The magnifications required expressed as a function of

the ten major presenting diseases are given in Table 7. When
the extreme magnification categories are combined to permit
a chi square analysis, a significant difference in the propor-
377
Number
of
cases
350
300
250
200
-
150
100
-I--
50
o -
2 3 4 5 6 8 10 15 20 Magnification
~~:: ~ ] Percentage of
total sample
IE-- 89% ~ distribution
IE-- 98% ~
Figure 6. Distribution of magnifications of spectacle

mounted low vision aids prescribed for the sam-
ple of 900 pre-screened low vision patients.
tior of magnifications used across diseases is demonstrated

(X = 115.57 for df = 36, p< 0.005). However, when the single
group of "myopic degeneration" is removed, the distribution
of the different magnifications is independent of the disease.
On average, the patients with myopic degeneration require
lower magnification than those presenting with other diseas-
es. These findings agree with the results presented in Table 4
which show the initial near visual acuities in myopic degener-
ation to be significantly better than the acuities with other
diseases.
In low vision practice patients presenting with severely

impaired visual acuities achieve less well than those present-
378
Table 6
The frequency of use of different magnifications for each of

the low vision aids supplied by the clinic.
LOW VISION AID MAGNIFICATION

2 3 4 5 6 8 10 15 20
Spectacle microscope 1 1
Aspheric magnifier 1 5 15 30 44 2
.......
Bar-type telemicroscope 6186 315 150 ;40 24 8
Full-field telemicroscope 9 15 9 21 16
Bioptic unit 1
Distance telescope with caps 1
Table 7
Distribution of magnifications required by patients presenting

with one of the ten major diseases from the sample of 900.
Pathology Magnification
3 4 6 8 10 15 20
Age related ~aculopathy 1 61 126 72 47 44 3 0 0

Nyopic degeneration 4 43 25 5 0 2 2 0 0
Diabetic retinopathy 0 14 32 19 8 4 0 0 0
Corneal condition5 0 19 15 8 5 5 0 0 0
GLaucoma 0 10 21 11 2 2 0 0 0
Disseminated 0 6 16 9 7 5 1 0 0
Macular dystrophies 0 5 18 10 6 2 1 0 0
Cataract 0 8 17 11 1 4 0 0 0
Optic atrophy 0 4 13 7 3 6 1 1
VascuLar conditions 0 3 15 5 6 3 0 U
ing with slightly better acuities. Despite the exceptionally

poor near visual acuities being excluded for screening from
the population sample under study this positive relationship
between initial and final near visual acuities is still present.
This correlation of r = +0.468 (p< 0.0001) is illustrated in the
frequency plot of Fig. 7. It is unlikely that this result is due
to inadequate magnifications being available among the trial
379
aids used for assessment because the higher magnifications of

15X and 20X were very infrequently used. It confirms the
clinical finding that most patients with partial sight will sac-
rifice a little resolution for the sake of gaining a little in the
field of view when reading [16]. This effect is further con-
firmed by the positive correlation of r = +0.460 (p< 0.0001)
which exists between the magnification of the prescribed low
vision aid and the patient's final near visual acuity (Fig. 8).
Such findings add support to the hypothesis that the task of
reading is performed by contextual shape recognition of
words rather than by individual character resolution.
Although LEGGE et ale [17] have shown that reading requires
a field of only four characters and a resolution of only 2
cycles/character, reading speed can be uncomfortable and
unacceptably slow at this level of visual degradation. Very
few of the patients in our sample had vision as poor as this,
but the results suggest that many patients with low vision
prefer a slightly increased field of vision at the expense of a
little resolution. Thirty years ago KELLER [18] suggested
that linear magnification calculations from near visual acui-
ties could be used to predict the magnification needed to
improve vision from a known level to a stated higher level.
For this purpose he produced his "A Series" of letter charts in
which changes in vision were expressed in a constant geome-
tric ratio. As a test of the validity of this principle we exam-
ined the relation between the initial near visual acuity and
the magnification required by a patient in their prescribed
low vision aid. The relation is shown in the scattergram of
Fig. 9. The positive correlation between these two measures
is to be expected. But there is a definite lack of correspon-
dence between the prescribed magnifications and the expect-
ed theoretical values using the principle of linear magnifica-
tion required to provide a corrected near visual acuity of N6.
On average, most patients require a higher magnification
than would be expected on the basis of simple theoretical
geometric principles. This could be due to the low vision
practitioner aiming to achieve a better visual acuity than N6.
Nevertheless, this does not explain the observation that
patients with the better initial near acuity appear to require
a much higher than theoretically expected magnification in
contrast with those patients presenting with the poorer acui-
ties. It is possible that the latter patients have had more
opportunity to learn to cope with their visual disability by
enhancing their ability to extract information from a very
poor quality visual image. Hence, while the theoretical pre-
dictions of linear magnification from visual acuity measures
380
may be used as an initial guide, they should only be used as a

very tentative approximation to what magnification is
required by a patient.
Final near
visual acuity
-,--r-,--r-,--r-'--r-'--I
10 I I I I 12131313131
_J __ L_J __ L_J __ L_J __ L_J __ I
I I I I I I I
8 I I I I 6 I 3 I 14 I 34 I 24 I 8 I
-,--r-,--r-,--r-,--r-'--I
6 I 17114155153187110416019 I
_J __ L_J __ L_J __ L_~ __ L_J __ I
I I I I I I 2 I 3 I
5 29 173 1107 I 67 I 71 1 46 I 1 I I
5 6 8 10 12 14 18 24 36 48
Initial near visual acuity
Pearson's r = 0.468 (p < 0.0001)
Figure 7. Relation between initial near visual acuity with

best spectacle correction at presentation and
final near visual acuity with prescribed low
vision aid (acuity measured in Times Roman
point size). The cell numbers indicate frequen-
cies from a sample of 900.
381
-,--r-,--r-,--r-,--r-,
10 1 1 3 1 2 1 4 1 5 1 1 1 1
_~ __ L_~ __ L_~ __ L_~ __ L_~
1 1 1 1 1 1 1 1 1
8 1 4 1 14 1 21 19 1 28 1 2 1 1 1 1
-,--r-,--r-,--r-,--r-,
6 2148114319115614216111 I
_~ __ L_~ __ L_~ __ L_J __ L_~
1 1 1 1
5 2 1 144 1 176 1 60 12 9 2 1
2 3 4 5 6 8 10 15 20
Magnification
Pearson's r = 0.460 (p < 0.000 1)
Figure 8. Relation between final near visual acuity and

magnification of low vision aid. The cell num-
bers indicate frequencies from a sample of 900.
382
Prescribed
LVA
magnification
-,--r-,--r-,--r-,--r-,
20 I I I I I I I I 1 I
_~ __ L_~ __ L_~ __ L_~ __ L_~
I I I
15 I I I I I I I 1 I I
-,--r-,--r-,--r-,--r-,
10 I I I I 1 I I 4 I 5 I
_~ __ L_~ __ L_~ __ L_~ __
I I I I I I I
I I I 4 I 4 I 6 I 27 I 37
-,--r-,--r-'--r-.."..===;f=="'l
6 I I 2 I 5 I 8 I 12 I 43
_~ __ L_~ __ L_~ __ L_
I I I I I I I I
I I 6 I 9 I 17 I 52 I 54 I 29 I 7 I
-,--r-,--r-'-- --r-,
4 I 8 I 24 I 83 I 65 I 89 8 I 6 I
_ ~ __ L _ ~ __ L _ -t.===;l'-==" __ L _ J
I I
I 27 I 51
-'--r-
I 1
10 12 14 18 24 36 48
Initial near visual acuity (Times Roman point size)
Pearson's r 0.619 (p< 0.0001)
Figure 9. Relation between initial near visual acuity

(spectacle correction) and prescribed low vision
aid (LV A) magnification. (Filled squares indi-
cate theoretically expected magnification to
provide a near visual acuity of N6.) The cell
numbers indicate frequencies from a sample of
900.
9. Some Practical Points
The limited range of low VISIOn aids used in this study was
based primarily on economic grounds.
For the ten year period of the study, the number of occa-
sions on which each patient attended the low vision clinic
never exceeded nine. This was unrelated to the initial near
visual acuity, the magnification of the prescribed aid or the
383
age of the patient. It was partly a function of the stability of

a pathology, but many patients with stable ocular conditions
made multiple attendances saying they were having difficul-
ties in using their vision aids. There is a need in low vision
management to identify those patients likely to need addi-
tional tuition and help in using special optical aids for read-
ing. Ideally, low vision clinics should be attached to rehabili-
tation units for the visually disabled and, in some instances,
special consideration should be given to the possibility of
domiciliary help in the rearrangement of home lighting to
improve reading performance [19].
In the ten year period of this study, most patients {i.e.

mode} attended the clinic on three occasions and 74% of
patients made no more than four visits. It is believed that the
number of visits made by each patient was kept to a minimum
by sending letters to patients enquiring whether they were
having any difficulties with their aid and required a further
assessment, or whether they no longer required the aid. By
keeping these follow up visits to a minimum, more new
patients were able to be seen in the clinic, thereby maximiz-
ing its availability to as many patients as possible.
10. Conclusions
This longitudinal study of a limited low vision service operat-

ed by one practitioner has demonstrated the positive benefits
from adopting a preliminary screening program to identify
those visually disabled patients who are most likely to benefit
from a full low vision assessment. The chosen screening
method and cut-off criterion was shown to be valid for the
limited service provision available. Others who wish to adopt
this approach need to select a different cut-off criterion
which more closely reflects the relative importance they
place on false positive and false negative misclassification
referrals. Methods for quantifying these beliefs and using
them rationally for making decisions about setting up new
clinical services or modifying the provision of existing servi-
ces have been described with examples by ASPINALL and
HILL [20]. The principle of preliminary screening is likely to
have value in many situations where the provision of a low
vision service is very limited such as in geographically remote
areas or in underdeveloped countries. It also has potential for
384
the private practitioner who wishes to offer a basic low vision

service to those of his patients most likely to benefit, but
with a minimal capital expenditure on a trial set of low vision
aids. Although only one manufacturer's range of appliances
(Keeler) was used in our study, there are many alternatives
now available and it is for each practitioner to decide which
appliances he feels happiest using. Whatever the choice, it is
clear from our study that a useful practical and efficient low
vision service can be provided with a limited range of specta-
cle mounted aids in addition to a few simple hand and stand
magnifiers.
11. References
1. Provision of Medical and Surgical Applicances,

(H.M.S.O., London 1971)
2. J. Silver, E. Gould, J. Thomsitt: The provision of low

vision aids to the visually handicapped. Trans.
3. E. Gunstensen: Visual aids for the partially sighted.

Br. J. Ophthalmol. 44, 672 (1960)
4. C.R.S. Jackson, E.B. Petrie: Partial Sightedness: An

Investigation into its Incidence and Causes; its Social,
Occupational and Educational Implications, (W.H. Ross
Foundation (Scotland), Edinburgh 1969)
5. P.A. Aspinall, A.R. Hill: Is screening worthwhile? In

Progress in Child Health, Vol. 1, ed. by J.A. Macfar-
lane (Churchill Livingstone, Edinburgh 1984) pp.
243-257.
6. C.R. Hart: Screening in General Practice (Churchill

Livingstone, Edinburgh 1975)
7. Initial Demographic Study - A Review of Available

Data on the Visually Disabled Population (Royal
National Institute for the Blind, London 1985)
8. J.E. Ross: The Functional Effects of Visual Disorder

(D. Phil thesis, Oxford 1983)
385
9. H.G. Robbins: The low vision patient of tomorrow.

Aust. J. Optom. 61, 54 (1978)
10. Annual Abstract of Statistics, Central Government

Statistical Office, Report No. 112 (H.M.S.O., London
1975)
11. J .H. Silver: Low vision aids in the management of vis-

ual handicap. Br. J. Physiol. Opt. 31,47 (1976)
12. E.E. Faye: Clinical Low Vision, 2nd ed. (Little, Brown
and Co., Boston 1984)
13. W.O.G. Taylor: ABC of diabetes: Organization of dia-

betic care. Br. Med. J. 285, 1210 (1982)
14. I. Lie: Relation of visual acuity to illumination, con-

trast and distance in the partially sighted. Am. J.


17. G.E. Legge, D.E. Pelli, G.S. Rubin, M.M. Schleske:

Psychophysics of reading ll. Low Vision. Vision Res.
25, 253 (1985)
18. C.H. Keller: Visual aids for the pathological eye.

Trans. Ophthalmol. Soc. U.K. 76, 605 (1956)
19. T .R. Cullinan, J .H. Silver, E.S. Gould, D. Irvine: Visual

disability and home lighting. Lancet 1,642 (1979)
20. P.A. Aspinall, A.R. Hill: Clinical inferences and deci-

sions. m. Utilities and the Bayesian decision model.
Ophthalmic Physiol. Opt. 4, 251 (1984)
Assessment of Vision of Deaf-Blind Persons: A Review
Lea Hyvarinen
1. Introduction
The deaf-blind population is a small group of people with spe-

cial needs in visual assessment. Although small in number,
this group shows great variation in its clinical picture. The
age of onset of the two impairments causes one important
variable. Since the deaf-blind person may be born deaf, hard
of hearing or with normal hearing; blind, partially sighted or
normally sighted, there are nine different groups of deaf-
blind individuals with respect to very early development. A
later loss of sensory functions may occur at any age. The
effect of the impairments on the development of different
functions is different if the function is lost during the first
year of life, during pre-school age before or after the devel-
opment of spoken language, at school age, or in senescence.
The deaf-blind population can be divided into three dis-

tinct groups: (1) patients with good language skills, spoken
language, American Sign Language (ASL) or any other sign
language; (2) patients with limited language skills; and (3)
patients with no language. This review covers assessment of
the vision of deaf-blind patients with good or limited lan-
guage skills.
If the deaf-blind patient uses auditory communication or

purely tactile communication, tactile signing, the assessment
is very similar to the assessment of a hearing patient. Prob-
lems arise whenever visual sign language is used, because we
use the same modality, vision, for communication when
assessing its function. The limited vision causes technical
problems in communication and at the same time there are
the usual problems related to interpretation from a spoken
language to a sign language. Both the technical and the com-
munication problems cause changes in the pattern of informa-
tion exchange and require some training before communica-
tion is fluent in the patient-interpreter-examiner triad.
387
2. Adaptation Problems
Luminance level and reflection from glossy surfaces is a com-

mon problem. Narrow range of visual adaptation is especially
frequent among patients with Usher syndrome. The patient
should be at the luminance level used during the examination
at least half an hour before the assessment starts. Otherwise
the vision of the patient is gradually adapting to the lumi-
nance level, which causes variation in the test results. This is
often experienced in laboratories when visual field is meas-
ured. If the patient comes to the laboratory on a bright day
and the measurement is started immediately, the slow
increase in the size of the isopters can be documented. If the
technician is unaware of the phenomenon he might interpret
it as a sign of poor co-operation. An automatic perimeter
will, of course, be unable to handle this situation. Hysteric
constriction of the visual field during the examination is well
known. The gradually increasing visual field of retinitis pig-
mentosa patients should become as well known.
At the beginning of the examination the first questions

should be: is the light level comfortable? (It would be best to
allow the patient to adjust the lighting to be comfortable by
using an attenuator or a dimmer switch.) Can you see the
interpreter well enough? Is the background good or does it
bother you? (A dark screen behind the interpreter improves
visibility of the signs in many cases.) Do any of the shiny
instruments cause dazzle? It is advisable never to use a
white coat or clothing with contrasting stripes or checks dur-
ing the visual assessment, because they may effectively dis-
turb the vision of many visually impaired people [sequence
1,2]*.
3. Communication Field
The distance at which the interpreter should sit and the size
of the communication field at that distance should be
assessed before any other measurements are done.
The space usually used for signing is approximately 50-60

cm in diameter. One degree of angle of vision is equal to one
centimeter at a distance of half a meter (57 cm) or two feet.
388
Thus a visual field of, for example, 15 degrees is 15 cm at

half a meter and 30 cm at 115 cm distance, which is a normal
communication distance. If a patient has a limited visual
field the signs have to be made smaller or the interpreter has
to step back until the signs are seen comfortably. If the deaf
person is used to lip reading, he may have difficulty in lip
reading at the greater distance. Attention should be paid to
good contrast on the interpreter's face by using focal light
(contrast pen to accentuate the lips would be helpful).
When the patient has chosen the communication distance,

the size of the communication field at that distance should be
measured by asking the deaf person to look at the interpret-
er's nose and respond to the appearance of the object used to
measure confrontation field around the interpreter's face
[sequence 3]*. When the interpreter has experienced the
measurement of the visual field, he is less likely to use too
large signs [sequence 4]*. If the interpreter has to move clos-
er to the patient during the assessment the examiner should
remember to point out that the communication field decreas-
es accordingly. If the visual field is very small, it is easier to
ask the patient to describe what he sees without moving his
eyes [sequence 5]*.
4. Communication and Spectacle Correction
Since the interpreter is at a certain distance, some patients

will need the corresponding refractive correction whenever
looking at the interpreter. If decreased accommodation or
aphakia require near correction at the communication dis-
tance, it should be easily available so that it can be quickly
put in place for communication. This is especially important
if the patient is strongly dependent on lip reading or fingers-
pelling, both of which require more accurate correction than
the big signs.
389
5. Communication During Subjective Refraction
Measurement of objective refraction should be done during a

preliminary examination because the dazzle caused by the
retinoscope may last for a prolonged period of time.
During the sUbjective refraction the patient has to know

which two moments to compare. Since the patient has to
keep looking at the chart he has to receive the information as
tactile information. The interpreter may move closer and
transfer the information or the examiner can provide the
information, for example by putting one finger in the palm of
the hand of the patient during the first alternative and two
fingers during the other alternative and then the patient may
answer "first" or "second". The whole procedure should be
explained before the process is started. Quite commonly a
patient may need some training before sUbjective refraction
can be measured reliably. The question "which lens is better"
is a difficult question to most visually impaired persons. It
seems to be even more difficult to many deaf visually
impaired patients. When neither of the two alternative
images gives clear vision it is hard to judge which of the two
poor quality images is better. The use of lenses strong
enough to create a more obvious difference in image quality
helps the patient to notice what he is asked to see.
6. Observation of the Patient
Although information is mainly received from the interpreter

one should pay attention to the mimicry and body language
that deaf persons use more often than hearing patients
[sequence 6]* even if they are fluent in ASL.
Observation of the patient's reactions is crucial whenever

dealing with a patient who has limited communication skills
or who is limited in communication because of fear or anxie-
ty. One has to repeat the questions and check whether the
answers and reactions contain similar inf orma tion throughout
the assessment. Quite often the impression one gets during
the first examination is later proven to be wrong. On the
other hand the patient may suddenly understand what the
examiner means with the questions and become very helpful
in explaining how he experiences the question [sequence 7]*.
390
7. How to Ask the Questions
Since the patient has to look to both the interpreter and the
test, questions should be phrased to fit this framework of
communication. With deaf persons one should usually
describe the general features of a task first and then ask a
specific question. Thus it is advisable to show the patient the
test, then guide his attention to the interpreter for more
information about the test and when that is clear, ask the
question. Sometimes well functioning deaf persons are accus-
tomed to guessing the later part of a sentence and do not
watch the interpreter long enough. Then the question must
be repeated until received correctly by the patient [sequence
8]*. Patients with small visual field may not notice that the
interpreter has started signing. In those cases the examiner
has to interrupt the communication (which actually is not
communication when the receiving part is missing) and guide
the patient to observe the interpreter.
Typically simultaneous interpretation starts as soon as a

speaker has said something and it is interpreted carefully to
correspond to the spoken language as closely as possible.
When discussing vision with visually impaired deaf persons,
one has to remember that specific terms do not have their
counterparts in sign language and mere fingerspelling of these
strange words will not make them easier to understand. Thus
one has to avoid difficult expressions, ask the interpreter
whether it is possible to transcribe the term or whether it
should be described.
Spoken language can linger around a question and may be

difficult to reorganize in sign language. "Economy of lan-
guage" is the term used for short, simple sentences that are
reasonably easy to interpret [Addendum 1]. Questions can be
given to the patient prior to the examination as home work
and they can be used as information material for special edu-
cators and families. If a question is difficult to interpret one
may ask the interpreter to speak aloud what he is signing, but
the contents if said exactly as signed may sound unintelligi-
ble. Whenever a question seems to be problematic it is advi-
sable first to repeat the question. If that does not help, reph-
rase the question. If rephrasing is not understood, it is best
to ask the patient what he thinks the question is all about. It
may be, for example, that the interpreter is using a sign with
both a concrete and an abstract meaning and the patient
knows only the concrete meaning. The whole question may
391
then be redone using concepts with which the patient is

familiar.
8. Measurement of the Different Visual Functions
Measurement of visual acuity seldom causes any problems. It

is the measurement that nearly all visually impaired deaf per-
sons have experienced a number of times. Measurement of
grating acuity is less often done and therefore may require
some training with the patient before the answers are relia-
ble.
Measurement of visual field for detailed analysis is

always a demanding task. The procedure has to be explained
in detail before the patient's head is positioned in the perime-
ter because at that time visual communication is lost. During
the measurement tactile information can be used to guide the
patient's fixation back to the centre, i.e. the patient can be
told that whenever the fixation moves from the fixation point
the examiner will tap him on the knee to remind him about
careful fixation.
Measurement of reading field, i.e. the number of letters

visible at one time without moving the eyes, is often difficult
when examining well functioning hearing persons and it is
more difficult when examining deaf persons who are used to
being visually alert and moving their eyes more than hearing
individuals. Comparison of the reading field when using
printed text and optical magnification to that when using a
CCTV is important in many retinal degenerations. Lists of
common words of increasing length may make the measure-
ment easier [sequence 9]*.
Contrast sensitivity is measured in the same way as in

hearing patients. There have been no notable difficulties in
the measurement of contrast sensitivity using low contrast
symbols (LH-5 contrast test). The experience of the use of
the different grating tests is still limited. In the experimen-
tal laboratory situation deaf persons do respond much like
hearing persons. It is possible to measure contrast sensitivity
even at low luminance levels in young Usher patients after
short training. The printed tests are often of limited value
because they do not have test material for patients with
392
severe loss of contrast sensitivity, common in the visually

impaired [sequence 10, 11, 1Z]*.
Visual adaptation is often affected early in Usher's syn-

drome and should therefore be carefully assessed both in the
examining room and outside in different lighting conditions.
The numerous different absorptive lenses, both nonphotochro-
matic and photochromatic may require several examinations
before a given patient has all the lenses he needs to function
at different luminance levels with his limited range of adap-
tation. We have used skiing camps to test absorptive lenses
for visually impaired deaf persons with limited communica-
tion. Other concrete situations could be similarly chosen to
give the patient experience with different absorptive lenses.
Color vision should be assessed for daily tasks. Rather

simple test material is sufficient if completed with anamnes-
tic information and advice to the patient [sequence 13]*.
9. Visual Illusions
One of the most difficult topics to discuss is the visual illu-

sions often seen by Usher syndrome patients. So far there is
no good description of the different illusions in sign language.
I have been reluctant to make information available in order
not to affect the information that should be gathered among
the RP population. The descriptions of the two Usher
patients [sequence 14, 15, 16]* are different and demonstrate
well how difficult it is to convey information related to illu-
sions.
10. Summary
Assessment of the vision of deaf-blind persons is no more dif-

ficult than any other time consuming clinical task. It
requires some experience in the communication of the deaf
popUlation and in the differences between sign language and
spoken language. Once the doctor has learned to work with
the interpreter, and the patients know that their effort is
appreciated, it may be possible to make a nearly complete
393
assessment of the vision of a patient who previously has been

able to answer only when tested for visual acuity [sequence
17, 18]* • Before we can expect pa tients to give proper
answers we must learn to ask the proper questions.
*The video material mentioned in this paper [sequences]

is available at the Helen Keller National Center, 111 Middle
Neck Road, Sands Point, N.Y. 11050. The video material was
produced in collaboration with Bernadette Wynne, M.A. and
Margaret Ransom, C.S.C. The project was sponsored by the
Helen Keller National Center and the Smith-Kettlewell Eye
Research Foundation.
11. Addendum 1: Questions About Vision
From the book Hyvarinen, L.: "Eyes and Vision", Vistest 1986,
with permission
Is the light here good, comfortable?
Is the interpreter at good distance? Should he be further

away? Closer?
Do you see his whole face without moving your eyes?
Now keep looking at his nose and tell when this white
ball becomes visible.
Can you see the signs this big?
Can you see fingerspelling?
Do you lipread?
Would it help if the interpreter had used lipstick?
Do you see well in dim light?
If you wait in dim light, do you start seeing better,

there, in dim light?
How long a time do you have to wait?

394
When you go in a dimly lit room, does your field of

vision change? Get smaller?
When sun shines on your face does it hurt your eyes?
When the sun has dazzled your eyes, how long does it
take, before you see as usual?
Do you use sun glasses?
Do you have special sun glasses? Several different

pairs?
Do you bump into people and things even when you have
your sun glasses on?
When you walk on the street or sit in a car, do the head-

lights bother you?
When you travel on buses and trains, is it difficult to

find your way because of too little light?
Is it difficult to go from a well lit place into a dark

place?
When you go into a shop, do you have to stop at the door

and wait before you start seeing in the shop?
Can you sit facing a window and see me?
Do candles or lamps disturb your vision?
When you read, how do you want to have the lights in

the room?
Is it difficult to move about in a crowd?
On a windy day, is it difficult to go in places where

branches are waving?
If you turn your head quickly, does that disturb your

vision?
Can you see the curb?
Can you see the edge of the stairs going down?

395
Can you see the poles of the traffic signs?
Do some parts of your visual field function poorly?
Do you lose small objects out of sight and after a while

they are back?
Has your reading changed? - become slower?
Do you still read books and newspapers?
Do you have any difficulties in reading?
Which size text is too small?
Is the text too poorly visible? - the ink not dark enough?
Do you use a magnifier?
Do you have special reading glasses?
What other visual aids do you have? Did you bring them
with you?
What visual aids would you like to try to use?

An Evaluation of Follow Up Systems in Two
Low Vision Clinics in the United Kingdom
Andrew J. Jackson, Janet H. Silver, Desmond B. Archer
1. Introduction
Blind and partially sighted individuals resident in the United

Kingdom have access to a hospital eye service which supplies
low vision aids on an "on loan" basis. SILVER and THOMSITT
[1] in 1977 showed that low vision aids could be obtained from
over 100 centers in different regions throughout the United
Kingdom and, although current figures are not available, this
number has almost certainly increased over the last decade.
Most large hospital ophthalmology departments have low
vision clinics which are staffed by ophthalmologists, optome-
trists and dispensing opticians. Such clinics provide a wide
range of low vision aids and recycle unused, discarded and
outgrown aids making the service more cost effective. In this
system strong emphasis is placed on follow-up examination.
Alternatives to the "in house" low vision service include

referral systems using ophthalmic and dispensing opticians
working in private practice, contracting services provided by
major optical companies and university teaching clinic servi-
ces. In the United Kingdom ophthalmologists are responsible
for the referral of the visually handicapped for low vision ser-
vices, although other health care professionals, e.g. special-
ized social workers may recommend the provision of a low
vision aid and refer the patient through the appropriate medi-
cal channels.
Moorfields Eye Hospital (MEH) situated in central London

made valuable use of a dispensing optical company until the
late 1950's when a hospital based low vision service was
established. The system was restructured in 1970 and 1983
when new methods of assessment and follow-up were intro-
duced. In 1985, 2600 new patients attended the MEH low
vision clinics and the new patient to review appointment ratio
was 1:1.
397
Prior to 1983 low vision services in Northern Ireland were

administered by a visiting dispensing optician employed by
one of the major optical companies and local ophthalmic opti-
cians who provided low vision aids at the request of an
ophthalmologist. In 1983 Northern Ireland's largest health
board established an "in house" regional low vision service at
the Royal Victoria Hospital (RVH) in Belfast. Approximately
400 new patients attend the RVH low vision clinic annually
and the new patient to review appointment ratio is 1:3.
The low vision service at MEH is held in high regard by

the ophthalmic profession and attracts patients from a wide
area with ill defined boundaries. Figure 1 illustrates the per-
centage of low vision patients (94 consecutive cases attending
MEH) who reside in London, the home counties and other
parts of the United Kingdom. Thirty eight per cent of the
sample were resident in the Greater London area. The low
vision service at the RVH serves the entire province and Fig.
2 shows the geographical distribution of 100 consecutive new
low vision patients. Thirty nine per cent of the sample were
resident in Belfast.
Department of Health and Social Services statistics for

1984 state that a total of 3364 individuals in the province are
registered blind and 1258 partially sighted [2] (population
1,578,500). Nineteen eighty two statistics place the numbers
of registered blind and partially sighted individuals resident in
England at 111,729 and 58,003 respectively [3] (population
46,956,400). Comparative figures for Scotland [4] and Wales
(1984) [5] indicate that 12,168 and 8,048 patients were regis-
tered blind and 3,271 and 4,876 were registered partially
sighted. CULLINAN [6] has shown that more realistic esti-
mations of the total number of blind and partially sighted
individuals within the U.K. would be obtained if one doubled
the figures available from the registration statistics. How-
ever, about 15% of those registered blind have an acuity of no
perception of light or perception of light [7] and do not
require the services of a low vision clinic. From these data
we estimate that approximately 7,500 residents in Northern
Ireland and 300,000 residents in England may benefit from a
low vision assessment.
The provision of high quality services to the consumer

invariably led to an increase in the demand for the service.
Both MEH in the past, and the RVH at present have been
faced with a "demand exceeds supply" situation and this
398
GEOGRAPHICAL DISTRIBUTION
OF NEW PATIENTS ASSESSED (ENGLAND)
(n - 84)
london - 38'110 d1
~
.+.N
s
HOlM Countl•• - 45'110
Provlnlcll Ar ... - 17'110 1
6
Figure 1. Geographical distribution of places of origin of

94 consecutive new patients attending MEH low
vision clinics.
coupled with a rigorous follow-up review system has led to an

escalation in waiting list times. The follow-up systems used
in the two centers are now different and this paper compares
and contrasts their advantages and disadvantages.
399
GEOGRAPHICAL DISTRIBUTION
OF NEW PATIENTS ASSESSED (N.!)
(n - 100)
Bellast area - 39 ..
Provincial areas - 61 ..
Figure 2. Geographical distribution of places of origin of

100 consecutive new patients attending RVH
low vision clinics (' represents individual
patients).
All available clinical data for 100 consecutive new patients

who first attended the Royal Victoria Hospital low vision
clinic between May and August 1984 have been reviewed and
compared with data from 100 consecutive new low vision
patients who were supplied with low vision aids (LVAs) on
first attending the Moorfields Eye Hospital clinics during a
similar period of time. Although the present MEH study does
not include data on the percentage of new patients not sup-
plied with LVAs a previous study of 1000 new attendants car-
ried out in 1974 showed that 30% of patients were not sup-
plied with LVAs [8]. The current RVH study showed that 25%
of the sample attending low vision clinics were not supplied
with LV As. Table 1 shows that the reasons for not prescrib-
400
ing LVAs in these two studies are comparable. Similar fig-

ures are published by PARKER [9] and ROSENBLOOM [10].
Table 1
A comparison of the number of patients supplied with Low

Vision Aids on first attending RVH, MEH Low Vision Clinics.
R. V.H. SURVEY H.E.H. SURVEY
1984 Ul4
(n = 100) (n = 1000)
81 Refraction only 101 Refraction only
17" Rejected lVA'a

i .... diate1y 20~ Rejected LVA'.
innediately
··........
. . . . . . . .- ............... .. . . .
............................:::: ......... '-:.'-:.-:-.-:-.~.
:- .. :-:-:-:-:-:-:-
:- :- :- :- :- :- :- :- :- :- :- :- :- :- :- :- :- :- :- ......,-
...-:-
••-:-
••7""
• ..,....,....
•• ~-.j
- .- .
.• • • •.•.•.....,~...,++. ,. '
.::
.•...................
• '''#~+m'................
.' .' ," .. " .' .' .' .' .'
" " .,' .. . ,','.' .......... .
:-:-:.:- :-:-:-:-:-:-:.
»:-:.:.:-:-:.:-:-:-: ........ :,'.' '.'.'.'.' . .'.'.'.'.'".',' ',' ..
,' '.'
· . . . . . . . . . .. . .................. . .
· . . . . . . . . ...................
,'
· .................... .
· ..................... .
. :-:-:-:-:-:-:-:-:-:-:-:-:-:-:-',',',',',',',',', .. , ' , ' , ' , ' , ' , ' , ' ,
........... . . . . .. . . . .
· ··.................
............. . . ......
........ . "
· .' .' .' .' .' .' .' .' . . . . . . . .. ." . . ................ .. . . . . .
· ~ . :. :.:. :.:. ~ ': :':':'::.::. :': :::::.: :':':'::::::::::::::::::::':".
......................
·· .'.......... .. . . . . .. ..................... .
. ............ .'.... '
........................ .
......... :........ . .. :-: ............
-:-:-:-:'.-:-:'.-:.'.':-:-:-:-:-'.:-: :-:-:-:-'.:-:-:-:-:-:-:-'.
......
·
.... .. ...................
.
.'.'
.'.'.'
.'
............................................. .
.'.'
. .
......................... . ,
· .................................................... .
::::::::::::::::::::::::::::::"::::: :.:.::::::::::::::::::::::::':.
·. . . .' . ........... .' ....... ...... '. ........................
· . . . . . . . . . . . · . .. . . .. .... . . .. . . . . . . . . . .
401
2.1 Age and Sex Distribution
In the RVH study 49 patients were female and 51 male and in

the MEH survey there were 60 female and 40 male patients.
The age distribution of the patients seen in the RVH and MEH
samples were virtually identical (Figs. 3 and 4). The peak
attendance of new patients occurred in the 7th to 8th decades
reflecting the prevalence of visual disability in the ageing
population [11]. Smaller peaks in the 2nd to 3rd decades can
be explained in that the majority of people who develop a vis-
ual disability in early life are more incapacitated and seek
early assistance.
l~ PATIENT AGE
DISTRIBUTION ON
l~ rIRST ATTENDING
R. V.H. LOW VISION
CLINIC
(N • 100)
2~
20 18
I~ . ....... .. .. .
. .
.
I~
. . . . . . ..... .. .
.
10
10
...................
. . . . . . . .
.. ....
. ......... ...... .
. . . . . . . . . . .. ..
. . . . ... . ..
AGE IlANDINGS (YEARS)
Figure 3. Age distribution of the 100 consecutive new

patients in the RVH study group.
402
PATIENT Ia
DISTRIBUTION ON
'5 FIRST ATTEIIlIN;
M.E.H. LDW VISION
D.INIC
(N = 100)
. . . .. 27
25
2D
15 ........
......... .
10
...........
.... .................
.... . .
,-"
Ia IIAIIlINOS (YEARS)
Figure 4. Age distribution of the 100 consecutive new

patients in the MEH study group.
2.2 Aetiology of Visual Loss
Tables 2 and 3 indicate how similar the two survey groups

were with regard to cause of visual loss. Seventy two per
cent of the RVH group and 77% of the MEH group had retinal
or choroidal disease as the primary cause of visual disability.
Senile macular degeneration, diabetic maculopathyand myop-
ic chorioretinal degeneration were the commonest clinical
entities. Forty five per cent of the RVH sample and 54% of
the MEH sample had senile macular degeneration. Medial
opacification and diseases of the optic nerve or visual path-
way were the next most common conditions causing visual
disability. Congenital defects including cataract, albinism
and nystagmus were significant causes of visual loss (RVH
13%, MEH 7%).
403
Table 2
Site of ocular pathology responsible for reduced visual acuity

(RVH study).
Patient
Site of leaion nUMber Major pathological
groupe
(n = 100)
Senile Macular
Retinal and degeneration 4;
choroidal diaease 7Z
Diebet ie
1118culopathy 10
Myopic
degeneration
Lena opacities
Opacification of
the aedie 10
Corneal aearring
GlaucoM.
Diaeaae of the
optic nerve and
viaual pathNay 18
Optic atrophy
Others
CONGENlT AL CAUSES - D
2.3 Visual Acuity
Distance and near visual acuities were remarkably similar in

the two sample groups as shown in Figs. 5 and 6. Sixty four
per cent of patients referred to the RVH low vision clinic had
best corrected visual acuities, in the preferential eye on first
attending, of between 6/18 (20/60) and 6/60 (20/200). Sixty
per cent of patients in the MEH survey were in the same cat-
egories. Fifty per cent of the RVH group and 55% of the
MEH group had near acuities of NIO or better (with plano to
+4.00 dioptre sphere addition).
404
Table 3
Site of ocular pathology responsible for reduced visual acuity

(MEH study).
Petbnt
nueber Major pathological
group.
(n " 100)
Senile •• culer
Retina! and deganeration
choroid.l di ••• e_ 77
OhbeUc
•• culopethy
Myopic
daganeretion
Opacirication of
the •• dh
Cornllal IIcarring
Db •••• of the
optic nerve and
"laual pathway is
....................... :.
Optic atrophy 1
CllCENI TAI.. CAUSES 7
2.4 Low Vision Aids Supplied
Seventy five patients in the RVH group and 100 patients in

the MEH group were supplied with LVAs on their first visit to
the clinic. Seventy six per cent of those supplied with LVAs
in the RVH group and 82% of the MEH group were initially
supplied with only one LVA. Eighteen point six per cent of
18% of the respective study groups were initially supplied
with two LVAs and 5.4% of the RVH group were supplied with
three LVAs.
The types of LVAs prescribed to both study groups were

approximately the same (Tables 4 and 5). Stand magnifiers
were the aids of choice with internally illuminated systems
being more popular in the higher powers. On average 1.29
LVAs were supplied to each of the RVH study patients for
whom an LVA was prescribed; this compares with a figure of
1.13 for the MEH group.
405
Table 4
Type and magnification rating of LVA's supplied to the RVH

patients at first appointment.
I R.Y.H. SILO... 1
HN;HlfICATIOH BAIIIlIt«:S
lOW VISION LDIII PCII«R MEDILtI POWER HIQi Pmt£1l

AIDS SlPPlIED
ON rIRST
VISIT
Tohl
(<• 2.5) (d,S _
," (>,"
Itond
"-gnirtere 1J
"ond Illu.ll111ted ,
,
IIIUllllneled 1 IllUlldnateo
"
M.gnHiere
Non IlIUllllneted Non IllUllllneted Non IllUlllinated 1
5pect.cle
"'-gnlHen
Near Vbion
Talncopic
~Hi.r.
Monocular. • Monocular>! 6
1.
Binoculars Binocuiur'l
NlJII[R (J" LVA'. SlPPl.IED = 97

/rIJ&:R (J" PATIENTS SUPPLIED WITH LVA'. :- 7S
AVERAGE MMlEA IF LVA'. PEA PATIENT :- 1.29
The above data indicate that the RVH and MEH groups
correspond well in terms of age, sex, visual acuities, ocular
pathology and the types of LVAs supplied. The personnel
staffing the two low vision clinics, although differing in num-
bers, were similarly qualified and experienced and used simi-
lar assessment techniques [12]. The methods of follow-up
were the only significant difference between the two clinics
in terms of patient management. The RVH low vision clinics
adopt a pre-booked follow-up system whereby all patients
supplied with LVAs are requested to attend for review
appointments at 6 months, 18 months and thereafter at yearly
intervals. Moorfields Eye Hospital on the other hand changed
its follow-up system from a pre-booked system to a combined
follow-up on request and postal recall system in 1983. The
remainder of this paper will compare the follow-up data
available from the two studies.
406
Table 5
Type and magnification rating of LVA's supplied to MEH

patients at first appointment.
I H.E.H.
MAGNIFICATION
STillY
BAJ«lIt«;S
1
LOW V[SION LOW 1£0[114 HIGH
AIDS SUPPPLIEIl
~<~;RST
Total
( <x 2i ) ( x 2! - x 5 ) ( >x 5)
Hand
Magnifiers 27 11 l' 2
Stand Illuminated 2 IllUlAinated Illuminated 12

Magnifiers .9
Non IllUlllinated 5 Non IllulJlinated 11 Non Illuminated 11
Spectacle
Magnifiers 2. 17
• J
Near Vision
Telescopic
Magnifiers 7 - 7 -
Distance Honocular8 - 'Ionoculars 2 HonocuIar~ 1

Telescopes 6
~inocu18r9 1 inoculars - Blnocular~ 2
NLJ&:R (J" LVA's SUPPLIED 113

NI.Ml£R (J" PATIENTS SUPPLIED NITH lVA's 100
AVERAGE Nl.JeER I:F lVA's PER PATIENT loU
3. Results
3.1 Follow-up Analysis
As a result of the different types of follow-up systems more

data are available from the RVH group over an 18 month
period than from the MEH group. Figures 7 & 8 chart the
progress of the 200 patients who presented as new patients in
the present study.
407
(R.Y.H. SlimY)
OIl
n£ VISUAL ACUl TIES AS
RECORDED I,. Tt£ BETTER EYE
(J" 100 J£W PATIENTS (1\1
FIRST A"EtrI)ING THE LaM
VISIIJN Q.INIC
'"
~ ......
20
1.
~
11·::·· <.~
............ ::<:~
10
6& 69
6/12
6/18
6/24
6/J6
6/6IJ
6/6lJ
}/6O
J/6IJ .... L...
1/60 PI.. .~
J
Yi-..l AcuiUn (Diat)
(R.V.H. 5TI.I>Y)
OIl
THE J£AR ACUITIES

(+4.00 Add) AS RECORDED
IN THE BETTER EYE or 100
'" '"
NEW PATIENTS ON rIRSl
ATTENDIte; Ttl: LOW VISION
Q.INIC
II!
! 20
i'
10
Near .\cui lie.
Figure 5. Distance and near visual acuities at first atten-

dance of 100 consecutive new patients in the
RVH study group.
3.2 RVH Low Vision Aid Follow-up
As previously stated 75% of the 100 new patients who attend-

ed the RVH clinics were supplied with LVAs. Eight of the
remaining 25 patients only required a refractive correction
and 17 further patients were not suitable for an LV A on the
grounds of poor motivation or grossly reduced vision. At the
6 month follow-up period 67 of the initial 75 patients supplied
with LVAs reattended for follow up. Three of the remainder
were deceased and an additional 5 were lost to follow-up.
Fifty nine of the 67 reattendants were considered successful
or partially successful LVA users and the remaining 8 gained
no significant benefit from the aids and returned them to the
408
. M.L"'. 5Tt.oV
THE VISUM. ACUIT 1[5 AS

REC1JIJ(D IN H£ linTER
,. EYE f7' 100 NEW PATIENTS

IJI fiRST ATTUDlfrrII THE
lOW VISION Cllfim:
.>:~. "
20
... ... u
10 ·•••.• .... 7~
~ •......• >••.. '>
YISIML M:UITI£S (OIST)
. M.E.H. STlIlY
.
THE NEAR ACUITIES
( ..... 0Il Add) AS RECORDED
IN THE BETTER [V[ IF 100
,. P"UENTS ON fiRST
A"[ttnfrrll n£ lOW VISION
"
ClIJiIiC
~
I 20
~ 1> 1>
if
II
10
H5,6 NB,9,lO H12,14 N18,24 P06,48
NEAR ACUITIES
Figure 6. Distance and near visual acuities at first atten-

dance of 100 consecutive new patients in the
MEH study group.
low VlSlon clinic. Thirty eight of the reattendants had been

coping well with the original LVAs supplied and an additional
14 requested additional LVAs. The remaining 6 patients
either had repairs, replacements or LVA exchanges carried
out at the 6 month follow-up period.
At 18 months, 44 of the 59 patients who had been suc-

cessfully using LVAs at 6 months reattended for follow-up.
Four of the remainder were deceased and a further 11 were
lost to follow-up and failed to respond to postal enquiries.
Forty two of the 44 reattendants were still successfully using
LVAs, whereas the remaining 2 finally rejected LVAs. Twen-
ty eight of the successful reattendants were coping well with
existing LVAs and a further 11 requested additional LV As.
The 3 remaining patients required LVA exchanges or repairs.
409
3.3 MEH low Vision Aid Follow-Up
The MEH data recorded in Fig. 8 refers to any follow-up

information available for any of the initial 100 patients who
were supplied with LVAs. Where more than one follow-up
visit was made to the clinic the most recent data was that
chosen to represent the follow~up information. Nine patients
initially supplied with LVAs rejected assistance. Of the
remaining 91 patients 46 were lost to follow-up. Forty five
were found to be successfully using LVAs and of these 19
required no further change. Thirteen of the remaining
patients required additional LVAs and a further 12 exchanged
the original LVA for a more powerful one. Only one reatten-
dant required repair facilities.
...,.. patient, 6/12 follow up 18/12 (011_ up
lIS Refraction
In Rejected
LVA"
1-.:11,t_ly
lOS Rejected
l'IA' •
.,ltt_tely
4i!lO ·~.~'~lly .
..-IlU,',
"(56" of tI.;...
initi.ally
~lied)
R.V.H. SlJRV(y
16' l..-t to
1984 to 1986 follow up
'-------,-.,-l .. : ,.... . . . .... I - - - - j

AJJ/ JfS/DBA
........
................ :: ...... , n Decelsed
within 18/12
Figure 7. Follow up data for 100 consecutive RVH new

patients.
410
3.4 Follow-up Comparison
This study shows that after 18 months 42 patients or 56% of

the initial RVH study sample who were supplied with LVAs
were still successfully using their LVAs. This compares with
45% of the MEH follow-up patients. Twenty one point eight
per cent of the RVH group who were supplied with LVAs were
lost to follow-up and a further 13.4% of those supplied finally
rejected LVAs. Only 9% of the MEH group rea ttended and
returned rejected LVAs and a total of 46% did not request
any follow-up appointments at the clinics.
An analysis of the visual acuities, both distance and near,

of those 18 month reattendants from both stuc:J,ies can be seen
in Figs. 9 and 10. Only 18% of the RVH reattendants had a
reduction in distance acuity, compared with 33% of the MEH
group. Twenty per cent of the RVH group had a drop in near
acuity over 18 months compared with 33% of the MEH group.
These findings may reflect the different waiting list times for
the two clinics. The RVH patient, who may have been on the
waiting list for up to one year before being seen as a new
patient, may have achieved a greater level of visual stability
when seen initially than their MEH counterparts who have to
wait for less than one month to be seen in the low vision clin-
ic once referred. These findings may also account for the
fact that on review very few RVH patients required increased
levels of magnification, whereas 12 of the 54 patients sup-
plied with LVAs at MEH and subsequently followed up
required more powerful LVAs.
A more critical analysis of success and failure data for

the RVH group of patients (Table 6) reveals useful informa-
tion in the lost to follow-up category. Ten of the 16 patients
lost to follow-up at 18 months were managing well when seen
previously at the 6 month period and there did not appear to
be any obvious reason why there should be a dramatic turn
around in their circumstances. If this figure is added to the
successful cases who reattended at 18 months the potentially
successful LVA users figure rises to almost 70%. Equivalent
information from the MEH study was not available although
there is no reason to suspect that the situation there would be
any different.
411
New patient. 6/12 follow up 18/12 follow up OutcOll8
a ftefrlletion " RefrlleUon

onl,
17% Rejected
LY",'.
1~i.tely
lor; Rejected
LVA'.
ulti.. tely
.is '~e~r~l1'y .
ueed LVA'.
()6" of thoee·
initially
.~l~~).
Lalit to
",::.' ...
,','
follow n Dec"ased
within 18/12
..... "
Figure 8. Follow up data for 100 consecutive MEH new

patients.
412
(R.V.H. STWY)
CHANGES IN VISUAl ACUITY IF THE

44 PATIENTS WHO REATTENlED FOR
18/12 FOLLOW UP APPOINTMENTS
AT THE R. V.H. LOW VISION CLINIC
0-
20
~
...
0-
C
10
1=~~~==~·=·=·=·=F·=·=·=·9r__-,____,-__-,____,~rofliM.C~
on the SMllen
2 ) chert
4-----4 ~
IlIIprov_nt Reduction in Visual Acuity
40 in VA
J4 CHANGE IN I«:AR ACUITY (+4.00 Add)

IF THE 44 PATIENTS IHl REATTENlED FOR
18/12 FOLLOW UP APPOINTMENTS AT THE
RVH LOW VISION CLINIC
)0
0- 20
::
Z
:
10
7
NUllb"r of Banding
2 Categar ies changed
4~~4--~~~~==============~
IlIIprov_nt
Reduction in near acuity
in near
acuity
Figure 9. Changes in both distance and near visual acui-

ties as recorded over an 18 month period for
RVH patients who reattend for follow-up.
413
(M.£.H. STlIlY)
CHANGES IN VISUAl ACUITY If" 48

OUT If" 511 PATIENTS FOR WHIII
FOLLOW UP DATA WAS AVAILA8I..E
AT THE M.E.H. LOW VISION CLINIC
3D 2B
III
!'" 20
....
z
l!l
....
C
A-
ID 9
B
NuoIber of linea
I=;:;=*;:=~~=*=::;==::r==:::;:==~-...., Snellen
change on the
chart
2
Reduction in Visual Acuity
CHANGES IN NEAR ACUITY (+4.00 Add)

If" "B OUT ~ 511 PATIENTS fill WHII4
fOLLOW UP DATA WAS AVAILABlE AT
THE M.E.H. LOW VISION CLINIC
2B
20
10
s
Nuoober Df
"
2 Bending Categariea
....---..
Inoprav-.t ~~------------------------------.~
Reduction in Near Acuity
changed
in near
acuity
Figure 10. Changes in both distance and near visual acui-

ties as recorded over an 18 month period for
MEH patients who reattended for follow-up.
414
Table 6
Success/Partial; Success/Failure (Survey of 18/12 follow-up

data at the RVH)
PeUent
Nueber.
Cn = n)
Making .. ceUent. 1.1 . .

of lYA'. for
education or
••ploy •• nt+
Meking excellent 1.1 • •

of LVA'. for other
ddly livtng
ectivitie. 10
,n
Found LVA'. u •• ful
for •••• nU.I visu81
hske 19
LVA'. u.ed only in

••• rg8ncl ••
loet to eneving
follow up .ell when
l . . t . . .n 10 ll.l!lj
-._-- ---- - -.- - - - - -- - -- - ---r'l:'i~:·r-,H·Z~~·~::':'·~~··:·~:I:;
Uncertain
plognoll18 B"
Returned
LVA'. (unable to
cope or of no further
u .. ) 10 u.n
9,4\
4. Conclusions
The RVH study group shows that a substantial proportion of

patients provided with LVAs require replacements, repairs,
exchanges or additional aids by 6 months, and both study
groups indicate that further changes will be required by 18
months. This is clear evidence that some form of follow-up
system must be adopted.
A pre-booked follow-up system requires each patient to

reattend for re-evaluation and provides a psychological boost
for those experiencing any difficulties with the LVAs. A fur-
ther advantage of such a system is that it encourages those
patients not using the original LVAs to return them to the
clinic for recycling. This sytem, nevertheless, tends to esca-
415
late waiting lists and increases "failure to attend" rates. Fur-

thermore many satisfied patients are unnecessarily reviewed
making poor use of scarce clinical time.
A follow up on request system overcomes many of the

problems relating to the poor use of clinical time, however,
elderly and often confused patients who experience visual
deterioration or damage their LVAs may not avail themselves
of the opportunity for reassessment. In addition, those
patients who keep but do not use LVAs are not identified and
a valuable resource is lost to the clinic.
The postal follow-up system is potentially valuable, how-

ever it produces an extra administrative workload for the
clinic and depends on the patients' ability to discern any
change in their visual status.
We believe a combination of the three systems is required

and that clinicians must decide which patients are capable of
rebooking appointments when there are any signs of change.
Other patients may be suitable for postal follow-up and a
third group including the elderly confused and the young,
whose requirements are changing rapidly, may be best suited
to a fixed follow-up system. Other systems making use of
training schemes and review by other associated professional
groups may also be appropriate in areas where such systems
can be easily organized.
416
5. References
1. J. Silver, J. Thomsitt: Low vision services in the U.K.

Health Trends 9, 73 (1977)
2. Great Britain, Department of Health and Social Servi-

ces: Personal Social Services Statistics for Northem
Ireland (DHSS, Belfast 1984) in press
3. Great Britain, Department of Health and Social Servi-

ces. Health and Personal Social Services Statistics for
England. (HMSO, London 1985) Tables 1 & 2, iv-v.
4. Statistics Branch of Social Services Group of Scottish

Education Department. Intemal Publication on Blind
and Partially Sighted Registration Statistics (Edin-
burgh 1985)
5. Great Britain, The Welsh Office: Health and Personal

Social Services Statistics for Wales. (HMSO, Cardiff
1984)
6. T.R. Cullinan: Visually Disabled People in the Com-

munity. Health Services Research Unit Report 28
(University of Kent at Canterbury, 1977)

1972)
8. J .H. Silver, E. Gould, J. Thomsitt: The provision of

low vision aids to the visually handicapped. Trans.
Ophthalmol. Soc. U.K. 94, 310 (l974)
9. J.A. Parker: Guidelines in low vision prescriptions.

Can. J. Ophthalmol. 4, 382 (1969)
10. A.A. Rosenbloom: Prognostic factors in low vision

rehabilitaiton. Am. J. Optom. Arch. Am. Acad.
Optom. 47, 600 (1970)
11. J. Silver, J. Jackson: Visual disability. Pt. I: Introduc-

tion and epidemiology. Ophthal. Opt. 22, 841 (1982)
417
12. J. Silver, J. Jackson: Visual disability. Pt. 9: Patient

management. Ophthal. Opt. 23, 804 (1983)
Vision Examinations of Handicapped Children
at the Oregon State School for the Blind
Sandra K. Landis, Terrel D. Dutson, William Ludlum
1. Introduction
Many residents at the Oregon State School for the Blind (OSB)
have multiple handicaps which limit the clinical evaluation of
their visual needs. Some students are uncomfortable being
transported to vision examinations. Others have become
extremely wary of doctors regardless of the professional's
desire to provide a service, because past experiences with
doctors have been necessarily unpleasant. Because so many
of the children at OSB shared these characteristics, the doc-
tors took the equipment to the institution in order to examine
the children in familiar surroundings.
Fifty children at OSB were given V1S10n examinations,

including a visual evoked response (V.E.R.). This was the ini-
tial step to introduce the institution to optometry, so its ser-
vices might be available to Oregonians with low vision in the
future.
The objectives for the project were: (1) To evaluate the

visual needs and to detect the unmet visual needs of the chil-
dren at OSB; (2) To prescribe low vision aids and other refrac-
tive materials as required; (3) To establish the need for a low
vision clinic and optometric services at OSB; (4) To provide
optometric services to multi-handicapped Oregonians.
Comprehensive examinations which are individualized to

meet the needs of multi-handicapped chidren are lengthy and
require non-standard equipment. Private practitioners gener-
ally cannot perform the thorough testing required to assess
the functional vision of this population.
Literature reporting on-campus evaluations at institu-

tions for exceptional children is limited. KELLY [1] stressed
"that it appears the more handicapped the children, the high-
er incidence of visual defects". A total of 519 children were
419
screened of whom 42% had visual defects. Of these 129 chil-

dren, 57 (44%) had convergent strabismus, 38 (29.5%) had div-
ergent strabismus, and 11 (8.5%) had nystagmus. Kelly cited
five reasons to perform a full ophthalmological examination:
(1) To confirm the diagnosis of a visual defect; (2) To recog-
nize anatomical defects; (3) To correct refractive errors; (4)
To record defective ocular movements and field defects; and
(5) To explain the visual defects to teaching staff, parents,
social workers, etc. To measure visual acuities, we included
another reason: (6) To perform a visual evoked response on
non-verbal patients.
JOHNSTON and BAILEY [2] describe the experience of

several optometrists who visited the School for the Blind in
Bali. These professionals screened 47 residents. None had
been prescribed low vision aids or spectacle corrections.
Twelve students had residual vision, and optical aids were
prescribed for four children.
2. Methods
The examinations performed on the children at OSB can best

be described as "individualized". Tests that were utilized on
each student were a visual acuity assessment, either by a ver-
bal response using the Feinbloom, Snellen, or Lighthouse tar-
gets, or by a Visual Evoked Response. Retinoscopy was an
invaluable test because of its objectivity. Dynamic, Dark
Room, and Monocular Estimate Method were used as
required. Cover tests and ocular motilities were performed
on each child. A refraction using a portable phoropter or
loose lenses was attempted to correct refractive errors. Low
vision aids were introduced. Ophthalmoscopy and pupillary
reflexes were assessed to confirm a diagnosis of a visual
defect. A classroom assessment and a consultation with the
classroom teacher to share the results occurred for each indi-
vidual.
420
3. Results
Of the 50 children screened, 7 (14%) had no useable vision,

and 2 individuals (4%) were "not testable". One girl was vio-
lently frightened and would rip off the V.E.R. electrodes
before the measurements were made. The second child
pinched his eyes tightly during our repeated attempts to
implement our tests.
Thirty-six (72%) of these children would benefit from

"functional" vision care. One girl who had been prescribed a
contact lens/handheld telescope system had never been
trained in proper handling techniques. She depended on her
mother to insert and remove the lens even though she had
sufficient dexterity required for the task. Because her moth-
er could not come to the school daily to remove and replace
the lens before and after swimming the child did not use the
low vision system. Proper follow up care and training would
have been a key to success for this visual aid.
Another young lady who benefited immediately by a func-

tional vision examination did not realize that a head turn with
an extreme right gaze decreased her nystagmus. She used
this posture unknowingly while mobile, and quickly improved
her visual acuity while stationary by using this posture.
Nineteen (38%) of the fifty children required a prescrip-

tion change. Two individuals benefited from low vision aids.
The visual defects found in the OSB population follow:

421
Table 1: OSS Results (1985)
Visual Defects # %
Nystagmus 23 46
Strabismus 18 36
-Convergent 10 20
-Divergent 8 16
Optic Atrophy 11 22
Congenital Cataracts 11 22
Leber Congenital Amaurosis 2 4
Band Keratopathy 1 2
Glaucoma 8 16
Retrolental Fibroplasia 5 10
Phthisis bulbi 1 2
Meglocornea 1 2
Optic Nerve Hypoplasia 4 8
Rubella Syndrome 2 4
Sclerocornea 1 2
4. Discussion
Objectives 1, 2 and 3 were met satisfactorily during this

study. Many children at OSB were discovered to have unmet
visual needs. This is evident when one realizes that 21 of the
50 children required a new prescription of low vision aids. A
review of the individual reports finds recommendations for 25
children (50%) to receive functional follow up care or vision
therapy to develop accommodative binocular and eye
movement/fixation skills. Eleven children received a recom-
mendation for light stimulation. A Low Vision Clinic on cam-
pus at this facility will provide these needed services.
Functional assessments of these children would benefit

the students directly by determining the most appropriate
prescription and teaching the utilization of the aid. The
classroom observations and consultations with the teachers
are also beneficial, for additional information is often discov-
ered during this activity. One child who had sclerocornea
422
benefited greatly when material was presented on an illumi-

nated background. This became a very important tool in his
educational setting when its benefits were realized by the
teachers and the student.
A visual acuity assessment provides valuable information

to teachers who must present properly sized print or objects
in an educational setting. This finding also helps parents
structure the most appropriate home environment. Prior to
this study, this measure was an unknown fact about many of
these children.
The final objective (to provide optometric services to

multi-handicapped Oregonians) will soon become a realized
goal. This August a one year pilot project will begin. A Low
Vision Clinic on the campus of OSB will provide secondary
care for Oregonians. Children who require a functional
assessment will be referred to this facility by optometrists,
ophthalmologists, edqcators, or OSB. The children who are in
the Eastern or Southern portions of the state who are unable
to travel to the OSB Clinic will be served through a "Loaner
Office" system. The eye care professionals will travel to an
office being loaned for a day by a practitioner and will per-
form the testing. Special equipment will be transported with
the examining team which will consist of the Clinic Director
and Pacific University College of Optometry interns.
In addition to the 54 full time students, OSB serves 150

children during summer workshops. Low Vision services and a
functional assessment will be provided.
The testing at the On Campus Clinic will be similar to

that used in this study with one major improvement. Prefer-
ential Looking will be used as an additional objective measure
of visual acuity [3,4]. This method will provide an even
greater understanding of a child's visual discrimination abili-
ties.
HOLCOMB et al.[5] proposed four recommendations to

better equip educators with information about visually handi-
capped children. The On Campus Low Vision Clinic should
meet each of these goals suggested by Holcomb, Dilibero, and
Ryan: (1) More involvement by the eye care professional
schools to make students aware of teachers' needs and to pro-
vide low vision screenings is suggested. This will be realized
by student interns from Pacific University College of Opto-
metry being involved in the evaluations; (2) more in-service
423
education given to educators by eye care specialists; and (3)

revision of eye report forms to provide more information on
functional use of vision will be met by the written reports,
classroom observations, and consultation sessions following
each examination; (4) the Low Vision Clinic will serve as the
liaison office between teachers and eye care specialists to
improve communication about the children's needs.
5. Summary
Visually and multi-handicapped children require individualized

testing procedures to assess their visual acuities and amount
of functional vision.
Preferential looking and a visual evoked response are

used to assess the visual acuities of non-verbal patients.
An On Campus Low Vision Clinic to serve Oregonians

with special visual needs will be available beginning August,
1986.
424
6. References
1. M.E. Kelly: Vision screening of handicapped children

in special schools in South Wales. Br. Orthopt. J. 38,
68 (1981)
2. A.W. Johnston, r.L. Bailey: School for the Blind in Bali.

Aust. J. Optom. 62,466 {1979}
3. V. Dobson, M.A. McDonald, P. Kohl, N. Stern, M.
Samek, K. Preston: Visual acuity screening of infants
and young children with the acuity card procedure. J.
Am. Optom. Assoc. 57, 284 {1986}
4. P. Kohl, R.D. Rolen, A.K. Bedford, M. Samek, N.
Stern: Refractive error and preferential looking visual
acuity in human infants: A pilot study. J. Am. Optom.
Assoc. 57, 290 {1986}
5. S. Holcomb, M. Dilibero, J.B. Ryan: Needs analysis for
teachers of the visually impaired. Am. J. Optom.
Physiol. Opt. 63, 281 {1986}
Accommodation in the Visually Impaired Child
Eva Lindstedt
1. Introduction
The accommodative function of the visual system may be

investiga ted by many methods, most of which, however are
elaborate and require special equipment. This applies, for
example, to studies by infrared or laser optometers, ultrason-
ic measurements, pachometry and Visually Evoked Potential
(VEP) measurements. HOWLAND [1] designed a special pho-
torefraction system by which it is possible to study refraction
and accommodation in infants.
Current clinical methods, the "push-up" method as well as

the "lens-flipper" method, are convenient only for use in the
sighted adolescent and adult, based as they are on the
assumption of a normally developed capacity for "detecting
blur".
Studies of visually impaired people infrequently report on

near vision or accommodative capability. The application of
different methods to the visually impaired child may be ham-
pered by the frequent occurrence of disturbing symptoms
such as nystagmus, opacities of the media and focusing dis-
ability. OTTO and SAFRA [4] described failing accommoda-
tion in patients with central retinal affections. Lindstedt
reported reduced near vision in cases of severely impaired
vision [5] and in cases of Down's Syndrome [6].
2. Methods of Vision Assessment at a Centre for Visually

Handicapped Children
At the TRC, Tomteboda Resource Centre for Visually Handi-

capped Children in Solna, Sweden, visual rehabilitation of
children has been practised for about ten years [7,8]. A mul-
tidisciplinary low vision team (ophthalmologist, optometrist,
psychologist and low vision teachers) has been working with
426
children and their parents, and with teachers visiting the cen-
tre. The ambition was pragmatic but a lot of experimental
and developmental work has been done, not least concerning
methods and procedures of vision assessment.
Our interest soon concentrated on simple clinical meth-

ods of vision assessment, specifically designed for the young,
visually impaired, sometimes multiply handicapped child. The
current methods of vision assessment (e.g. Visual Acuity)
were found less convenient for our little clients. The types of
symbol used were unsuitable for the developmental level of
small children, and the testing distance was often less com-
fortable [9-11].
The introduction of LH vision tests and the construction

of the BUST-LH playing cards test made the assessments
more successful [12]. Free choice of viewing distance at near
made the procedures still more convenient, particularly for
the youngest children. Thus assessments were made possible
in children from about 20 months of age. Children with
severe visual impairments as well as additional handicaps,
e.g. hearing disability and mental retardation, could also be
assessed by the new procedure. When convenient, conven-
tional methods were also used.
To be able to "read" the assessment results, interpreting

them in Snellen Visual Acuity (V.A.) equivalents, conversion
tables were constructed. Lebensohn presented a conversion
table of a similar kind [13]. The design and manifold use of
the conversion/comparison tables are described in a booklet
[11].
When practising routine vision assessments at a distance

and at near, making use of the conversion/comparison tables,
a remarkable discrepancy was noticed between distance V.A.
and near V.A. in many of the visually impaired children, near
V.A. being more reduced than distant V.A. The resolution
power of the visual system cannot vary, and the difference
noticed must be attributed to an inability on the part of the
child to utilise his actual visual resolution/acuity at short dis-
tances. We thus believe the discrepancy to be due to a failing
accommodative capacity.
A controlled study of the conditions is to be published

later. The findings, however, seem to justify a report.
427
3. A Simple Method of Accommodation Assessment in

Children
Visual acuities at a distance and at near are assessed by a

method convenient to the child: the child is assessed with
correction for a refractive error. Preferably several assess-
ments are made. The V.A.s are expressed in Snellen equiva-
lents and compared. If the V.A. at near is considerably worse
than the distance V.A., this is taken as a sign of failing
accommodation.
Notes: The V.A. at near is assessed at a viewing distance

chosen by the child as most convenient when presenting "top
performance" at near. The viewing distance is measured
from the sheet of the testing material to the front of the cor-
nea or, in children wearing glasses, to the posterior surface of
the correction lens. This method of measuring the focal dis-
tance of the optical system of the eye is certainly inexact.
The magnitude of the error will vary, depending on the type
of refractive error, power of correction lens, fitting of glass-
es, and the viewing distance. The error will be greatest when
the viewing distance is very short. Considering all these
sources of error, the method is nonetheless quite useful clini-
cally, as the discrepancy between the V.A.s at distance and at
near is very often so great that even a scientifically signifi-
cant error is, for practical purposes, of limited importance.
4. Report on 52 Visually Impaired Children, Aged 2-6

Years
Fifty two children performed different vIsIon tests at the

TRC, Tomteboda Resource Centre for Visually Handicapped
Children, Solna, Sweden. The assessments were carried out
by a low vision team in the presence of the parents in several
sessions during a 2-3 day stay at the centre. They took place
in 1985.
For this report data were collected from the current clin-
ical forms used during routine examinations. Aphakic chil-
dren were excluded, as were three children whose results
were indefinite.
428
Sex and age. There were 19 children aged Z-4 years

(36%) and 33 children 5-6 years (64%). The sexes were equal-
ly represented (Z6/Z6).
Refraction. Thirty five children had refractive errors of

+/- 2 Diopters or more (67%), 17 had refractive errors of +/-
5 Diopters or more (32%). Data refer to the best eye only.
Diagnoses. The diagnostic panorama was varied. All

children except one, however, suffered congenitally or peri-
natally acquired visual impairments. Disease/damage of the
retina or optic nerve and visual pathways and centres domi-
nated. Minor additional handicaps were noted in 13 children
(25%). (Severely multiply handicapped children are attended
to by another centre in Sweden.)
Table 1
52 Visually Impaired Children, aged 2-6 years. Diagnoses.

Eye diseases No. Notes
Deletary myopia
Malformations of the globe syndromes 2
Albinism
Lens disorders (catav., sub lux) rubella 1
Retinochorio'iditis toxoplasmosis
Retinopathy of prematurity RLF 3
Hereditary retinal dystrophies achromatopsia 3
Atrophy of the optic nerve, hereditary syndromes 6.
visual pathways and centres 16 (30%) perinatal damage 6.
hydrocephalus 2. bypo-
plasia 2
Congenital nystagmus UND ...!

52
Additional visual disorders
Nystagmus 46 (90%)
Squint 35 (68%)
429
Oculomotor dysfunctions. Thirty four children (66%) had

a squint and nystagmus was observed in 46 (90%). In addition,
several children had focusing problems leading to abnormal
position of head and gaze during focusing. Very often there
was adduction of the focusing eye. This happened also in
children who had only one functioning eye. This habit of
adduction of the eye is a nuisance when the children ought to
be wearing glasses. Surgery to accomplish a more favourable
position of the eye when "blocking the nystagmus" is strongly
indicated in such a case. Even a partially successful opera-
tion may make it possible for the child to use a badly needed
optical correction.
Visual acuity. V.A. was assessed by different methods, as

a rule several assessments were carried out and a mean value
obtained. Data (Figs. 1,2 and Table 2) refer to the best eye
with correction for refractive errors.
Table 2
52 visually impaired children aged 2-6 years. Visual acuity.
VISUAL ACU ITY
distance at near
No No
~-~ W (19 i') 33 (b5~)
~-% 24 15
-%20
;::. 18 4
52 52
Distance visual acuity was assessed by the LH single or

line test, or the HVT test. In a few cases Ffooks' distance
vision test or Sheridan's test was used. The tests were per-
430
formed at 3 m or, if necessary, at 2-1 m in which cases the

V.A. values were calculated accordingly. Correction glasses
were used by the child.
Near vision was recorded in Snellen visual acuity equiva-

lents. A conversion table was used for converting test results
at near to an reduced Snellen acuity.
The near vision tests used were BUST playing cards, LH

playing cards, LH near vision test (single or line) and Ffooks'
near vision test. Most of the children were assessed by two
or more methods.
The assessment results, though approximate, were consid-

ered significant, because they were confirmed by cross-
checking using different tests.
The child was allowed to choose viewing distance during

the assessment, which aimed at measuring the "top perform-
ance" at near.
Viewing distance was found to be very short in all chil-

dren, most of those choosing a viewing distance of 10 cm or
less.
Table 3
52 visually impaired children aged 2-6 years. Viewing dis-

tance chosen at near performance.
viewing distance No
15 - i1 em 7
iO - 6 em
27}
. 45 (87~)
.::::. 5 em 18
431
Binocularity is not recorded. The severely visually

impaired children, as well as the youngest children, were una-
ble to manage the TNO or Freesbee tests. Those who man-
aged as a rule failed. Thus data concerning near vision and
accommodation refer to the best eye.
Accommodation. The visual acuities as a rule were lower

at near than at far. This was true for the majority of chil-
dren (Fig. 1). Considering the approximate character of the
values in young children and the possibility of errors of meas-
urement, the divergency between the V.A.s at distance and at
near remains very remarkable. As can be seen from Fig. 1,
the resolution utilised at near was often only half or one-third
of the resolution/acuity used at a distance.
V.A.
at near
Ys 0.1
X. 0.6
%. Os
~ 04
~
%'0 0.3
;{.
~D 02
~
~ 0.1
v
1/
,•
~
:v
0.•
l( pc
~ 0.•
V ap
~ o. 4
)to
¥o Q.
V
D.• .,q,
o 0
o 0.2 0.3 0.4 0.5 0.6 0.1
I I I I I I I I I I I I I I I
o XD ;to XoXXoXo ~ %, X. %oX % Xz Xc Ys
Distance V. A.
Figure 1. 52 visually impaired children aged 2-6 years.

Visual acuity. Best eye, with correction.
432
;/.
iDO
,, __ V /
.......
90
I
/'"
,,
I
80
70 /
60
,I
50 I /
III
,,
40
30
20 !I
10 I
o 1/
o 0.1 0.2 0.3 0.4 0.5 0., 0.7
- - - At near - Distance
Figure 2. Cumulative frequencies of visual acuities at

distance and at near.
Near correction. Near correction was fitted in many of

the children, but, as the follow-up is still in progress, no
report is offered on the results. In many cases, however,
immediate relief was experienced. The fitting {not seldom of
bifocals} was very demanding and had to be done ad hoc, tak-
ing into account the vastly varying types of problems,
{refractive errors, abnormal gaze, practical needs etc.} of
each child.
5. Discussion
In addition to impairments of visual acuity, a majority of the

children reported had refractive errors and suffered from
oculomotor dysfunctions including nystagmus, squint and
focusing disability. We would like to add: and deficiency of
accommodation. What is evident from the study, however, is
that many of the visually impaired children were unable to
utilise all their actual visual resolution/acuity at near.
433
The development, in the sighted child, of visual

resolution/acutiy and accommodation seems to proceed
quickly through the first months of life [14]. The capacity for
"detection of blur" may be postulated as the trigger mecha-
nism of developmental sequences. A child, from birth suffer-
ing from reduced visual capability/potential relative to "nor-
mal", might fail in "detecting blur", thus lacking the sensory
motor stimulation and feedback which should give rise to
development of accommodative accuracy.
There is good reason for paying greater attention to the

near visual acuity of infants and young children, and for
developing and refining appropriate methods of assessment [9,
15]. The fitting of glasses for near should be considered as a
means of promoting visual acuity development in the very
young visually impaired child [5].
When accommodation fails or is incorrect, hyperopia will

be of greater significance as a vision reducing factor also at a
distance.
When entering school and confronted with text-reading

tasks, the visually impaired child may drastically experience
the disabling consequence of a greatly reduced visual acuity
at near. This might be all the more of a shock to a child of
moderate visual impairment who until then has been able to
manage daily visual tasks fairly well.
The World Health Organization (WHO) definition of "cat-

egories of visual impairments" states a visual acuity of less
than 6/20 (20/66) as a possible cause of visual disability [16].
A child assessed to have a distance V.A. of 6/20 (20/66) or
above and complaining of visual disability as a rule is consid-
ered "fussy" and left without remedy. The sample of visually
impaired children reported includes a considerable number of
children with a distance V.A. of 6/20 - 6/9 (20/66 - 20/30).
As can be seen from Fig. 2, all of the children actually used a
V.A. of only 6/20 (20/66) or less at near.
When studying the 18 children with a distance V.A. of

6/20 - 6/9 (20/66 - 20/30) we found no difference from the
total sample as regards the diagnoses, the occurrence of
refractive errors and of squint or nystagmus. There was,
however, great discrepancy between distance and near visual
acuities: all except two of the children using less than 50%
of their visual acuity at near and 9 of them using less than
434
40% of their actual visual acuity at near. A failing accom-

modation ability might very well be an important reason why
these children experience considerable visual disability in
spite of a fairly good distance visual acuity.
A child who is complaining of visual disability should be

assessed properly for near visual acuity in addition to the rou-
tine distance visual acuity assessment.
6. References
1. H.C. Howland, B. Howland: Photorefraction: A tech-

nique for study of refractive state at a distance. J.
Opt. Soc. Am. 64, 240 (1974)
2. H.C. Howland: Infant eyes: Optics and accommoda-

tion. Curro Eye Res. 2, 217 (1982/3)
3. K.E. Brookman: A retinoscopic method of assessing

accommodative performance of young human infants.
J. Am. Optom. Assoc. 52, 865 (1981)
4. J. Otto, D. Safra: Accommodation in amblyopic eyes.

Metab. Ophthalmol. 2, 139 (1979)
5. E. Lindstedt: Early Visuo-oculomotor development in

visually impaired children. In: Early Visual Develop-
ment - Normal and Abnormal, ed. by L. Hyvarinen, E.
Lindstedt. Acta Ophthalmol. (Suppl.) (Copenh.) 157,
103 (1983)
6. E. Lindstedt: Failing accommodation in cases of

Down's Syndrome. Ophthalmic Paediatr. Genet. 3, 191
(1983)
7. E. Lindstedt: Assessment, counseling, and training of

integrated visually impaired children. J. Visual
Impairment & Blind. 73,351 (1979)
8. E. Lindstedt: Early vision assessment in visually

impaired children at the TRC, Sweden. Br. J. Visual
Impairment 2, 49 (1986)
435
9. L. Hyvarinen, E. Lindstedt: Assessment of Vision in

Children, (SFR Tal §Punkt, Stockholm 1981)
10. E. Lindstedt: How Well Does a Child See (ELISYN,

Stockholm 1984)
11. E. Lindstedt: Conversation/Comparison Tables for

Vision Tests: Instructions and Comments (ELISYN,
Stockholm 1986)
12. E. Lindstedt, L. Hyvarinen: BUST-LH Playing Cards:

Manual (ELISYN, Stockholm 1985)
13. J.E. Lebensohn: Visual acuity tests for near: Implica-

tion and correlations. Am. J. Ophthalmol. 45 (4 pt. II),
127 (1958)
14. J. Atkinson: Human visual development over the first

six months of life. A review and a hypothesis. Hum.
N eurobiol. 3, 61 (1984)
15. L. Hyvarinen, E. Lindstedt, eds: Early Visual Develop-

ment - Normal and Abnormal. Acta Ophthalmol.
(Suppl.) (Copenh.) 157, (1983)
16. World Health Organization: International Classifica-

tion of Impairments, Disabilities and Handicaps. (WHO,
Geneva 1980) p. 80.
Abnormal Arm Tone, Cigarette Smoking and
Use of Blood Pressure Medication in a
Sight Enhancement Clinic Population
M.E. Paetkau
1. Introduction
There are three pigmented tissues originating from neural

crest which frequently degenerate during aging in industrial-
ized societies: the pigmented epithelium of the eye, the sub-
stancia nigra of the thalmus and the pigmented cells of the
organ of Corti. Each of these tissues once developed does not
regenerate following loss. They must last the insults of a
lifetime for the elderly to see clearly, hear well, and have
smooth coordination. Is degeneration of these tissues interre-
lated? Are they influenced by common risk factors?
Three diseases relate to degeneration of the pigmented

tissue of the thalmus: Parkinson's disease, paratonia, and
essential tremor. The expected prevalence of these diseases
is shown in Table 1 [1,2,3]. Each of these diseases may result
in tremors and/or abnormal arm tone. Although tremors and
abnormalities of arm tone have individual characteristics [4],
for statistical reasons, this study combines the patients with
abnormal arm tone relating to diseases of the thalmus and
contrasts them to those with normal arm tone. These diseas-
es also have common factors. Prevalence of Parkinson's dis-
ease is found twenty-four times the expected among those
with essential tremor [5]. Parkinson's disease shares with
paratonia high incidence of the gabeller reflex and the nucho-
cephalic reflex [2].
Age related maculopathy is found in 9.6% of white males

and 6.9% of white females age 70 and over [6]. The Framing-
ham study of patients who developed age related maculopathy
showed association with weak hand grip many years previous
to the development of disease [7]. This study questions
whether there is increased prevalence of abnormal arm tone
among patients with macular degeneration. Combined degen-
437
eration of the pigmented tissues of the thalmus and retina

might account for the Framingham finding. Also questioned
is if hearing loss is more frequent in either patients with age
related maculopathy or abnormal arm tone. Abnormal hearing
is predicted among 2.5-50% of those aged 65 and over [8].
Light colored eyes have been associated with hearing loss [9]
and age related maculopathy [10,11]. Other risk factors asso-
ciated with age related maculopathy include: age, smoking
habits [12], hypertension [13,14], family history of age related
maculopathy [10], chemical exposure [10], history of lung dis-
ease [7]. Abnormality in lipids has been suggested in age
related maculopathy [15]. The presence of corneal arcus was
noted as it is associated with lipid abnormalities [16]. Expo-
sure to light has been suggested as a factor in macular degen-
eration.
The interaction of the three diseases and diabetes is

examined. The expected prevalence of diabetes is 9/100 in
those aged 70 and over [17].
The low vision clinic at the University of Alberta is the only

clinic doing visual rehabilitation in the Northern half of
Alberta with a population of 1 million. Patients are referred
by regional ophthalmologists. History was ascertained by one
of two trained interviewers. Smoking history was recorded in
pack years to ascertain lifetime exposure. A pack a day for
one year equals one pack year. Passive smoking was not esti-
mated as there is no way to quantify lifetime exposure.
Patients were asked if they had ever been on hypertensive
medication and if they had ever had lung disease or a diagno-
sis of diabetes. They were asked if they had a relative with
blindness associated with macular degeneration, other or
unknown causes. They were asked if they had ever worked
with or been exposed to non specific chemicals and if they
were a ware of any hearing loss.
Patients aged 65 and over seen in the low VISlon clinic

during a two year period were classified as to whether visual
loss was due to age related maculopathy or other causes.
438
The presence or absence of any tremor was noted. While

patients were requested to relax, each arm was passively
moved to determine if arm tone was normal or abnormal.
Normal arm tone was present if the arm moved easily with no
rigidity, cog-wheel motion, intermediate opposition or catch-
ing. Two observers noted arm tone and eye color. The cornea
was examined with the +10.00 D ophthalmoscopic lens to
determine if it was clear or otherwise. Corneal arcus was
noted if a distinct white ring was present separated from the
edge of the cornea by a clear zone. Confirmation of referred
retinal diagnosis was done visually where possible. Those in
whom retinal diagnosis was impossible due to more anterior
pathology were noted and classified as non age related macu-
lopathy. Data analysis was done using the statistical package
for the social sciences.
3. Results
A total of 323 patients were enrolled. Age and gender distri-

bution is shown in Table 2. Visual loss related to age related
maculopathy and other causes is shown in Table 3. The most
frequent causes of other visual loss were diabetic retinopathy
(19), glaucoma (13), high myopia (7), cataract (4). Rarer
causes of visual loss relating to trauma, genetic and metabol-
ic diseases of the eye accounted for the other 33 cases.
Table 1
Expected prevalence of degenerative diseases of the thalmus
Age
Parkinson's 65-74 639 / 100,000
75+ 1148 / 100,000
Paratonia - 65-69 6%
70-74 10%
75-79 12%
80+ 21%
Essential Tremor 1 - 10 in 100,000
439
Table 2
Age by gender distribution of study population
Male Total
~ - - Female
65-69 7 16 23
70-74 22 43 65
75-79 25 62 87
80+ 56 92 148
110 213 323
Table 3
Age by sex distribution of patients with age related maculo-

pathy and visual loss from other causes
ARM OTHER
M p- M F
65-69 3 10 4 6
70-74 14 29 8 14
75-70 19 44 6 18
80+ 48 78 8 14
84 161 26" 52
Abnormal arm tone was present in 50 patients. Table 4

shows those with abnormal arm tone by age. Table 5 shows
the expected percentage of paratonia by age [2] and the per-
centage of those with abnormal arm tone by age and smoking
habit. Higher than expected percentages of those with abnor-
mal arm tone was found for those smoking 35-124 pack years
aged 70 and over. Table 6 shows the expected percentage of
paratonia and the percentage of those with abnormal armtone
440
by age and gender. Rates were double for males age 7S-79.
Table 7 sho;rs the relationship between smoking habit and
arm tone (x = 9.17, p<0.0102). Table 8 shows tha"i smoking
habits were significantly associated with gender (x = 97.11
p<O.OOOS). Those smok~g 3S-124 pack years were more likely
to have lung disease (x = 8.7i, p<0.02) and for males more
likely to have corneal arcus (x = 7.04, p<0.03). The interac-
tions of the three diseases associated with pigment tissues of
the brain and risk factors is shown in Table 9. Women with
abnormal arm tone were more likel[ to have good hearing
than those with normal arm tone (x = 7.39, p<O.O~ and to
have been treated with hypertensive medications (x = 4.01,
p<O.O~). Abnormal arm tone related to the presence of trem-
or (x = lSl.91, p<O.OOOS). Tremor was so disabling that
approximately 1/3 of those with abnormal arm tone were una-
ble to hold telescopes. There were no other relationships
between the three pigment tissue diseases. Having a light
colored ir~ was associated only for men with macular degen-
eration (x = 16.74, p<O.OOOS). It was not associated with use
of blood pressure medication. The age related maculopathy
group had the following characteristics: 76% used medication
for high blood pressure, S7% had hearing loss, 29% had a rela-
tive with visual loss due to maculopathy or unknown cases,
8% admitted chemical exposure, lS% had a diagnosis of ~a
betes, 26% had corneal arcus. Age related ~aculopathy (x ; ;:
lS.55, p<0.002) and tremor (women only - x = 10.15, p<0.02)
were both associated with older age. Sixteen patients had all
three degenerative conditions.
Table 4
Patients with abnormal arm tone by gender and age
Abnormal Arm Tone

M F Total
65-69 1 0 1
70-74 1 4 5
75-79 7 7 14
80+ 14 16 30
TI 27 50
441
Table 5
Expected % of paratonia *(Jenkyn) and % of abnormal arm-

tone (AAT) by smoking habit
STUDY POPULATION
+
Expected Never Smoked - 35 Pack Years = 35 Pack Years
Age % AAT Total % AAT Total % AAT Total %
65-69 6% 1 11 9% 0 7 0% 0 4 0%
70-74 10% 1 27 4% 1 21 5% 3 15 20%
75-79 12% 6 44 14% 4 28 14% 4 13 31%
80+ 21% 18 87 21% 3 33 9% 9 26 35%
26 169 15% 8 89 9% 16 5!l 28%
Table 6
Expected percentage of paratonia *(Jenkyn) and % of abnor-

mal armtone (AAT) by gender
STUDY POPULATION
MALE FENALE
Expected
Age % AAT Total ;!; AAT Total ;!;
65-69 6% 1 14% 0 16 0%
70-74 10% 1 22 6% 4 43 9%
75-79 12% 25 28% 62 11%
80+ 21% 14 2§.... 25% 16 ...1.L 17%
23 110 21% 27 213 13%
442
Table 7
Relationship between smoking habit and armtone (Chi square

test p<O.Ol02; 7 pipe smokers with normal armtone not
included)
Normal Abnormal
Arrutone Armtone
Non-Smokers 143 26
1-34 pack-years 81 8
+ 35 pack-years 42 16
Total 266 50
Table 8
Relationship between gender and smoking habit (Not shown -

7 male pipe smokers; Chi-Square = 97.11 P < 0.0005)
Never 1 - 34 35 - 124
Smoked pack years pack years
Female 152 47 14
Male 17 42 44
Total 169 89 58 316

443
Table 9
Summary if risk facts and degenerative disease and P value

using Chi-Square test
." .
. .'"."
.:
"
:l!.... g:,
....."
>,0
0
.
..
.Y.
u •"
'"o •~ ~,g
. ,. .... .. '"
o •
.... " ~ .~
~
~
g
.<:
u
-::
0
:.
..
2(
~ ...~
'"
E
....." ""
oo'"
~
.,~
Age related
0.0002 NS
"'u
0.0002 NS NS 0.06 NS NS
"
O~OOO5 NS
"'-"
NS
H
NS
maculopa thy M NS
Abnormal .02 0.05 NS NS

0.06 0.06 NS NS NS NS NS Ns
armtone F
NS NS
Tremor 0.02F NS NS NS 0.04 NS NS NS NS NS NS NS
F
Hearing NS NS NS NS 0.04 NS NS NS NS NS NS NS
M
NS - not significant
* - negative association
M - association for male only
F - association for fe:nale only
4. Discussion and Conclusions
Aging degeneration of the three pigment tissues of the brain

occurs independently for each tissue. In this study risk fac-
tors are not commonly shared. Some risk factors associated
in the literature with macular degeneration were also associ-
ated with abnormal arm tone including smoking and hyperten-
sion. In the BALTIMORE study [10], patients with diabetes
were eliminated due to common shared risk factors such as
hypertension and diabetic retinopathy [14]. Our control group
was visually impaired. Many had diabetes. Risk factors may
be common for several diseases such as exposure to chemicals
and optic atrophy.
Although we cannot halt the passing of time or alter our

genetic heritage, perhaps incidence of tremor and abnormal
arm tone among the aged can be reduced by smoking less than
35 pack years. Smoking 35 pack years or more was signifi-
cantlya male custom during the life span of those now 65 and
over. The outcome of the changed smoking habits following
1940 where women were encouraged to smoke may lead to
excessive numbers of the elderly with abnormality of arm
tone.
444
The etiology of Parkinson's disease and related diseases

[18] is varied including virus triggering, autoimmune process,
carbonmonoxide poisoning [1], street drug use [19], psychiat-
ric drug use [1] and minor stroke relating to hypertension, but
not cigarette smoking [20]. Smoking heavy amounts increases
carbon monoxide [21] and lead blood levels [22]. Systemic
lead poisoning of rabbit retinal pigment epithelium has been
shown [23]. There may be a similar reaction in the substancia
nigra and in the pigmented epithelium of the eye to chronic
chemical poisoning. Damage from a multitude of chemicals
in cigarette smoke may account for our findings. The use of
hypertensive medication was associated with women having
abnormal arm tone. Whether the damage to the thalmus
relates to the vascular insults of hypertension itself or chemi-
cal poisoning from one of the commonly used drugs must be
determined.
Pathologic studies of the choroid of age related maculo-

pathy have not shown consistent findings of atherosclerosis in
all eyes [24]. Perhaps macular degeneration and degeneration
of the thalmus are the end result of several different mecha-
nisms of damage. Separating patients into groups according
to those influenced by genetics, atherosclerosis, hypoxia and
chemical poisoning may be helpful in understanding these dis-
eases.
Rehabilitation of the visually impaired with profound

tremor is severely limited as aids are difficult to hold. Trem-
or has a high social cost including early retirement and
embarrassment [25]. The use of a stand magnifier occasional-
ly facilitates reading. Telescopes and microscopes may be
mounted on spectacles for these patients. Prevention of
abnormality of arm tone should be encouraged.
Warning: excessive cigarette smoking may lead to abnor-

mal arm tone which is associated with tremor in old age.
Warning: the one who falls in love with the man in the
cigarette ad must love well, for in old age she may be his
hands and if he is blue eyed, his vision.
445
5. Acknowledgements
I wish to thank Dr. S. Warren for help in design, Drs. T. Tea-

rum and M. Grace for statistical assistance, Dr. D. McLean
and Dr. R. Ranayawa for neurologic advice, Leslie Moyer,
Deb Bignell and Lorna Hensel for interviewing patients,
Elaine Olynyk for data entry, Dr. G. Molnar for encourage-
ment. The research was supported by the Muttart Diabetes
Research and Training Centre.
6. References
1. B.S. Schoenberg, D.W. Anderson, A.F. Haerer: Preva-

lence of Parkinson's disease in the biracial population
of Copiah County, Mississippi. Neurology 35, 841
(1985)
2. L.R. Jenkyn, A.G. Reeves, T. Warren, R.K. Whiting,

R.J. Clayton, W.W. Moore, A. Rizzo, I.M. Tuzun, J.C.
Bonnett, B.W. Culpepper: Neurologic signs in senes-
cence. Arch. N eurol. 42, 1154 (1985)
3. A.F. Haerer, D.W. Anderson, B.C. Schoenberg: Preva-

lence of essential tremor: Results from the Copiah
County Study. Arch. Neurol. 39, 750 (198Z)
4. L.P. Rowland: Merritt's Textbook of Neurology, 7th

ed. (Lea & Febiger, Philadelphia 1984)
5. J.J. Geraghty, J. Jankovic, W.J. Zetusky: Association

between essential tremor and Parkinson's disease.
Ann. Neurol. 17,329 (1985)
6. B.E. Klein, R. Klein: Cataracts and macular degenera-

tion in older Americans. Arch. Ophthalmol. 100, 571
(1982)
7. H.A. Kahn, H.M. Leibowitz, J.P. Ganley, M.M. Kini, T.

Colton, R.S. Nickerson, T.R. Dawber: The Framing-
ham Eye Study. II. Association of ophthalmic patholo-
gy with single variables previously measured in the
Framingham Heart Study. Am. J. Epidemiol. 106, 33
(1977)
446
8. P.W. Alberti: Hearing loss: What's the underlying

cause? Diagnosis 3, 31 (1986)
9. R.N. Kleinstein, M.R. Seitz, T.E. Barton, C.R. Smith:

Iris color and hearing loss. Am. J. Optom. Physiol.
Opt. 61, 145 (1984)
10. L. Hyman, F. Ferris, S. Fine, A.M. Lilienfeld, A. Var-

ma: Risk factors in senile macular disease. Invest.
11. J.J. Weiter, F.C. Delori, G.L. Wing, K.A. Fitch: Rela-
tionship of senile macular degeneration to ocular pig-
mentation. Am. J. Ophthalmol. 99, 185 (1985)
12. M.E. Paetkau, T.A.S. Boyd, M. Grace, J. Bach-Mills, B.

Winship: Senile disciform macular degeneration and
smoking. Can. J. Ophthalmol. 13, 67 (1978)
13. R.D. Sperduto, R. Hiller: Systemic hypertension and

age related maculopathy in the Framingham Study.
Arch. Ophthalmol. 104, 216 (1986)
14. W.C. Knowler, P.H. Bennett, E.J. Ballintine: Increased

incidence of retinopathy in diabetics with elevated
blood pressure: A six-year follow up study in Pima
Indians. N. Engl. J. Med. 302,645 (1980)
15. V. Landolfo, L. Albini, S. De Simone: Senile macular

degeneration and alteration of the metabolism of the
lipids. Ophthalmologica 177, 248 (1978)
16. J. Pe'er, J. Vidaurri, S.-T. Halfon, S. Eisenberg, H.

Zauberman: Association between corneal arcus and
some of the risk factors for coronary artery disease.
Br. J. Ophthalmol. 67, 795 (1983)
17. A.P. Bender, J.M. Sprafka, H.G. Jagger, K.H. Muckala,

C.P. Martin, T.R. Edwards: Incidence, prevalence and
mortality of diabetes mellitus in Wadena, Marshall and
Grand Rapids, Minnesota: The three city study. Dia-
betes Care 9, 343 (1986)
18. R.C. Duvoisin: Is Parkinson's disease acquired or

inherited? Can. J. Neurol. Sci. 11, 151 (1984)
447
19. J •W. Langston, P. Ballard: Parkinsonism induced by

1-methyl- 4-phenyl-1,2,3,6-tetrahydropyridine (MPTP):
Implications for treatment and the pathogenesis of
Parkinson's disease. Can. J. Neurol. Sci. 11, 160
(1984)
20. A.H. Rajput: Epidemiology of Parkinson's disease.

Can. J. Neurol. Sci. 11, 156 (1984)
21. U. Spohr, K. Hofmann, W. Steck, J. Harenberg, E. Wal-

ter, N. Hengen, J. Augustin, H. Morl, A. Hock, A.
Horsch, E. Weber: Evaluation of smoking-induced
effects on sympathetic, hemodynamic and metabolic
variables with respect to plasma nicotine and COHb
levels. Atherosclerosis 33, 271 (1979)
22. J.L. Hooper, J.D. Mathews: Extensions to multivariate

normal models for pedigree analysis. IT. Modeling the
effect of shared environment in the analysis of varia-
tion in blood lead levels. Am. J. Epidemiol. 117, 344
(1983)
23. D.V.L. Brown: Reaction of the rabbit retinal pigment

epithelium to systemic lead poisoning. Trans. Am.
Ophthalmol. Soc. 72,404 (1974)
24. W.R. Green, S.N. Key: Senile macular degeneration:

A histopathologic study. Trans. Am. Ophthalmol. Soc.
75, 180 (1977)
25. W. Koller, N. Biary, S. Cone: Disability in essential

tremor: Effect of treatment. Neurology 36, 1001
(1986)
The City Study - Preliminary Findings
J.H. Silver
1. Introduction
The City Study is a multidisciplinary longitudinal study into

the vision of people entering retirement. The overall aim is
to discover any associations between visual disability and oth-
er factors, to define any predictors, and to investigate the
natural history of disorders threatening vision in older people.
The principal researchers are a physician, an ophthalmologist,
a statistician and an optometrist.
2. Background
In the western world most visual disability is caused by the

degenerative changes that are associated with age, with more
than half of all recognised visual loss being associated with
cataract or senile macular disease [1]. In relatively few
instances is there a dramatic event, retinal detachments,
acute angle-closure glaucoma or massive diabetic haemor-
rhage into the vitreous being perhaps the most conspicuous
exceptions. Sight threatening disease in later life frequently
remains undiagnosed until much vision has been lost. The eld-
erly consider loss of vision or hearing to be an inevitable con-
sequence of old age and may therefore ignore early signs.
There is evidence that loss of vision has a connection with
falls in elderly people leading to a loss of mobility [2]. A spi-
ral can be created by the inability to read the instructions on
medication which reduces compliance with medical regimes
leading to greater confusion and an exacerbation of any other
medical condition.
There is a series of myths about the consequences of old

age. As well as belief in the inevitability of visual loss, there
is also the sense that sight can, in some way be used up, or
"strained" and therefore use of the eyes should be avoided.
Or that use can in some way exacerbate the disease process.
449
Thus activities such as reading may be reduced or abandoned.

Sight loss may be seen as the ultimate effect of reading in
poor light in youth or some similar long delayed retribution.
Dazzle caused by oblique light bouncing off opacities in

the media encourages people to wear a tinted filter or reduce
light overall creating extra difficulties in low-contrast situ-
ations. In countries like the United Kingdom where medical
help is free but doctors may be overworked and busy, elderly
people see their complaints as relatively trivial and there-
fore, "do not bother the doctor". In other countries medical
assistance may be expensive, and therefore avoided for that
reason. The threat of costly new spectacles is another deter-
rent to those on a limited income. There is also an ostrich
attitude of not wishing to hear possibly bad news.
However, without doubt the largest single factor prevent-

ing elderly people seeking help with deteriorating vision is
that they are so overwhelmed with other problems, such as
general illnesses, that it is only when poor sight is the major
or only handicap that a consistent effort is made to seek help
[3]. Cullinan reported that many housebound elderly people
who are in regular contact with a general medical practition-
er or receiving frequent visits from nurses or social workers
may consider that it is inappropriate to discuss a sight prob-
lem with anyone other than an ophthalmologist or optome-
trist. The situation is, therefore, that many elderly people
are isolated, have a number of physical and financial prob-
lems of which diminished sight is perceived as a relatively
unimportant one, and are not in contact with eye care profes-
sionals who they see as the only people able to tackle what
are probably insoluble problems.
If methods could be ascertained to predict relatively ear-

ly which people are likely to need help later, problems could
either be avoided or alleviated. It was to this end that the
City Study was established by a team from St. Bartholomews
and Moorfields Eye Hospitals, both of which have close con-
nections with the City of London.
450
3. Subjects and Methods
Large employers in the London area were approached and,

with the co-operation of the medical officers, individuals who
were nearing retirement and their partners were invited to
participate in the study. The purposes of the study, i.e. to
"predict and prevent eye problems in retirement", were made
very clear but no treatment was promised although if previ-
ously unrecognised abnormalities were found, volunteers
would be directed to the appropriate source of help. Volun-
teers would be seen at intervals of three years for twelve
years.
Before attending, the volunteers completed a detailed

questionnaire, including medical, optometric, alcohol, tobac-
co, drug, hearing and family history. Details were requested
of working career, travel etc. Subjects were characterised by
age, sex, social class, eye color and ethnic background. At
each visit the volunteers are seen by an optometrist, a physi-
cian and an ophthalmologist. The following data was collect-
ed at the first visit:
The optometrist was the first member of the team to see

the subject. Current spectacles were examined and dated.
Visual acuity for distance and near was recorded and best vis-
ual acuities obtained. Where near vision was less than N5
with a standard +4.00 dioptre addition, the necessary mini-
mum magnification level for N5 was obtained. (Distance
vision was measured indirectly at 6 metres with a back-lit
standard Snellen chart, near vision with a Faculty of Ophthal-
mologists Times New Roman reading chart illuminated with a
standard tungsten bulb to between 500 and 600 lux [4]. Cen-
tral visual fields were plotted using a Freidmann Mark II Visu-
al Field Enhancer.
With the patient, the physician reviewed the question-

naire, and expanded details where needed. He took resting
blood pressure and fasting blood samples. The blood was later
analysed for sugar, cholesterol and triglycerides.
The ophthalmologist assessed the intraocular pressure

using a Goldmann applanation tonometer, and completed a
full dilated pupil slit-lamp examination of the lens and fun-
dus. Photographs were taken of the disc and macular areas.
451
It is planned that the above procedures be repeated at

intervals of 3 years. All the data is being coded and analysed.
The first follow up is due to commence in the summer of
1986.
4. Findings
Our volunteers were drawn from every level of large organi-

sations including The Post Office, London Transport, British
Telecom, the Inland Revenue and Marks and Spencers.
In all 1080 volunteers were seen between June 1982 and

October 1985. A few people, mainly spouses, were seen who
were outside the preferred age range of 55-65 years - their
data has not been considered. And some results were spoiled
by, for example, bad film or occasional staff lapses.
Several of the organisations involved have a retirement

age of 60, therefore of our sample 72% were aged 54-59
years, and only 28% were 60-65. Forty three percent were
women, 57% men. Thirteen point eight percent, mostly sen-
ior managers, were classified as Social Class I or II, 40% were
Class m non-manual or clerical, 10% Class IV skilled manual,
24.6% Class V semi-skilled manual, and 11.6% Class VI and
VII labourers [Fig. 1]. Thus the cohort included manual work-
ers, senior managers and clerical staff. The distribution
reflected the urban nature of our sample. There was rela-
tively little previously recognised eye disease, for example
0.7% had been treated for glaucoma (although 4.7% had a
first degree relative with the disorder).
Thirty five per cent had never used tobacco, and 37% had
abandoned the habit, but 14% smoked 15 or more cigarettes
per day, and 14% less than that. Of those who had smoked at
one time, 18% had been heavy or fairly heavy smokers, and
19% light smokers. Of the entire cohort 36% claimed to
drink alcohol very occasionally or not at all, 55% were light
drinkers (1-2 British measures/day), 6% moderate drinkers
(3-8 measures), and 3% heavy drinkers (over 8 measures).
One British measure is the equivalent of a glass of wine or a
gill of spirits. Seventy six per cent reported "good vision" in
childhood, 16% declared that they had been "longsighted", or
"shortsighted", but there seemed to be a good deal of confu-
452
OCCUPATION
SOCIAL CLASS
.1&2
24·60%
.405
03
06&7
CITY 86
Figure 1. Breakdown by social class.
sion about the actual meaning of the two terms. The other
8% either did not respond or specified another condition.
One individual claimed never to have seen an optometrist

or ophthalmologist, and 40 (4%) do not now consult one at all,
but 40% have an eye examination every 2 or 3 years, 17%
more frequently, 19% every 3 to 4 years, and 20% even less
frequently (Fig. 2).
The majority of the sample (54%) had started wearing

glasses before they entered the presbyopic age group,
although 32% started at around 50 years (Fig. 3). Another
peak occurs in the teens and early twenties.
4.1 Physician's Findings
Two point three per cent were diabetic, and 2.5% had diastol-
ic blood pressure of over 100 (Fig. 4).
One per cent had a fasting blood sugar of more than 5

mmols/litre and the serum cholesterol and triglycerides
placed the cohort between the levels quoted by SHAFER [5]
in Finland and Japan.
453
EYE EXAMINATIONS
CITY 86
Does
not gO yrs yrs yrs yrs yrs yrs+
Figure 2. Frequency of eye examinations.
4.2 Optometric Findings
In the younger (54-59 years) group 94.5% of the total eyes

were found to have normal vision, i.e., 6/9 (~0/30) or better
and 96% N5, with correction when necessary (Fig. 5). (Of the
group just one man, aged 58, achieved these levels without
any correction; both eyes were emmetropic for distance, and
pupils were normal.) In the 60-65 group 92% were 6/9 (20/30)
or better and again 96% N5. If visual disability is defined
according to World Health Organisation recommendations [6]
then 3.6% of the younger group and 4.2% of the older would
have significant visual impairment if dependant upon that
eye, although nearly all of them had sufficient vision in that
eye to read newsprint given appropriate low vision aids.
4.3 Fields
The data are not yet available.

454
200
180
160
140
120
100
80
60
40
20
o
5 10 15 20 25 30 35 40 45 50 55 60 65
Figure 3. When first wore glasses.
4.4 Ophthalmological Findings
There was a high prevalence of lens changes rising from 23%

in the age group 54-59 years to 28% in the age group 60-65
years (Fig. 6). In the majority changes were bilateral. There
were slightly more cortical than nuclear cataracts, and the
difference increased in the older group. In many cases chang-
es were slight and did not significantly reduce acuity (Fig. 7).
There is a positive association between smoking ciga-

rettes and nuclear opacities (but not the cortical type). It
applies only to present moderate (15-24 cigarettes per day) or
heavy (25+ per day) smokers, or previously heavy smokers
(Fig. 8). The association is not confounded by other variables
such as the use of beta blockers or steroids. Of the 162 eyes
in which changes in the nucleus was observed, 20% had a visu-
al acuity of less than 6/9 (20/30), 34% were 6/9 (20/30), and
the remainder were 6/6 (20/20) or better. The retinal find-
ings are currently being coded, and the preliminary findings
suggest that drusen were present in 4.4% of our subjects who
455
180
160
140
120
100
80
60
40
20
o ,
10 20 30 40 50 60 70 80 90 100 110 120 130 140
Figure 4. Diastolic blood pressure.
are therefore at risk of developing disciform macular degen-

eration [7]. However, there is some uncertainty about the
consistency of the coding, and this figure may be modified.
The intra-ocular pressure was over 20 in 14% of the sam-

ple (Fig. 9).
456
54 - 59
R
58 o 6/ S - 6/6
0 6/9
100
06i12
12 5 10 4
0 ml 6/ 18
AGE
.0 6/36
6124
SUO
60 - 69 0 6/ 60
400 . U;SSTHAN
6 / 60
JOO
211 215
200
100
9
0
AGE
Figure 5. Corrected visual acuity and age.

457
500 fI1 NONE

400
EJ NUCLEAR
fm)CORTICAL
300 Ell POSTERIOR
SUB CAPSULAR
200
100
o
54 - 59 AGE 60 - 65
Figure 6. Lens changes (numbers of people) by age

458
LENS OPACITIES
NONE + NUCLEAR CORTICAL POSTERIOR MORE THAN TOTAL

CONGEN SCL + SUB - 1 TYPE
MATURE CAPSULAR
VISUAL ACUITY
(DISTANCE)
6/5 - 6/6 77 .0% 9.2% 9.0% 1.1% 3.8% 100.0%
6/9 69.8% 11.8% 8.7% 2.2% 7.6% 100.0%
6/12 47.8% 21. 7% 15.2% 6.5% 8.7% 100.0%
, 6/18 36.8% 31.6% 5.3% 26.3% 100.0%
6/24 36.4% 18.2% 45.5% 100.0%
6/36 95.2% 4.8% 100.0%

6/60
6/60 64.9% 10.8% 10.8% 2.7% 10.8% 100.0%
Less than 6/60 100.0% 100.0%
TOTAL 73.7% 10.0% 9.1% 1.6% 5.5% 100.0%
Figure 7. Lens opacities by visual acuity (all eyes)
6/6 6/9 6/12 6/18 6/24 6/36 6/60 <6/60 TOTAL
SMOKING HISTORY
NON SMOKER
QUANTITY SMOKED
NONE 65.3% 27.5% 2.0% .6% 1. 7% 2.3% .6% 100.0%
PAST SMOKER
QUANTITY SMOKED
1 - 14 72.9% 24.0% 2.1% .5% .5% 100.0%
15 - 24 58.8% 33.6% 2.5% 1. 7% 1. 7% 1.7% 100.0%
25 + 62.5% 27.1% 2.1% 2.1% 2.1% 4.2% 100.0%
PRESENT SMOKER
QUANTI TY SMOKED
1 - 14 61.5% 31.8% 3.4% .7% .7% .7% 1.4% 100.0%
15 - 24 63.8% 30.5% 2.9% 1.0% 1.9% 100.0%
25 + 48.5% 39.4% 3.0% 3.0% 6.1% 100.0%
TOTAL 64.6% 28.9% 2.4% .8% .3% .9% 1.7% .4% 100.0%
The City Eye Study 1985
Figure 8. Visual acuity (Right Eye) by smoking

459
[]1H3
160 156 19 14- 16
140 0 17- 19
120 0 20-22
100 0 23 - 25
60 0 26 - 28
60 0 29 +
40
20
0
54·59 AGE
160
140
120
100
80
60
40
20
60 - 65 AGE 60 - 65
Figure 9. Intra-ocular pressure and age
5. Discussion
The connection between smoking and lens changes is the first

important finding in the City Study. It will add another point
to the already strong argument of the anti-smoking lobby.
Disciform macular degeneration is increasingly a treatable,
or at least an arrest able disorder particularly if diagnosed
early [8], and an at-risk group has been identified.
Although a large proportion of people do have regular eye

examinations, a significant number do not, and thus a possible
first-line screen is not always present. Recent legislation has
de-registered dispensing opticians in the United Kingdom, and
thus made it easier for people to obtain spectacles without
consulting a qualified person. Most of this first review took
place under the previous system, when despite the availability
of free eye examinations {indeed the eye examination is still
460
free), many of our volunteers are content to keep a prescrip-

tion for many years.
The fact that more than half our group were wearing
spectacles pre-presbyopia suggests that emmetropia is rela-
tively unusual. The peak in early adult life presumably repre-
sents the majority of the myopes. Quite possibly with over
50% being non-manual relatively small hypermetropic and
astigmatic errors cause asthenopic symptoms, but these cor-
relations are not yet available.
The Study is having an educational function of its own. In

the relaxed and informal atmosphere that prevails in the Eye
Department of Barts during the sessions our subjects ask
questions very freely. Any self-selected sample must be
viewed suspiciously, and it might be argued that our findings
may be influenced by this factor. Certainly it must be con-
ceded that these volunteers, feeling that they have access to
sympathetic and interested professional advice, may well be
more likely to seek help early. The relatively low prevalence
of pre-existing eye disease suggests that few if any volun-
teers were hoping for "extra" advice or treatment.
Inevitably with a study of· this nature an enormous

amount of data becomes available, and any analysis of it
tends to pose more questions than there are answers.
6. Future Papers
Full accounts of the findings on smoking and lens changes
1. from the epidemiological viewpoint (Cullinan/the

City Study Group) is in preparations, and
2. another from the ophthalmological position (Flaye/

The City Study Group, 1986) is in press.
3. a fuller investigation of the refractive errors against

age, previous correction, and visual acuity is present-
ly being coded, as is the prescription form (single
vision, bifocals, etc), which should give useful infor-
mation both to the concerned professions and the
industry.
461
4. the retinal data is being examined for possible associ-

ations
5. an analysis of the results of referrals is in prepara-

tion
6. any other association are being actively sought.
It is however, the long-term results which will be most

interesting and most valuable.
7. Acknowledgements
I am grateful to the other members of the City group (T.R.

Cullinan, D. Ehrlich, D. Flaye and K. Sullivan) for their gen-
erosity with data, encouragement, and criticism - especially
the last; to the Medical illustration Department of Moorfields
Eye Hospital for graphs etc; and to Carol Clarke for clerical
support.
8. References

England and Wales, 1948-1962, Reports on Public
1966)
2. J.C. Brocklehurst, A.N. Exton-Smith, S.M. Lempert

Barber, L. Hunt, M. Palmer: Fracture of the Femoral
Neck, Two Centre Survey of Aetiological Factors, No.
1 (Departments of Geriatric Medicine, University Hos-
pital of South Manchester and University College Hos-
pital, London 1976)
3. T.R. Cullinan: The Epidemiology of Visual Disability.

Health Services Research Unit, Report 28 (University
of Kent at Canterbury, 1977)
4. F.W. Law: Standardization of reading types. Br. J.

Ophthalmol. 35, 765 (I 951)
462
5. A. Shaper: Ischemic heart disease - epidemiology and

possibilities for primary prevention. Med. Int. 826
(1985); J .H. Silver, E.S. Gould, D. Irvine, T .R. Cullinan:
Visual acuity at home and in eye clinics. Trans.
6. World Health Organization: The Prevention of Blind-

ness, WHO Technical Report Series, No. 518 (WHO,
Geneva 1973) p. 10
7. A.C. Bird: Recent advances in the treatment of senile

disciform macular degeneration by photocoagulation.
Br. J. Ophthalmol. 58, 367 (1974)
8. The Moorfields Macular Study Group: Treatment of

senile disciform macular degeneration: A single-blind
randomised trial by argon laser photocoagulation. Br.
J. Ophthalmol. 66, 745 (1982)
The Silver Pages: Are They Easier to Read?
Denise A. DeSylvia, Thomas R. Corwin
1. Introduction
In recent years considerable attention has been focused on

the increasing number of people over the age of 55. Current-
ly over twenty percent of the population of the United States
is over 55. By the year 2020 the number is expected to be
twenty-five percent [1]. Government agencies, housing
projects, corporations and private organizations are among
the vast number of groups striving to meet the needs of this
group. The Silver Pages, a telephone directory listing servi-
ces and products for the elderly, produced by Southwestern
Bell Publications, is an example of a publication specifically
designed to be used by adults over 55 years of age.
Any printed material requires a high degree of visual

function, especially telephone directories. With this in mind,
Southwestern Bell Publications elected to use larger print.
Discussions with the company's research department indicate
that this was the extent of the consideration given to lessen-
ing the visual demands of the directory [2]. As well, elderly
citizens were asked to comment on the visibility of various
advertisements. This was prior to selling space to vendors
and service providers and was used more for marketing than
determining readability.
Because the Silver Pages is a publication specifically for

use by the elderly, of whom a significant number have some
reduction in visual function, we thought it important to assess
the book and test if it is easier to read. We wanted to evalu-
ate print size, print contrast and readability in comparison to
a standard Yellow Pages directory.
464
2. Subjects
Subjects were between the ages of 24 and 83. Ten were over
55 and eleven were under 55. Two subjects, one age 47 and
the other age 83, had moderate low vision and used hand mag-
nifiers during the testing. The 83 year old is a patient at the
New England College of Optometry and the other is a staff
member of the Vision Foundation in Watertown, Massachu-
setts. Of the remaining subjects all had 6/7.5 (20/25) or bet-
ter acuity at near and far. SUbjects were either New England
College of Optometry staff, students and faculty or clients of
the Somerville 60+ Health Center in Massachusetts. Not all
subjects in the 55 or over age category performed all tasks
due either to time constraints or fatigue. Nine subjects did
Task One, six did Task Two, five did Task Three and seven did
Task Four.
3. Testing Materials
A standard Yellow Pages directory was adapted so that it was

the same size and weight as the Silver Pages. Each book's
print size and ink contrast were measured. The findings are
in Tables 1 and 2. The print was measured with a standard
millimeter rule under 12X magnification using a CCTV. Two
measurements of each type print were taken. The first meas-
urement was the actual letter size. Capital letters "T" and
"L" were used and lower case letters "a" and "0" were used.
From this a Snellen equivalent was calculated as well as the
visual angle. The second measurement was the number of
characters per inch. This value was divided into one thou-
sand, giving another Snellen Acuity. It is interesting to note
that the visual angles measured in the Silver Pages approach
the minimum visual angle for reading efficiency as deter-
mined by LEGGE et al.[3]. The minimum is 0.3 degrees. Let-
ters from the Yellow Pages have a visual angle of much less
than the 0.3 degrees needed for efficient reading. Areas of
grey and black inks were evaluated for contrast. Contrasts
for black ink on paper were similar, the Yellow Pages having
seventy-nine percent contrast and the Silver Pages having
seventy-eight percent contrast.
465
Table 1
Book Characteristics
Yellow Silver
Size 9 in x 10 3/4 in 9 in x 10 3/4 in
Weight 16 oz 16 oz
Thickness 12 nun 11mm
Paper color yellow buff (c ream)
Avg background
20.95 ft-L 21.43 ft-L
(paper) luminance
Avg black (print)

4.38 ft-L 4.68 ft-L
luminance
Avg grey (artwork)

12.48 ft-L 17.08 ft-L
luminance
Avg red (artwork)

luminance 15.25 ft-L red not used
Black ink contrast 79% 78%
Grey contrast 40% 20%
Red contrast 21% not used

466
Table 2
*All print was measured in mm under 12X magnification using

a CCTV. **Rule of 1000 - number of letters per inch divided
into 1000. *** Visual Angle of letter when viewed at 40 cm.
Yellow Silver
INDEX ~ Lower case All caps

*Measured 1. 75 1.25 2.00
***Visual Angle .25 0 .179° .286 0
Snellen Acuity 20/60 20/43 20/69
**Rule of 1000 20/45 20/65
Tele;2hone numbers
Measured 1. 75 2.75
Visual angle .25 0 .394°
Snellen Acuity 20/60 20/95
Rule of 1000 20/48 20/71
General Listings - Names All caJ2s I both books

Measured 1. 75 2.75
Visual Angle .25 0 .394°
Rule of 1000 20/40 20/82
General Listings - Addresses
~ Lower Case ~ Lower Case

Measured 1.5 1.25 2.0 1.5
Visual Angle .215 0 .179° .286° .215°
Rule of 1000 20/38 20/52
4. Procedure
One half the subjects in each age category were tested first
on the Yellow Pages for all tasks and the other half were
tested first with the Silver Pages. Subjects used available
light and were allowed to use any spectacles, bifocals or low
vision aid that they wished. Subjects read at their own pre-
ferred working distance.
1. Task One: Indexes from each directory, using the

words only, were read by all subjects. The listings
contained 3124 characters each. Spaces between
467
words were counted. Subjects were evaluated by the

time it took to silently read each index.
2. Task Two: Subjects located and recorded the entire

name, address and telephone number of five catego-
ries from each directory. Subjects were evaluated by
the time it took to record the listings. The total
number of characters which the subject was required
to copy was the same for the Yellow Pages as the Sil-
ver Pages, and consideration was given to the sUr-
rounding material or layout.
3. Task Three: Subjects located five listings from

each directory. The listing located in the Silver Pag-
es was the same as the one in the Yellow Pages. Sub-
jects were timed during this activity.
4. Task Four: Subjects copied a column of telephone

numbers from each directory. They were allowed one
minute to record as many numbers as possible.
5. Results
For tasks one, two and three the average times in seconds
were calculated for the Yellow Pages and Silver Pages and
the average differences were obtained. The difference
between times for each subject was calculated and a median
difference was determined. For task number four, the aver-
age number of characters was determined and an average dif-
ference and average median was determined. The median
difference and average difference were similar in both ages
for all tasks, except for task number three. The average dif-
ference was less than the median difference for the over 55
age group. A percentage change was determined for all tasks
using the median difference.
All four tasks showed a greater efficiency with the Silver

Pages in both age groups. The percentage difference for the
over 55 age group was greater than the under 55 age group
for all tasks. Task Three (locating listings) best simulates the
use of a telephone directory, and the most marked difference
between the Yellow Pages and Silver Pages was found during
this task. There was a 48.2% difference for the over 55 age
468
group. Data for all tasks are presented in Table 3. The per-
centage change for all tasks are in Fig. 1.
50r---------.---------.----------.--------~
40
30
Percent
20
10
O~-L+
Over Over
55 55
TASK H2 TASK #3
Figure 1: Improvement
6. Conclusion
The four tasks selected indicate that the Silver Pages are
easier to use. The contrasts of each book are similar and the
sizes and weights are the same. The major difference is the
print size. Tasks One and Two involve only print and its read-
ability. The Silver Pages directory was more efficient for
both age groups, the over 55 age group more so than the
under 55. The Yellow Pages are below the 0.3 degree mini-
mum visual angle for optimum reading. LEGGE et al.[3] con-
cluded that for letter sizes below the minimum angle, the
decline in optimum reading efficiency is probably a factor of
acuity limitations. Since the print size is the major differ-
ence between books, it is not surprising that for tasks involv-
ing print readability only, the Silver Pages are more efficient
and that subjects under 55 do better than over 55. Subjects
under 55 still may have the ability to accommodate and can
increase the visual angle by bringing the material closer.
469
Tasks Two and Three had larger percentage differences

between the Silver Pages and the Yellow Pages. Again print
size is likely to be a significant factor, but this task also
included location of an item. The Yellow Pages has more
listings under each category. This makes the task more com-
plex and print size is no longer the only factor. Overall the
Silver Pages is more efficient than the Yellow Pages. When
print size is the most significant difference between books,
the increase in efficiency is modest. When a location task is
involved, the differences in efficiency are considerably great-
er.
Table 3
Results
Median Median Percen t

Yellow Silver Difference Change
TASK III Average time in seconds:
Under 55 133.91 124.45 9.45 3.1
Over 55 159.38 139.44 24.67 6.0
TASK 1/2 Average time in seconds:
Under 55 304.45 271.45 60.18 20.00
Over 55 440.2 347 93.20 23.7
TASK 113 Average time in seconds:
Under 5S 170.73 97.82 73.82 35.7
Over 55 217.80 106.00 111.80 48.2
TASK 114 Average number of characters per minute:
Under 55 101.27 107.18 7.73 1.9
Over 55 70.13 79.25 9.13 12.00

470
7. References
1. C. Kirchner, C. Lowman: Elderly blind and visually

impaired persons: Projected numbers in the year 2000.
J. Visual Impairment & Blind. 73,69 (1979)
2. Conversation with R. Blake and M. Shuck, Research

Department and Public Relations, Southwestern Bell
Publications (1985)

25, 239 (1985)

25, 253 (1985)
A Hierarchy of Perceptual Training in Low Vision
Jocelyn Faubert, Olga Overbury, Gregory L. Goodrich
1. Introduction
A growing concern in low VISIon care is whether people

afflicted with a visual impairment can adapt to their condi-
tion and relearn to perform lost functional abilities. A purely
sensory-physiological approach to this issue is restricted
because 1) low vision patients often have below what is nor-
mally assumed as the basic necessary sensory input for many
functional tasks (e.g. reading) and 2) in many cases, such an
approach assumes a lack of plasticity past a critical period of
acquisition. An alternative approach is that there is some
useful plasticity or ability to relearn at all ages even though
they may differ quantitatively and/or qualitatively.
In 1950 GIBSON [1] summarized many studies demon-

strating that perceptual judgement tasks can be improved
either by practice or training for adult observers. Other
reports in the perceptual literature show that the visual sys-
tem has a remarkable ability to compensate for artificially
distorted images [2,3]. The low vision patient is often faced
with the dual task of coping with the onset of visual loss and
the adaptation to visual aids. A recent report [4] has demon-
strated that an individual, restored from blindness in adult-
hood, can readjust and acquire limited perceptual abilities at
a time in life which is long past the critical period of child-
hood visual development. It is also well accepted in the field
of low vision, though not well researched, that some training
programs can help a number of individuals to use visual aids
more efficiently and perform perceptual tasks with better
ease. In the late 1970's GOODRICH and coworkers [5,6] dem-
onstrated with a variety of techniques that low vision
patients can improve the use of their residual vision.
QUILLMAN et al.[7] demonstrated that the Frostig Fig-

ure Ground test was a useful predictor of reading efficiency.
This prompted researchers from our laboratories to look into
the efficiency of this test in predicting perceptual function
capabilities under training and non-training conditions [8,9].
472
An interesting finding was that the Frostig Figure Ground was

a good predictor of how well patients will adapt to their visu-
al aids. For instance two females of the same age with visual
acuities of 6/21 (20/70) in the better eye and under 6/60
(20/200) in the worse eye, who were prescribed the same cor-
rection upon a clinical assessment (+8.00 D spectacles),
scored very differently on the Frostig. The one who scored
best came back a few weeks later very satisfied with the pre-
scription while the other was quite dissatisfied with the aid.
A possible explanation for these results is that some kind of
functional hierarchy is present where the ability to perform a
"lower" level visual task is a precondition and/or a predictor
of "higher" levels of visual performance.
BARRAGA and her colleagues [10,11] have attempted to

systematize the study of visual impairment in childhood
development. They proposed a hierarchy of perceptual devel-
opment in the normally sighted child. They also state that
the visually impaired individual follows the same pattern but
is slower to progress from one level to another. The question
we asked was whether a hierarchical model such as the one
proposed by Barraga and her colleagues has any implications
for the impaired adult visual system and whether it could
explain some of the results mentioned above. To determine
this the following steps are warranted. First, the develop-
ment of testing and training materials which are specific to
the different levels. Second, it is important to determine
what level or levels of the visual hierarchy are affected by
visual pathology. Thirdly, whether such a functional model is
sequential in the adult system is of theoretical and practical
interest. In other words, if an observer is functionally
impaired at level 5, does this presuppose a functional impair-
ment at all subsequent levels or is it level/task specific? If
sequential, it is theoretically possible that training one level
may improve the adjustment of subsequent levels.
Recently an attempt has been made to develop testing

and training materials which represent 8 stages based on
BARRAGA's hierarchy [12,13]. This report is a description of
the testing materials developed so far and some preliminary
findings. The following stages represent the hierarchy in
question:
1. Visual attention: the ability to localize visual targets

is tested.
473
2. Efficient eye movements: the patient's ability to

track a moving light is tested.
3. Manipulation of concrete objects to match model: an

example of this kind of task can be represented by
the block design subtest of the Weschler Adult Intel-
ligence Scale (WAIS).
4. Copy/draw shapes from a model: stimulus configura-

tions from the Bender Gestalt test of visual percep-
tion and similar test patterns can be used.
5. Match single element picture to complex picture:

elements such as the picture completion subtest from
the WAIS can be used in a modified fashion so that
the test is multiple choice.
6. Figure-Ground discrimination: the baseline measure

for this level is the Frostig Figure Ground test.
7. Letter and word recognition: materials have been

specifically developed for the last two levels.
8. Reading efficiency.
2. Methods and Results
The first two levels of the hierarchy are essential for any
progress in functional vision to occur. What is assessed by
these levels is the patient's ability to perceive gross targets
such as a card or bright light. On the second level a basic
assessment is made of the ability of the patient to follow the
target mentioned. The task does not involve sophisticated
eye movement analysis such as saccades but rather the test is
one of smooth pursuit. The question is whether there is
enough light perception and visual stability to perform this
task. The great majority of the people seen so far can per-
form these tasks without great difficulty. The third level is
presently in development but a task such as the block design
subtest of the WAIS fits the description of this level well.
The unique difficulty present at this level is the concept of
3-dimensions and handling of concrete objects on several
planes simultaneously. It is often the case in low vision
474
patients that binocular vision is not the best functional vision

to use. Perhaps assessment of depth perception may be
required for this level.
From the third level on, many sublevels of difficulty can

be envisaged in the hierarchy. For instance at the fourth lev-
el (copy/draw), the reproduction of a dot pattern (see Fig. 1)
was designed along with a straight line pattern demonstrated
in Fig. 2 and part of the Bender Gestalt test (see Fig. 3) has
also been used.
Sample COpy IDRA \V Testing Materials

Fill in the bottom grid t.o match the model grid on top.
• 0 0
•
0
• • 0
0
• • 0
• 0 0
•
0 0 0 0
0 0 0 0
0 0 0 0
0 0 0 0
Figure 1. Dot matrix configuration task from the copy/

draw level. The observers are asked to repro-
duce the pattern seen in the upper grid in the
bottom one.
475
Figure 2. Straight line configuration task from the copy/

draw level. The observers are asked to repro-
duce these 4 patterns on a separate sheet of
paper.
Clearly the difficulty level is different for each test pat-

tern. This represents one of the main difficulties in designing
testing and training materials of different levels. A concen-
sus of what is difficult or not warrants a large amount of data
collection.
Several testing materials have been developed for the

fifth stage of the hierarchy. Figures 4, 5, 6 and 7 demon-
strate some of the tasks of various difficulty.
The bars test (Fig. 4) should be the easiest to perform

followed by the target match (Fig. 5) and the tests derived
from the WAIS picture completion test which we transformed
into a multiple choice task (Figs. 6 and 7).
476
Figure 3. Bender configuration task from the copy/draw

level. The observers are asked to reproduce
these 4 patterns on a separate sheet of paper.
The sixth level inevitably involves the Frostig Figure

Ground test mentioned previously. An example of a different
type of figure-ground task is given in Fig. 8. This is a sample
of the usual test which contains 25 numbers.
The task is to trace with a pencil from one circled num-

ber to another. At a given time the target number becomes
the "figure" and the surround is the "ground". The "figure"
and "ground" components in this test interchange roles during
testing.
Figures 9 and 10 demonstrate letter recognition tasks and

Fig. 11 is a word recognition task.
In the letter recognition the observer is requested to

identify a target letter, for example the letter "i" in Fig. 9
and "y" in Fig. 10. The word tasks require the observer to
find every word in the sentence within the bottom paragraph
of meaningless words (Fig. 11).
Preliminary data in the form of Spearman Rho correla-

tions of 25 patients who had completed most of these tests,
demonstrate weak correlations except for the conditions of
--
477
--
--
--- --
-- ---
--
-- --
--
Figure 4.
-- -
Missing line exercise from the target match
level. The observers must identify the pattern
in the 4 bottom choices which match the top
pattern.
Frostig Figure Ground and all the tests in levels 7 and 8 which
produced correlation coefficients between +0.56 and +0.80.
In other words, we confirmed the results obtained from previ-
ous studies stating that the Frostig Figure Ground is a good
predictor of reading performance and related tasks. The pro-
duction of testing materials is not complete and final conclu-
sions on the validity of each measure and how they relate is
premature.
478
Sample TARGET MATCH Training Materials: 4M Size
The tuk Js to match the sample figure in the left colUmD with tbe aame figure iD
tbe samples on tbe right.
EB e ® 0 EB
c c- C ::J C
-
/- -l ~ I /-
u u u n n
-< -<
- >- >-
- -<
+-- -t / ~ +--
6 V' V 6 /\
E E C :3 ::J
--- ,......., ......... ---
,......., ..-.. ~
Figure 5. Symbol tasks from the target match level. The

observers must identify the correct symbol in
the 4 choices on the right to match the target
symbol on the left (from GOODRICH and MEHR
[14]).
479
Figure 6. Flower pattern from the picture completion

test (WAIS) modified to allow a multiple-choice
response for the target match level.
480
Figure 7. Car model from the picture completion test

(WAIS) modified to allow a multiple-choice
response for the target match level.
o
o o
o
finish
Figure 8. Sample of tracing number task for the figure-

ground level. The observer must trace from one
number to another. The actual test contains 25
numbers.
481
Patient Namt' or Code: _ _ _ _ _ Date: _ _
Low Vision Clinic (check 1):
II Palo Alto [ I Berkeley [ I Houston [ ) New York II Montreal
Find the "i's":
111111 11111 IiIII 1111 IIllil lllli illll llill 1111 llil
Ilill 11111 III 1111111 lill lllill III 1llliII III 11111
III ill llllil 1111 lilll 1111 111111 lill 1111 llli 1111
liIIII llill 111111 llllil 1111 llli 111111 1111 illll III
Number of "j's" found: _ _
Time to complete: _ _
Comments:
Figure 9. Example of letter search task. Observer must

identify the "i"s embedded within the "l"s.
482
Patient Name or Code: _ _ _ _ _ _ Date: _ _
II Palo Alto II Berkeley II Hou,ton II New York II Montreal
Find the "y's":
ggggygg gggg ggy gggg ggygg ggyggg

gyggg ggy gggggg gyggg gggygg gggg
ggggg gygg gyggggg ggggy gyg ggggg
gggg gggggy gygg gggygg gggg gyggg
l'\umber of "y's" found: _ _
Time to complete: _ _
Comments:
O.2m
Figure 10. Letter search task where the observer must

identify the "y"s embedded within the "g"s.
483
Get some extra milk, eggs, and fish

w hen you shop today.
Got Get Go gone some soon extra

exert melt milk malt eyes eggs east
and end fish fine what where when
you year shop ship today toast
Min _ Sec_
Jack, who always cries, and his

friend, who doesn't, were lost.
Jack Joke Jake who why how away

always after cry cries crisis and
this her his kind friend find who
don't doesn't where were last lost
Min _ Sec_
Figure 11. Example of word search task. The observer
must identify the words in the nonsense para-
graph below which make up the target sen-
tence.
3. Discussion
Testing low vision patients with evaluative tools such as the

ones just described can provide insights for future training
designs on the basis of the different approaches taken by the
observers. An example of this can be given by comparing the
performance of two patients whom we will identify as X and
Y. X and Yare both males of approximately the same age
484
with 6/120 (20/400) in the better eye. X had a good Frostig

result and Y had a bad result. The approach taken by X could
be identified as "thinking his way through". When tracing a
star he assumed correctly that there were so many even
points and thus could trace more appropriately. It was clear
that Y could not adjust to the Frostig task. When they were
tested on the letter and word recognition tasks, patient Y
could not do them at all. The results of patient X are demon-
strated in Figs. 12, 13, 14 and 15.
Patient Name or Code; _ _ _ _ _ _ _ Date: _ __
[ I Palo Alto [ I Beckeley [ I Hou,ton [ I New York [ I Montreal
Find the "i's":
111111 11111 Qill] llll llllil lllli @IDllill llli QlID

~lllH 11lIll1fill@lllilllllllllillllllllll
lll(i!l)llilil 1111 lilll 1111 111111~111(]]J])IIJI
lillllllillllllll1111il llU@TI)llllll 1111 illll 111
Number of "j's" found: _ __
Time to complete: _ __
Comments:
Figure 12. Results of patient X in the "i" letter search

task.
A demonstration of his strategy for identifying the "y"s

was shown to us by the patient and can be seen in Fig. 14. He
would look for the "tail end" of the "y" which was very differ-
ent from the "g". In the word recognition task it is clear
from Fig. 14 and 15 that patient X was tracing from one con-
sonant to the other such as a ltd" and a "t", to determine the
correct word. For example, when searching the word
485
Patient Name or Code: _ _ _ _ _ _ _ Date _ __
Low Vision Clinic (check I):
! I Palo Alto II Berkeley [ 1Houston! I New York [ I Montreal
Find the "y 's" :
ggggygg gggg@gggg ggyg@ggg

gyggg@gggggg@gggMgg gggg
ggggg@gg@gggg gggiY@g ggggg
gggg gggggy(fi'gg~g gggg@ggg
Kumber of "y's" found: _ __
Time to complete: _ __ y) 880

Comments:
- O.2m
Figure 13. Results of patient X in the "y" letter search

task.
"doesn't" in the phrase, a horizontal line was drawn between

the "d" and "t" of "don't" and "doesn't" which, by the differ-
ence in length, helped recognition of the appropriate word. X
also was able to read the standard text in 9 minutes and 31
seconds.
What this tells us is that 1) an individual who theoretical-

ly should not have been able to read was able to do so with
highly strategic maneuvers, and 2) training observers to be
more attentive to pattern details of words may enhance their
performance. Ideally, after systematic training this could
become second nature and not necessitate conscious attention
on the part of the reader.
An attempt was made to describe the ongoing activities

of a long term joint project between two low vision service
centers. It is hoped that a systematized approach to testing
486
Get some extra milk, egg~, and fish

when you shop today.
Min _ Sec_
Jack, who always q.-j~s, and ~ji?

friend, who doesn't, were lost.
~Joke Jake~w4how away

~a~er cry(&ri~~--Erisis~
this he hi_skince-~nd~
don't ~here~last~
Min _ Sec_
Figure 14. a) Results of patient X in the word search
task.
training visual functions of low vision individuals will bring

about dividends in the future. This would enable related pro-
fessionals to understand the dynamics of low vision and to
speak "the same language" and thus improve services in this
area.
487
Dan, who made the fire, needs t'wo

or th1ree more logs.
B~bs~~howC~od~~J
~ ther@.?--..r~r~ feeds seedin:eeus)
t~es----\ to (two
~----...,
or are,--- -------.
and . four.
C .'\
~thr~ove sore~egs !~
---- Min _ Sec_
Figure 15. b) Results of patient X in the word search

task.
488
4. References
1. E.J. Gibson: Improvement of perceptual judgements as

a function of controlled practice or training. Psychol.
Bull. 50,401 (1953)
2. R. Held: Plasticity in the sensory-motor systems. Sci.

Am. 213(5),84 (1965)
3. I. Kohler: The formation and transformation of the vis-

ual world. Psychol. Issues 3, 28 (1963)
4. S. Carlson, L. Hyvarinen: Visual rehabilitation after

long lasting early blindness. Acta Ophthalmol.
(Copenh.) 61,701 (1983)
5. G.L. Goodrich, R.D. Quillman: Training eccentric

viewing. J. Visual Impairment & Blind. 71, 377 (1977)
6. J .G. Holcomb, G.L. Goodrich: Eccentric viewing

training. J. Am. Optom. Assoc. 47, 1438 (1976)
7. R.D. Quillman, E.B. Mehr, G.L. Goodrich: Use of the

Frostig Figure Ground in evaluation of adults with low
vision. Am. J. Optom. Physiol. Opt. 58, 910 (1981)
8. O. Overbury, B. Conrod: Perceptual, psychophysical

and functional correlates of success with low vision
aids. Paper presented at the Annual Conference for
the American Association of Workers for the Blind,
Orlando, Florida, July 1982
9. O. Overbury, B. Conrod, M. Trudeau: Perceptual train-

ing in low vision. Paper presented at the Annual Con-
ference for the American Association of Workers for
the Blind, Nashville, Tennessee, June 1984
10. N.C. Barraga, M.E. Collins: Development of efficien-

cy in visual functioning: Rationale for a comprehensive
program. J. Visual Impairment & Blind. 73, 121 (1979)
11. N.C. Barraga, M. Collins, J. Hollis: Development of

efficiency in visual functioning: A literature analysis.
J. Visual Impairment & Blind. 71, 387 (1977)
489
12. G.L. Goodrich, E.B. Mehr, O. Overbury, M.J. Harsh:

Training materials to optimize residual vision. Invest.
13. G.L. Goodrich, O. Overbury, E.B. Mehr, R.D. Quillman:

A low vision manual -- the VET Book. Invest. Ophthal-
mol. Vis. Sci. (Suppl.) 27, 107 (1986)
14. G.L. Goodrich, E.B. Mehr: Eccentric viewing training

and low vision aids: Current practice and implications
of peripheral retinal research. Am. J. Optom. Physiol.
Opt. 62, 119 (1986)
Low Vision Performance as a Function
of Task Characteristics
Shelly Marmion
1. Introduction
A crucial component of the rehabilitation process for low

vision persons is the enhancement of visual function through
the optimization of residual vision. Intervention strategies of
potential benefit are numerous. CORN [1] has proposed a
model of low vision visual functioning describing the process
as having three distinct components or dimensions, each of
which can be subject to intervention. These include (1) Visual
Abilities, consisting of the five physiological components of
vision, (2) Stored and Available Individuality, consisting of
aspects of the individual which impact on performance, and
(3) Environmental Cues, consisting of object attributes which
determine their visibility. Intervention techniques related to
this third dimension have considerable potential in terms of
both practical value and widespread applicability, being per-
haps the easiest and least expensive of interventions to
effect. However, despite this potential, much is still known
about the relative benefits of specific environmental modifi-
cations to the enhancement of residual vision, and the extent
to which facilitative effects are consistent across differing
visual conditions and tasks.
The limited number of studies investigating the effects of

environmental conditions on low vision visual performance
suggest the following:
1. There is considerable interdependence among stimulus

characteristics such that when a task is made more difficult
along one stimulus dimension (e.g. lower illumination), other
stimulus attributes become more critical to performance
[2,3].
2. Task characteristics tend to be more critical to the

visual performance of vision impaired persons [4-7].
491
3. There appear to be considerable individual differences

as to what constitutes the optimum environmental and task
conditions among the low vision population [5,8,9].
4. Relationships exist between the site of visual impair-

ment and performance difficulties on specific visual tasks
[4,10].
5. The extent to which a given environmental variable

affects performance of a low vision individual may be some-
what task specific [6-9,11].
The purpose of the present study is to investigate the

effects of several stimulus characteristics on the perform-
ance of vision impaired subjects across various tasks related
to many kinds of real world visual functioning. Several
research questions were addressed:
1. What is the relative strength of various stimulus

effects?
2. Would such a ranking of stimulus effects remain con-

stant across different visual tasks?
3. To what extent do such characteristics interact with

one another? Are such interactions consistent across tasks?
4. How do various group performances vary relative to

these variables? Are the stimulus effects the same for a low
vision group as for a group with simulated visual acuity loss-
es?
2. Method
2.1 Subjects
Two distinct populations were tested in separate phases of

the project. The first involved the testing of sighted individ-
uals under conditions of simulated visual impairment. Sighted
subjects consisted of 50 students enrolled in a general psy-
chology course (26 females and 24 males). These subjects
were tested while wearing specially treated spectacles which
simulated acuity losses within a range extending approxi-
492
mately from 6/60 (20/200) to 6/180 (20/600) and comprised

the Simulated-Loss group. The second phase involved the
testing of 43 low vision clients residing in Mississippi (25
females and 18 males), who comprised the Low-Vision group.
2.2 Tasks and Variables
Three visual performance tasks measuring aspects of visual

functioning such as visual search, pattern recognition, and
visuomotor control, were developed and administered. Tasks
and manipulated variables were as follows:
1. Landolt-C Search Task. The subject's task was to

search through an array of Landolt CIS printed on a small
card, locating CIS of a designated orientation, and marking
them. Performance was measured in terms of speed and
accuracy. The task, which relates to tasks involving visual
scanning (such as reading), was performed under three levels
of illuminance (50, 200, and 300 footcandles or 500, 2000 or
3000 lux), three levels of contrast (0.92, 0.86 and 0.72), and
three target sizes (12, 18 and 24 points).
2. Rotary Pursuit Task. The task involved tracking, with

a handheld stylus, a lighted target as it moved along a prear-
ranged pattern for 30 second trials. Performance was meas-
ured in terms of the amount of time during each trial the sty-
lus made contact with the moving target. The task was
performed at two different target speeds and the above three
levels of contrast.
3. Pattern Identification Task. Stimuli consisting of

either the letter C or E were presented singly to subjects via
a memory drum. The task was to view each stimulus and to
(a) identify the letter and (b) determine its orientation (up,
down, right, left). Viewing distance from the drum was indi-
vidually adjusted so that task difficulty was more comparable
for subjects with differing levels of impairment. Two presen-
tation modes were employed: moving-target and stationary-
target. Three contrast levels (0.92, 0.86 and 0.72), two back-
ground conditions (black and white), three stimulus sizes (12,
18 and 24 points), and three illuminance levels (50, 200 and
300 footcandles or 500, 2000 and 3000 lux) were manipulated.
Several considerations influenced the choice of variables,

the range of values, and other methodology involved in the
493
tasks. The variables themselves were chosen from among

those environmental and task variables which are reasonably
modifiable in real world work situations, in order that results
and conclusions might have practical implications for the
workplace. The range of values selected for each manipulat-
ed variable was guided by the following criteria: (1) related-
ness to real world task variable ranges (e.g. print sizes, etc.),
(2) the need to avoid ceiling or floor effects within a hetero-
geneous sample, and (3) maximization of the sensitivity of
measurement. Because a different range is optimal for dif-
fering levels of impairment, the strategy of relative place-
ment from the task was used. That is, subjects with greater
acuity impairment were placed closer to the task than those
with less, in order to approximate the same range of difficul-
ty for each subject. Although this technique allows the con-
founding of distance (and thus, viewing field angle, etc.) with
stimulus values, interpretation of results can remain straight-
forward if viewed in the context of relative differences in
stimuli, rather than absolute values.
2.3 Procedure
An estimate of each subject's distance acuity was obtained

using an Illiterate E Acuity Chart, to determine placement of
subjects in the Pattern Identification Task, adjusting task dif-
ficulty relative to acuity. Simulated-Loss subjects wore the
adapted spectacles to simulate low vision acuity.
Subjects were first given instruction and practice on the

Rotary Pursuit Task to eliminate a possible learning curve on
this task. Following this practice, subjects began the testing
sequence, which lasted approximately one hour, including a
short rest break midway through the sequence. Tasks were
divided where needed into segments of no more than 6 min-
utes in length and were presented in a counterbalanced order,
as were variable levels within tasks.
494
3. Results
3.1 Landolt C Search Task
The dependent measure consisted of a combined time and

accuracy score and was subjected to a 2X3X3X3 analysis of
variance, using the variables of Group, Stimulus Size, Con-
trast and Illumination. Significant main effects were
obtained for Group [F=7. 77(1,63)p<O.007], Stimulus Size
[F=19.19(2,126)p<O.OOI], and Contrast
[F=19.48(2,126)p<O.OOI], with better performances exhibited
for the Simulated-Loss group and under conditions of larger
stimulus size and higher contrast levels. The main effect for
Illumination was not significant, even though the plotted
function appears almost identical to that of the contrast
effect, because of the greater between-subject variability
(especially among the low vision group) for this variable.
Significant interactions between the group variable and

other variables were obtained for Group X Contrast
[F=3.91(2,126)p=O.022], Group X Size X Contrast
[F=3.25(4,252)p=O.OOI], and Group X Size X Illumination
[F=3.01(4,252)p=O.OI9], and indicated that although the Low-
Vision group appears to be more influenced by stimulus size
than the Simulated-Loss group, and less influenced by con-
trast and illumination when graphed, all effect sizes are, in
fact, statistically greater for the Simulated-Loss group
because of the greater performance variability of the Low-
Vision group. Significant interactions between stimulus vari-
ables were Contrast X Size [F=3.6(4,252)p=O.OI2], and Size X
Illumination [F=2.62(4,252)p=O.036], indicating that both Con-
trast and Illumination had differential effects at different
levels of Size.
3.2 Rotary Pursuit Task
The dependent measure was the number of seconds for which

the subject maintained contact between the handheld stylus
and the moving target. A 2X2X3 analysis of variance was
computed for Group, Target Speed and Contrast variables.
Significant main effects were obtained for Group
[F=14.03(1,87)p<O.OOI) and Target Speed
[F=129.63(1,87)p<O.OOI], with better performances exhibited
by the Simulated-Loss group and for the slower target speed.
Significant interactions occurred for Contrast X Target Speed
495
[F=6.08(2,174)p=O.003], and Group X Contrast X Target Speed

[F=4.46(2,174)p=O.003], indicating that Contrast had the pre-
dicted effect only for the slower Target Speed and only for
the Simulated-Loss group.
3.3 Pattern Identification Task
The dependent measure for this task was the number of cor-
rect responses, which were analyzed in a 2X2X3X3X3 analysis
of variance for Group, Task (identification and orientation),
Illumination, Size, and Contrast variables. Significant main
effects were obtained for Size [F=210.47(2,168)p<O.OOl], Task
[F=89.75(1,84)p<O.OOl], and Illumination
[F=22.93(2,168)p<O.OOl]. Performance was better for larger
stimuli, higher illumination, and for the identification task.
No significant differences were found for Groups or Contrast.
Significant interactions between Size and several other

variables occurred: Size X Task [F=14.48(2,168)p<O.OOl], Size
X Contrast [F=lO.24(4,336)p<O.OOl], and Size X Illumination
[F=6.48(4,336)p<O.OOl]. These indicated that, in general, no
variable effects were exhibited for the largest stimulus size,
but became increasingly more likely with smaller (more diffi-
cult) sizes. Two significant interactions between Groups and
the variables of Illumination [F=15.17(2,168)p<O.OOl] and Con-
trast [F=3.63(2,168)p=O.03] indicated that again, the stimulus
variables had a consistent effect only for the Simulated-Loss
group. A significant Group X Size X Contrast interaction
showed that only for the Medium sized stimuli did Size have a
systematic effect for both groups, and that the Simulated-
Loss group generally was more affected by the Size variable.
Separate analyses were computed for each group for the

two background conditions. For the Low-Vision group, sever-
alof the effects and interactions were slightly larger for the
black background condition compared to the white back-
ground condition, and overall performance was somewhat
superior for the black background. For the Simulated-Loss
group, performance was generally better, but more variable
for the black background, with Size and Task having slightly
decreased effects and Illumination having a greatly increased
effect on performance. Presentation modes were also com-
pared for the two groups. For the Low-Vision group, the
Moving Target condition resulted in greater errors, while the
reverse occurred for the Simulated-Loss group.
496
3.4 Legally Blind Sample
Those individuals whose functional acuity was found during

testing to be less than 6/60 (lO/lOO) were included in an addi-
tional analysis to determine whether stimulus effect patterns
were consistent for this more greatly impaired group. In gen-
eral, more errors were made by this group, but results were
consistent with those of the Low-Vision group. However, the
smaller sample size and greater group variability resulted in
smaller effect sizes overall.
4. Discussion
The low vision population is extremely heterogeneous because

of the great variability between persons on many dimensions
which undoubtedly influence visual performance. These
include etiological variables, such as the type and number of
visual disabilities, age of onset, prognosis, background and
experience characteristics, and a whole range of subject
characteristics such as personality, cognitive functioning,
motivation, etc. Such characteristics combine in almost infi-
nite variety to produce almost as many distinct patterns of
visual functioning as there are low vision persons. This con-
siderable heterogeneity becomes problematic to the study of
low vision performance, because it is difficult to identify
consistent variable effects statistically. Although the visual
performance of many low vision persons may be affected in
similar ways by specific environmental and task variables,
these consistencies are often masked by the many differences
that are also apparent in their performances. It is because of
these difficulties that the use of a sample of subjects with
simulated visual acuity losses is useful. The simulation of
reduced acuity allows the investigation of variable effects on
visual performance without a multitude of other factors cre-
ating additional "noise" which might mask such effects.
Simulated-Loss subjects are impaired ONLY by the reduced
acuity, not by multiple disabilities, not by impaired sensory
development, not by inadequate education, and so on. Thus,
the effect of a stimulus characteristic on visual performance
is a more "pure" reflection of the disability produced by
reduced acuity and the functional disability perhaps most
characteristic of low vision persons.
497
4.1 Variable Effects Across Groups
Results of the two subject groups and the legally blind subset
do not greatly differ. There are only minor instances where
the findings relative to each group are actually in conflict,
with most differences in findings being ones of degree rather
than of type. In general, performances by the Simulated-Loss
group were superior to those of the Low-Vision group and the
Legally-Blind group, probably because individuals in this
group were on the average higher functioning, had no secon-
dary disabilities, were more accustomed to performance test-
ing situations, and did not have a "vested interested" in evalu-
ation outcome.
The smaller to non-existent stimulus effects exhibited by

the Low-Vision and Legally-Blind groups are apparently due
primarily to the much greater performance variability of
these groups, both in terms of within subject and between
subject variation. It appears that environmental and task
characteristics influence the visual performance of most low
vision persons in much the same way that they influence
those with simulated impairments. However, the variable
effects are somewhat less consistent from one subject to the
next and, for any given subject, from one time to the next.
4.2 Variable Effects Across Tasks
Size proved to be the most potent of the variables, resulting

in significant main effects for all groups on each task for
which it was included. In addition, Size significantly inter-
acted with virtually all other variables, most consistently for
the Simulated-Loss group. The Contrast variable had a some-
what weaker effect on performance, and was considerably
more task dependent (i.e., the strength of effect varied from
task to task). Across groups, Contrast had the greatest influ-
ence on the Landolt-C Search Task. The Illumination variable
had a significant influence only on the Simulated-Loss group
performances, most particularly on the Pattern Identification
Task.
498
4.3 Variable Interactions
Consistency of interactions was greater across tasks than

across groups. Size interacted consistently with other vari-
ables such that as Size decreased, other variables such as lllu-
mination and Contrast increased in influence, most consis-
tently for the Simulated-Loss group. However, only half of
these interactions reached significance for the Low-Vision
group; none for the Legally-Blind group. A Contrast X Target
Speed interaction occurred on the Rotary Pursuit Task for all
groups, but yielded differing patterns for each.
When the Pattern Identification Task was varied, as with

different background conditions and presentation modes, the
strength of variable effects differed, suggesting that they are
somewhat task dependent, at least for the Simulated-Loss
group. For example, the Size effect, always large, was fur-
ther increased for the White Background condition, while lllu-
mination most influenced performance for the Black Back-
ground. Contrast most greatly affected performance on
Moving Targets, while the Size X Contrast interaction was
greatest for the Stationary Target condition.
4.4 Summary of Results
1. Subjects with simulated losses in acuity performed

better on all tasks than low vision persons, even though aver-
age acuity loss was significantly greater for this group. Their
performances were also less variable and were influenced by
task variables in more systematic and predictable ways.
2. The subset of low vision subjects who tested as legally

blind exhibited lower performances on average and greater
performance variability.
3. Group comparisons (e.g. Simulated-Loss vs. Low-

Vision) yielded somewhat different patterns of variable
effects, but differed more in terms of degree than kind.
4. Of the several stimulus and environmental variables

manipulated, stimulus size had the greatest and most system-
atic effect on performance.
5. Virtually all other variables exhibited some effect on

performance, but such effects were task specific. In general,
499
as task difficulty increased (e.g. smaller stimulus size, etc.)

the likelihood of various stimulus effects also increased.
Since all variables impacted on task difficulty, this resulted
in numerous interactions, as would be expected.
6. The relative strengths of variable effects are fairly

consistent across tasks, while absolute strengths are more
task dependent. Thus, while Size always had a greater influ-
ence on performance than contrast, the specific potency of
each depended on the task involved.
4.5 Implications for Environmental Intervention
The fact that all variables did, in one context or another, sig-
nificantly affect visual performance should be considered a
fairly convincing argument for the use of such strategies as
intervention techniques for maximizing low vision visual per-
formances. While not every variable can be manipulated for
a specific real world task, generally one or more could be
incorporated into most tasks to enhance their visibility.
Individual performances varied, exhibiting differing

amounts of reaction to specific variables, especially among
the low vision population. Thus, environmental intervention
strategies should be as flexible as possible, in order to take
into consideration the varying needs of specific clients. Opti-
mally, this would include assessment techniques to provide
information as to the specific environmental variables which
most influence a client's visual performance. However, such
assessment should be as task specific as possible, since stimu-
lus effects tended to be task dependent.
A general strategy in the absence of individual assess-

ment would be to change those characteristics which can be
modified for a given task, in the direction which generally
enhances visibility (e.g., increases in illumination, contrast,
and size, decreases in target speed, visual complexity, etc.)
but in ways which maintain maximum flexibility whenever
possible. It should be remembered that while the relationship
between a variable and stimulus visibility may be fairly
straightforward, relationships between variables and visual
performance are more task dependent. For many tasks,
"more" is not always "better". For most individuals, stimulus
effects will follow a curvilinear function relative to most vis-
ual tasks, with moderate levels of the variable having a more
500
facilitative effect on performance than levels at either

extreme. Because of such task dependence and because of
the variability in individual needs across and within low vision
persons, flexibility in environmental modifications or
enhancements is especially desirable.
5. Acknowledgements
This study was supported by the Rehabilitation Research and

Training Center Grant G008103981 from the National Insti-
tute for Handicapped Research, Department of Education,
Washington, D.C.
6. References
1. A.L. Corn: Visual function: A theoretical model for

individuals with low vision. J. Visual Impairment &
Blind. 77, 373 (1983)
2. S.N. Greenberg, L.E. Krueger: Effect of letter orienta-

tion and sequential redundancy on the speed of letter
search. Memory and Cognition. 11, 181 (1983)
3. H.C. Weston: Light, Sight and Work, 2nd ed. (Lewis,

London 1962)
4. C.C. Krischer, R. Meissen: Reading speed under real

and simulated visual impairment. J. Visual Impairment
& Blind. 77, 386 (1983)
5. L.H. Lehon: Development of lighting standards for the

visually impaired. J. Visual Impairment & Blind. 74,
249 (1980)
6. S.W. Smith, M. Rea: Proofreading under different lev-

els of illumination. J. Illum. Eng. Soc. 8, 47 (1978)
7. S.W. Smith, M. Rea: Performance of a reading test

under different levels of illumination. J. Illum. Eng.
Soc. 12, 29 (1982)
501
8. I. Lie: Relation of visual acuity to illumination, con-

trast, and distance in the partially sighted. Am. J.
9. S.Miles: Productivity and Comfort of the Visually

Impaired Worker as a Function of Low Vision Aid
Usage and Illumination/Color Contrast Modifications.
(Mississippi State University, Rehabilitation Research
and Training Center on Blindness and Low Vision, 1984)
10. G.E. Legge, G.S. Rubin, D.G. Pelli, M. M. Schleske:

Psychophysics of reading IT. Low vision. Vision Res.
25, 253 (1985)
11. P.R. Boyce: Age, illuminance, visual performance and

preference. Light. Res. Tech. 5, 125 (1973)
National Long Term Care Facility Survey
Sheree J. Aston, Monica Beliveau, Anne Yeadon
1. Background/Statistics
According to the 1980 Census, people over the age of 65 (eld-

erly), constitute approximately 11% of the total population of
the United States - over 25.5 million individuals. In the year
2000, the elderly population will total 32 million. Approxi-
mately 80 per cent of those over 75 (currently 10 million -
40% of the elderly population) are likely to experience some
degree of functional visual disability [1]. Prevalence esti-
mates of blindness and severe visual impairment all indicate
one clear characteristic: these conditions occur primarily in
older people. Many of these older, visually impaired people
are currently residing in long-term care facilities.
To ascertain the prevalence of visually impaired individu-

als living in institutionalized settings, two studies are gener-
ally cited: The National Nursing Home Survey [2] and the
Health Interview Survey [2]. The 1977 Nursing Home Survey
estimated that about 5% of all older people can expect to
spend some time during their life in such a setting [3]. The
study reported that about 3% of the residents were "unable to
see" and another 26% were partially or severely visually
impaired, giving a total of approximately 344,000 nursing
home residents who were visually impaired. The Health
Interview Survey indicated the existence of approximately 1.4
million severely visually impaired persons in the non-
institutionalized population, of whom 71% were over the age
of 65 [4]. By combining the figures of these two studies, it is
approximated that there are a minimum of 1,344,000 serious-
ly visually impaired elderly persons, 25% of whom are in nurs-
ing homes.
In 1980, the National Center for Health Statistics (NCHS)

conducted the National Master Facility Inventory (NMFI) and
identified 233,065 nursing homes in the United States. During
this same time, the survey estimated 952,600 full-time equiv-
alent employees worked in these nursing homes. Despite the
factors that elderly persons constitute approximately 70% of
the blind and visually impaired population, and that almost a
503
million employees in nursing homes are serving over a quarter

of a million of this population, this group is virtually ignored
in terms of current rehabilitation services [5-7].
2. Review of the Literature
Numerous studies have investigated the ways in which the

physical environment can impact an older individual's daily
living attitudes and habits [1,8-11]. Environments which are
visually dull, confusing or dangerous inevitably exacerbate
the sense of confinement and dependency, and the resultant
withdrawal tendency. Such behaviours may be viewed by both
resident and service providers as mental confusion or the ini-
tial signs of progressive senility.
RUSALEM's 1969 study [12] of blindness in nursing homes

established that a considerable number of residents were not
recognized as visually impaired. Those who were recognized
were treated as if they were completely dependent, when, in
fact, their primary physical limitation was their inability to
see. Other investigators have concluded that the majority of
blind and visually impaired residents are neglected by staff,
and that the staff's lack of awareness often leads them to
assume a need for total dependency on the part of such
patients [13].
Little research has been conducted to determine whether

nursing home personnel have an understanding of the feelings
and needs of visually impaired elderly residents, or if they
encourage, or have the training, to teach independent living
skills to such residents [13]. Most of the staff working direct-
ly with elderly individuals in long term care facilities have
not studied gerontology [14], nor have they been trained in
any aspect of rehabilitation of the blind and visually impaired
[13]. Vision rehabilitation personnel (e.g. rehabilitation
teachers and orientation and mobility specialists) are rarely
utilized by the nursing and medical-care system [15,16] and
their potentials as valuable components of the medical team
are invariably unrecognized and untapped.
No model training package or program is readily available

for institutional staff to help them in their work with elderly
visually impaired residents in nursing homes. In the absence
504
of needed curricula and support, the functional implications

of visual and other sensory disabilities tend to be ignored by
long term care service providers [17]. Further, long term
care residents often learn to accept the disability of vision
impairment as a normal part of the aging process [5] and have
little incentive to help themselves.
Because of the tremendous documented need, and the

dearth of available information specific to individuals work-
ing with the elderly visually impaired institutionalized popu-
lation, a nationwide survey of nursing homes was conducted
to assess their current practices and needs for alleviating the
problems related to their visually impaired residents.
3. Sample and Method
A survey was designed to assess the current practices in long

term care facilities related to their elderly visually impaired
residents. The instrument consisted of detailed questionnaire
checklists covering such areas as: the number of elderly resi-
dents (65-74), (75-84) and (>85), the number of visually
impaired residents; specific administrative policies regarding
vision examinations, screenings, vision rehabilitation; how
vision problems are detected, provision for visually impaired
residents, how staff are informed of a visually impaired resi-
dent's status and needs, any adaptations made for the resi-
dents, inservice education practices, specific areas which
need to be addressed, which personnel should receive inser-
vice training, and problems faced in providing inservice edu-
cation. The survey was sent to 2000 randomly selected long
term care facilities chosen from the 1982 National Master
Facility Inventory (NMFI). The sample was representative of
the national distribution of institution by type, size and loca-
tion. The surveys were sent out with a cover letter explain-
ing the nature and source of the request for information. The
survey was directed to the administrator of the facility. The
initial mailing resulted in 302 responses (15.1%). A second
mailing resulted in a response rate of 25.7% (513 total). The
responses were entered, tabulated, and analyzed for content
and representativeness of the initial sample.
505
4. Results
The demographic characteristics of the nursing home popula-

tion in terms of age and vision impairment are exhibited in
Tables I and 2. The majority of residents (81.6%) were over
the age of 65 (see Table 1). The breakdown by age group was
as follows: (65-74) 18.46%, (75-84) 39.18%, and (>85) 42.5%.
An estimated 15.16% of residents in the nursing homes were
visually impaired or totally blind. Of those individuals, 13.7%
were reported to be totally blind and 84.3% visually impaired
(see Table 2). These estimated numbers are comparable to
nationally reported figures.
Table 1
65 - 74 75 - 84 85 +
# % # % # %
5,591 18.46 11 ,863 39.18 12,873 42.36
Summaries of the survey questions regarding vision poli-

cies are shown in Tables 3 and 4. Few nursing homes had
administrative policies or requirements for vision examina-
tions and screenings. The majority (67%) did not have a poli-
cy on vision examinations while only 4% required a current
vision examination for every new admission, 4% required a
current vision examination only for those admissions known to
be visually impaired, and 20% admit the individual and then
perform a vision examination within a one year period (usual-
ly as needed). When asked if the facilities screened their res-
idents for vision problems, 35% responded yes, 65% responded
no. Of those facilities which reported no, the most common
reasons were: no policy requirements (45%), residents are
unable/unwilling to cooperate (14%), inaccessibility of trained
eye care specialists (10%) and lack of trained staff (9%).
506
Table 2
Totall~ Blind (B)

# X %
724 2.08 2.39
Severel~ Visuall~ Im~aired
# X %
3868 11.11 12.77
BIS VI
# X %
4592 13 15.16
Tables 5, 6, 7 and 8 report information on the identifica-

tion and services for visually impaired and blind nursing home
residents. Considering the lack of required vision examina-
tions and screening, it is not surprising that vision problems
are most frequently detected through staff observation, the
resident and family and friends. Once the impairment is
identified, most facilities made available the following provi-
sions: vision examination, non-optical aids, rehabilitation ser-
vices, and prescription of low vision optical aids. Nursing
home staff were usually informed of a visually impaired resi-
dent's progress and needs through the nursing supervisors, the
resident him/herself, family members, case conferences and
staff meetings. Several adaptations were made for visually
impaired and blind residents of the nursing homes. They
included: special recreation activities or modifications,
tactual/color/large print identification signs, and special
eating/dining room arrangements.
507
Table 3
4. DOES YOUR FACILITY

(CHID{ ONLY ONE)
18 Responses Require a current vision examination for every

-4% new admi.ssi.on
22 Responses Require a current vision examination on~ for

------;r% those admi.ssions known to be visually impaired
104 Responses Admit the individual, then perform a vision examination

2Ofo
16 Responses Within one month
~
12 Responses Within three months

-~
01 Responses Within nine months

-1%
10 Responses Within one year
-~
141 Responses As needed

~
340 Responses Have no policy on vision examinations

----r:r%
The results of the survey questions concerning in-service

education needs and problems are summarized in Tables 9-12.
When asked if their staff would benefit from in-service edu-
cation materials related to the care of the elderly visually
impaired residents, an overwhelming 93% of the nursing
homes responded yes. Of those who replied yes, the greatest
staff needs were in the areas of how to: teach residents
adapted activities of daily living techniques; adapt recrea-
tional activities and increase participation; increase staff
knowledge of functional implications of specific vision disor-
ders, to orient the resident to familiar/unfamiliar areas;
teach independent, safe ambulation skills, and help residents
deal with their emotions towards vision loss. The consensus
of facilities wanted nurses, nursing aids, dining room staff,
508
Table 4
5. DO YOU PJiRIODICALLY SCRIIDl YOUR RE3lDlliTS FIlR VISION PROBLEl'IS
173 Responses Yes

)5%
~ Responses No
.!.~2!2.z..is this because:
207 Responses There are no policy requirements

~%
_ ; Responses Most residents do not need it
_~% Responses t-!any residents are unable/unwilling to cooperate
17 Responses Staff do not have the time

~%
_~ Responses Lack of trained staff
45 Responses Inaccessibility of trained eye care specialists

~
67 Responses Other (please specify)
-----ut
1% 01 in process of setting up annual screening
1% Of residents unable to cooperate or withstand exam
1% OT family education needed
1% (j2" no one will provide in hexne service
5'% 2; if problems, screen or refer to a specialist
1% (jf short tenn resident
1% N no specification
1% ~ families tend to this and medical needs
21> TT screened as r.eeded
1% 03 patients are not transportable by car
1% 04 residents are taken for exams annually
1% 02" no reimbursement by medicare/medicaid
1% 02 expense to resident
1% 02 taken care of before admittance
1% '04 resident responsibility
1% Of ~ public aid
1~ Of on advice of ph¥sician
1% i:[ will be routinely screened
in-service educators, and psychosocial personnel to receive

such in-service training. In-service education is most fre-
quently provided by a combination of the director of nursing,
external consultants, in-service educator, or through outside
conference/workshops. The most frequently cited problems
currently faced in providing in-service education were limited
staff time, lack of information materials and "how to" guide-
lines, staff scheduling patterns, and a limited or non-existent
budget.
509
Table 5
6. IF RESIDENTS HAVE VISUAL PROBLElIIS, HOW ARE THEY MOST FR.EX;!UEN'rLY

DErEXjTED.
(PLEASE CHECK NO MORE THAN THREE)
193 Responses Routine general medical examinations

-38%
52 Responses Periodic vision screenings
--1b%
113 Responses Examinations by eye care specialist
-2Cf,
334 Responses The resident him/herself

-~6%
239 Responses Family or friends

--;r7%
410 Responses Staff observations
-81%
o Responses Ombudsman program/patient advocate
-b%
07 Responses Resident council
--1%
02 Responses Other ( please specify
--1%
1%01 Routine nurse assessment
1%01 Refers to optometrist/ophthalmologist
510
Table 6
7. WHEN THERE ARE INDICATIONS OF VISION IMPAIRMmT, DOES YOUR
FACILITY MAKE PROVISION FOR ANY OF THE FOLLOWING.
(CIOOK ALL THAT APPLY)
377 Responses Low vision examination qy ophthalmologist/

~ optometrist
221 Responses Prescription of low vision optical aids (e.g.
-----;j:5% magnifiers~ telescopic lenses, high powered
spectacles)
345 Responses Non-optical aids (e.g. sun visors, large print
-----"0;, books, writing guides)
90 Responses Training in the functional use of optical aids
-'8%
207 Responses Environmental modifications
-----;j:2%
257 Responses Rehabilitation services (e.g., ADL, Orientation
-52% & Mobility)
265 Responses Referral to community resources (e.g. State or

----;-4% private agenc,r for the blind/visually impaired
29 Responses Other (please specifY)
-~
2% 12 with permission of family members

1% ~ on suggestions qy optometrist
1% 03 talking books
1% Of government publications
1% ~ support group
1% or lighting
1% 02 taken to optom/ophthal, they do not come to facility
1% or proviSions for eye exam by optometrist/ophthanologist
1% Of counselor for visually impaired
1% 01 will have optometrist in two weeks
1% ~ contact attending M.D.
511
Table 7
8. HOW IS STAFF INFORMED OF VISUALLY IMPAIRED RE3IDmTS' STATUS,
PROORmS AND NEEDS.
(CIlJ!X]( ALL THAT APPLY)
299 Responses Staff meetings

~
Case conferences
297 Responses Review of medical charts

-59%
Physician consults
374 Responses NurSing supervisors

-75%
243 Responses Social service staff
-48%
187 Responses Outside professional consultants/referrals
-:;7,(,
375 Responses Resident him/herself
~5%
337 Responses Family members/friends

--r;7'/o
16 Responses Ombudsman program/patient advocate
-':1%
26 Responses Resident council
-5%
11 Responses Other ( please specify
---""2'/0
Patient/nursing care plan meetings

Recreational therapist
Sign on bed
Assignment sheets
Family members
Activity board and nurses station
Admission review and assessment
Staff notes
512
Table 8
9. PLEASE CHEX;K IF ANY OF THE FOLLOWING ADAPTATIONS HAVE BEEN MADE
FOR YOUR BLIND AND 'lISUALLY IMPAIRED RESIDmTS
110 Responses Special room/floor assignment

~%
70 Responses Special room/hallwaa- lighting

-16%
230 Responses Special eating/dining room arrangements
-5'3%
265 Responses Tactual/color/large print signs for
~1% identification of rooms, hallwaa-s, other
landmarks
275 Responses Special recreation activities or modifications
-----0'3%
162 Responses Specific fire/disaster plan
~7%
19 Responses Other ( please specify )

--4%
1% 01 As specified by ANSI code

1% 02 Raised letters on doors
1% 02 large print and talking books
1% or Volunteers
1% Of Not valid answer
1% Of Special seating for activities
1% Of large print reading materials
1% 02 Raised letters on telephone
1% ~ Special adaptive eating equipment
1% Of Special tape recorder with tapes for listening
1% Of Orientation to facility and room
1% Of Not equipped to handle blind people
1% 0) Sandpaper on door knobs
513
Table 9
10 • DO YOU F.:EL 'l'IIA'r YOUR STAFF WOULD BENEFIT FROM IN-iilllVICE
EDUCA'rION MATERIAlS RlILATED TO TEE CARE OF 'I'IIE ELDERLY VISUALLY
IJIIPAlRED RESIDENTS
457 Responses Yes

-9"3%
35 Responses No
-7%
If yes, what staff needs would you like addressed.

(Check All That Apply)
1'10 How to conduct basic vision 8creeni~

-:55%
115 How to interpret resal ts of eye examinations and eye
~4% doctors' notations in medical charts
.305 HoW' to increase staff knowledge of functional implicatiolld
-"63% of specific vision disorders
293 How to orient the resident to familiar/unfamiliar areas
-1)1%
25'7 How to ~ide the visually impaired resident
-----,3% (familiar/unfami lar environments)
286 How to teach independent safe ambulation skills wi thin the

--59% resident's room and section (including use of w/e, cane,
walker, etc.)
325 How to teach residents adapted ADL techniques (eating,
~7% personal care, bathing, grooming, organi9.ation of personal
possessions )
216 How to increase knowledge of special aids, appliances, and

~% technology
314 How to adapt recreational activities and increase

--';5% partiCipation
215 How to meet the writing and reading needs of blind and
45% visually impai red residents
227 How to make environmental adaptations

-47%
203 How to help the resident understand hiS/her eye disorder
-42'/>
Z78 How to help residents deal with thei r emotions toward vision
--')8% loss
189 How to develop/implement a team approach to provide

-)9% rehabilitation services to blind and visually impaired
residents
214 How to involve family/friends/volunteers in the

---"'44% rehabilitation process
234 How to increase knowledge of special available financial and

~~ service resources (federal, state, private) for meeting the
needs of blind/visually impaired residents
(){ Other (please specify)

-2'/>
1% 01 Eating meal.
1% N How vision loss effect ADL
1% ll) All of the above
1% N Review the anatOll\Y of the eye
1% QI How to talk to a blind person
514
Table 10
11. WHAT PERSONNEL WOULD YOO WANT TO REX:EIVE SUCH TRAINING

(CHEX;K ALL THAT APPLY)
435 Responses Nurses

~
435 Responses Nursing aides
-g-5~
42 Responses Physicians
--gf,
152 Responses Rehabilitation specialists (e.g., a.T., P.T.)

~3%
209 Responses Psycho-social personnel

~~
168 Responses Administrators

-,,~
326 Responses Person(s) responsible for in-service education

--71% of staff
235 Responses Dining room staff

---'1~
75 Responses Support/secretarial staff
~6%
101 Responses Security and maintenance staff

---:<2',t
171 Responses Volunteers
-,,%
31 Responses Other (please specify)
-7%
Recreational activities staff

Personal aides
Social service designee
All staff
Chaplain
Rehabilitation aide
Famil,v
Housekeepers, laundry
Dietary
515
Table 11
12. WHAT PROBLl'Ml, IF ANY, DO YOO FACE IN PROVIDING IN-llmVICE
EDUCATION
(CJID]( ALL THAT APPLY)
291 Responses Limited staff time

---os,:
194 Responses Staff scheduling patterns
~~
19 Responses Under-staffing
-4~
Staff turnover
~Responses Lack of staff expertise/knowledge
4,: Responses Lack of informational materials and "how-to"

guidelines
~~Responses Limited/no budget
14 Responses other (please specify)

-':If,
Cost
No instlllctor
TopiCS of higher priority
lack of resources in area.
Extra meetings create overtime
Knowledgable instrtwtora
Inseni ties needs to be given on two shifts
Staff not intergrated
Too IDal\Y other reqllired in services
Big state
I.eck of expertise
516
Table 12
13. HOW IS IN-SERVICE EDUCATION PROVIDED AT YOUR FACILITY

(cmx;K ALL THAT APPLY)
260 Responses Special in-service Educator

--'-2%
348 Responses Director of Nursing
~CJf,
55 Responses Director of Personnel

·-11%
133 Responses Director of Social Services

-~7%
108 Responses Medical Director

-22%
330 Responses External ConsQltants

----06%
222 Responses Offering time and/or fQnds for outside
-45% conferences/workshops
~ Responses Other (please specify)
5. Conclusions
The total number of blind and severely visually impaired indi-

viduals (b/svi) elderly in nursing homes in the United States is
conservatively estimated to be 325,000. (This figure was
determined by multiplying the average number of b/vi per
facility (13) times the number of nursing homes reported by
the 1982 NMFI (25,000)). Because most residents are identi-
fied as having a vision impairment through self-reporting,
staff observation, family and friends, rather than through
required, periodic or standard examinations and/or screen-
ings; the population reported is probably greatly underesti-
mated. Many visually impaired and blind residents remain
undetected and therefore unserved due to the lack of require-
ments for eye examinations upon nursing home admission and/
or periodic vision screening of residents. Although the major-
ity of nursing homes made provisions for visually impaired
residents, these provisions {low vision evaluations, optical and
517
non-optical aids and rehabilitation services} are only made to

residents when they are determined to be visually impaired.
There is a great need for nursing home staff to receive in-
service education on vision impairment. Properly trained
staff will be able to identify and care for their visually
impaired residents in an efficient, routine and timely manner.
The results of this survey, the 10 interdisciplinary team

meetings with nursing homes, the interviews with visually
impaired nursing home residents, and the curriculum develop-
ment workshop, will provide the basis for the needed work-
shop materials.
The comprehensive, interdisciplinary-based in-service

materials developed for nursing home personnel, will include
informational materials and "how to" guidelines on: vision
screening, functional implications of eye disease, recreation
activities, environmental activities, alternate methods of
performing activities of daily living, safe travel, orientation
and mobility, and psychosocial implications of vision loss.
6. References
1. J.L. Fozard, S.J. Popkin: Optimizing adult develop-

ment: Ends and means of an applied psychology of
aging. Am. Psychol. 33, 975 (1978)
2. The National Nursing Home Survey, 1977 Summary for

the United States, Vital & Health Statistics Series 13
No. 43 (National Center for Health Statistics, Hyatts-
ville 1979)
3. R. Kastenbaum, S.E. Candy: The 4% fallacy. Int. J.

Aging Hum. Dev. 4, 15 (1973)
4. Vision Problems in the US (National Society to Prevent

Blindness, New York 1980)
5. R.A. Scott: The Making of Blind Men (Russell Sage

Foundation, New York 1969)
6. L.G. Hiatt: Designing for vision impaired older people.

In Housing for the Elderly ed. by V. Regnier, J. Pynos,
Garland, 1982.
518
7. F .A. Koestler: The Unseen Minority (David McKay,

New York 1976)
8. W.lttleson: Some issues facing a theory of environ-

ments and behavior. In Theory Development in Envi-
ronment and Aging, ed. by P. Windley, T. Byerts, F.
Ernst (Gerontological Society, Washington D.C. 1975)
9. M.P. Lawton: The impact of environment on aging and

behavior. In Handbook of the Psychology of Aging, ed.
by J .E. Birren, K. W. Schaie (Van Nostrand Reinhold,
New York 1977)
10. K.F. Riegel: The dialectics of human development.

Am. Psychol. 31, 689 (1976)
11. J. Parr: The interaction of person and living environ-

ments. In Aging in the 1980's: Psychological Issues, ed.
by L.W. Poon (American Psychological Association,
Washington D.C. 1980)
12. H. Rusalem: A study of the incidence of blindness in

homes for the aged and nursing homes. N. Outlook
Blind 63, 168 (1969)
13. R.J. Wine burg: The Value of [nservice Training for

Staff Who Serve Elderly People in a Nursing Home
(Ph.D. University of Pittsburg 1980)
14. K.T. Tangorra: Your attitudes towards the elderly.

Nurs. Life September/October (1982)
15. I.M. Burnside: A nurse's perspective: Blindness in

long-term care facilities. N. Outlook Blind 68, 145
(1974)
16. C.D. Dickson: Issues in caring for blind patients. J.

Practice Nurs. November (1975)
Rights of Low Vision Children and Their Parents
Jennifer Leigh Hill
1. Introduction
This paper will deal with the rights - medical, educational and
rehabilitative - of a very special population of children. Chil-
dren that are "at risk" because of their recognized visual
impairment, as well as being "at risk" because some of their
basic needs are not being met.
Statistics for 1984 indicate that there are over 3,000

legally blind children under age 19 in Canada [1]. PRATT [2.]
suggested that there are four times as many partially sighted
children as there are legally blind. If this is in fact true,
there may be as many as 12.,000 low vision children in Cana-
da.
Are these children being served in a manner that recog-

nizes the fact that they are first children with vision or are
they being placed in programs for the visually impaired and
being taught to be "blind"? Are low vision children part of
the "forgotten population"?
GENENSKY [3], himself a partially sighted individual,

stated emphatically:
There is nothing wrong with being totally blind,

and there is nothing wrong in providing the
totally blind with services appropriate for their
blindness. It is, however, grossly unfair to
regard the partially sighted • • • as being blind
and to offer them services that are appropriate
only for the totally or functionally blind. What
the partially sighted need most of all is to be
classified differently from the functionally
blind. Until this difference is recognized by our
government and society, the partially sighted
will continue to be neglected.
520
2. Medical Rights
The most basic right of the low vision child is that of accu-
rate diagnosis of the problem. The diagnostic procedures may
range from basic vision screening to more complex diagnostic
work-ups. This basic right is most evident if the child is
thought of as being "visually at risk" [4]. For all children,
neonatal and infant screening programs are of prime impor-
tance in detecting a vision loss. There should be mandatory
vision screening prior to school entrance.
In 1985, PENNOCK and SHAPIRO examined the vision

screening programs in public schools in British Columbia.
This study was a follow-up investigation of a similar study
conducted in 1980 [5]. The authors found that the majority of
the schools still relied on far-point assessment by means of an
eye-chart. They concluded:
{I)t is as probable today as it was five years ago

that many children with visual anomalies (other
than myopia) are not identified in school
screening programmes. Thus many children
attempt to learn with some discomfort and/or
difficulty in classrooms of this province.
The findings of PENNOCK and SHAPIRO [6] are probably

representative of Canada as a whole. One exception is in the
Province of New Brunswick, where all Grade One children are
provided a "modified clinical examination" by optometrists
traveling to schools in a mobile vision van. The program
found, during the 1984-1985 school year, that 7.8% of the
children had visual problems requiring immediate care, and
that 14.0% of the children would need to be seen within six to
nine months [7].
Vision assessment for the multihandicapped child is often

a neglected area. CRESS et al. [8] discussed the need for pol-
icies regarding vision screening for this population. They
stated that "most states do not include persons with severe
handicaps in vision screening programs, even when such
screening is mandatory for the nonhandicapped". Regardless
of the individual's handicap, every person has a right to com-
prehensive professional eye examinations [8].
521
The multihandicapped are difficult to assess, and many

ophthalmologists and optometrists have been frustrated by
the fact that these children do not respond to the standard
testing procedures [9]. However, within the past several
years, a plethora of ar~icles and books have delineated meth-
ods that are applicable for assessing the vision of these spe-
cial needs children [9-18].
No child should be considered "untestable", regardless of

age or the multiplicity of the problem. Many children are
diagnosed as being "cortically blind", even though they have
not undergone an electrophysiological measure, such as the
visually evoked response (VER) [10]. Many children have been
placed in educational settings "accompanied by inadequate
reports giving some indication of visual classification and an
unintelligible description of the specific impairment" [17]. It
is the right of the parent (and teacher) to know the extent of
the vision loss. With this information, realistic goals for the
child can be set.
Once a diagnosis of the problem has been made, the obvi-

ous need for the child is treatment. If the vision problem can
be ameliorated (e.g. corrective lenses, low vision aids, sur-
gery, etc.) then the child should be treated by the appropriate
individual. If the child cannot be assisted medically, the child
should be referred to the appropriate agency.
Too often the attitude of the eye specialist is "nothing

can be done". The doctor makes the diagnosis - "your child is
blind" - and the child and the parents have to go home and
"live with it". I cannot stress enough the importance of fur-
ther follow-up by those that have training in the fields of
rehabilitation and/or education. Referrals should be made as
early as possible. Too often the referral comes at the point
the child is ready to enter school, at the age of 5 or 6. This is
too late. Early intervention is critical, as it plays a signifi-
cant role in ameliorating the effects of the loss and prevent-
ing the occurrence of additional problems (e.g. motor prob-
lems, cognitive deficits).
Children being educated in programs designed specifically

for the visually impaired are not always provided with the
medical intervention they require. HOFSTETTER [19] exam-
ined the visual needs of 60 visually handicapped students in
the Indianapolis public school system and at the Indiana
School for the Blind - children diagnosed as having a signifi-
522
cant visual loss. The students werE:: "known to have at least

some light perception and no other grossly disabling handicaps
serious enough to prevent the attaining of reliable responses
to visual tests". Hofstetter found that 45% of the students
could have been helped significantly by special visual aids and
rehabilitative guidance. Another 20-40% would probably have
gained from such assistance. Only 15% were judged to not
have needed assistance by means of optical aids.
Even more alarming were his findings related to routine

vision care. Over 40% of the children who were wearing
glasses did not have the correct prescription. The refractive
error magnitudes exceeded the criteria for referral in con-
ventional school screenings. He stated, "the specially classi-
fied group appears to be receiving no better, or even less,
vision care than the rank and file school population".
In his concluding remarks, Hofstetter stated "in a very

real sense we may be teaching them (i.e. those recognized as
being visually impaired) how to be blind rather than how to
utilize such residuals of vision as may be at their disposal".
Visually impaired children have the right to receive the "spe-
cialized visual attention that is available and capable of
enhancing not only their quality of life but also their opportu-
nities to be self-sufficient and self-supporting" [19].
In my work with visually impaired children, I have found

that treatment - basic treatment - is often not offered to all
children in Canada, particularly if they have more than one
problem. Many multihandicapped children are denied treat-
ment, (e.g. surgery, corrective lenses, etc.) because of the
severity of their handicaps. Ophthalmologists and optome-
trists must use the same treatment criteria for patients with
and without handicaps [8].
Many low vision children, both in residential and day

schools, could benefit from referral to a low vision clinic.
However, some clinics will not accept children. If there is not
a clinic available locally, geographical distance must not pre-
clude referrals. Ophthalmologists must refer children to clin-
ics run by optometrists, and vice versa. The needs of the
child must come first. Too often the parents are not made
aware that visual aids may dramatically improve the visual
status of their child. If they are told "nothing can be done",
they do not know to question the verdict. Eye practitioners
must act as advocates for this unique population.
523
Before a child is seen at the Low Vision Clinic, input

from the child's teachers and the child's parents is crucial.
Particularly critical is input from the teacher of the visually
impaired working with the child, for he/she is knowledgeable
about the child's visual abilities and the child's visual needs.
In an "ideal" situation, the examining doctor, or his assis-

tant, would conduct an "on-site" visit to the home, the school
or the work place prior to the assessment [20]. In a more
"realistic" situation, the information should be gained through
a Clinic Referral Form. Pre-examination information is crit-
ical to the successful prescription of a device and its long-
time usage.
For those children that are seen by a Low Vision Clinic,

adequate fOllow up must be available. The referring doctor
needs a follow up report, but even more important, the
parents and the child's teachers need to know the results of
the assessment, and if an aid has been prescribed, how the aid
should be used [21]. Children are just as likely to suffer from
the "top drawer" syndrome [22] if adequate follow up is not
provided.
Doctors must convey the assessment information, by

means of written reports, to the child's teachers and not
expect the information to arrive "through some osmotic pro-
cess" from the parent to the educator [23]. Since educational
decisions will be made on the basis of the inf ormation gained,
the evaluation results are "too important to trust to chance"
[23]. All children seen by a low vision clinic must be followed
carefully. As their visual needs change with maturity and the
visual demands change over time, a visit to a low vision clinic
should not be a "one-shot deal".
KELLEHER, in 1979, wrote the following:
It has been well documented ••• that there is a

tremendous lack of comprehensive low vision
services throughout the nation. Teachers, coun-
sellors, social workers, and other non-eyecare
professionals have begun dispensing low vision
aids to visually impaired persons out of frustra-
tion at the difficulty of obtaining appropriate
low vision services for their clients and stu-
dents. [24]
524
His comments, unfortunately, remain accurate in the mid

1980s. There is an urgent need for more low vision clinics
across Canada. There is an urgent need for pediatric low
vision specialists to serve these students.
Parents and children have the right to medical counseling

regarding the condition underlying the vision problem. Many
parents and children that I work with do not know the etiolo-
gy of the eye problem. I have had many parents say to me
that their child has had eye surgery - but cannot tell me what
the surgery was. Doctors must spend time talking to the
parents and to the children - in language that they can under-
stand. Often a simple diagram will help. Parents need to
know more than that their child is blind and glasses will not
help.
Parents need to know the cause of the eye problem and to

know if treatment can alleviate the problem. They need to
know what services are available for their son or daughter,
who the service providers are, and how to access the services.
They need to be made aware of services not only directly
related to the loss of vision, but also those, such as occupa-
tional and physical therapy, speech and language training, and
audiological assessment, that may be needed by their child.
Parents and children need to learn about environmental modi-
fications that may be beneficial and about equipment (both
optical and non-optical) that may be appropriate. They also
need to know about eligibility criteria for financial assistance
(e.g. the "blind" pension, income tax reductions).
To be able to provide this information to the parents,

ophthalmologists and optometrists must work to become more
informed. If eye practitioners are not able to do anything
more for the child, their involvement must not end there.
They are obligated to refer the children to those persons or
agencies that are able to provide further assistance. Eye pro-
fessionals too often have the attitude "we are interested in
vision but have little interest in blindness" - after all, blind-
ness represents failure and no one likes failure [25].
BRILLIANT [26] stated that the attitude and the words

"nothing more can be done" must be "struck from low vision
vocabulary". FREEMAN [27] suggested that there are two
ways that the eye professional can approach the low vision
child:
525
First is the image of the "omnipotent practi-

tioner". From high atop the professional moun-
tain, the doctor will lay down the edict as to
whether or not the child will see •••The other
attitude, the one which puts most children at
ease, is that the doctor is a helper whose knowl-
edge will aid the child's achievement in the
sighted world.
In all provinces there are provincial and/or local educa-

tional and rehabilitative services for visually impaired chil-
dren. The majority of these services extend from birth to age
21. The programs not only assist in the basic education of the
child {i.e. the three Rs} they also provide students with train-
ing in a wide variety of compensatory skills - vision stimula-
tion, orientation and mobility, daily living skills, vocational
training, to name a few. Ontario, Quebec and the Atlantic
Provinces have residential programs for the visually impaired.
Some residential schools also have home-based programs for
preschool children, school-age and multihandicapped children.
In all provinces, the Canadian National Institute for the Blind
has regional offices. Many of these offices employ Children
and Youth Counsellors and for older children, many of the
offices have Employment Counsellors.
Another area of counseling that is often lacking is that of

genetic counseling. I have students on my caseload that have
never been told that their eye problem was genetic in origin.
PAGON [28] suggested that "each family with a genetic eye
disorder needs to be evaluated to determine the specific diag-
nosis and its mode of inheritance. All individuals at risk in a
given family should be examined and counseled". The low
vision child has the right to be educated regarding the cause
of the vision problem, and how the condition may be passed
on to his/her offspring.
Those working with visually impaired individuals, in

either the educational setting or a rehabilitative setting, may
be in a unique position to provide some assistance. CROSS
stated:
Personnel employed by agencies and associa-

tions for the blind often spend much time with
patients and have extensive contact with rela-
tives. Consequently, his emotional rapport may
exceed that of the doctor and his patient, and,
526
since counseling and guidance are already par~

of this relationship, workers among the blind
have considerable opportunity and responsibility
to ensure that all who can benefit from genetic
counseling are advised of its availability. [29]
3. Educational/Rehabilitative Rights
All low vision children have the right to an appropriate edu-

cation. In Canada, there is no federal equivalent to the
American "Education for all Handicapped Children Act" (Pub-
lic Law 94-142) that ensures educational services to children
regardless of the handicap. Some provinces (e.g. New Bruns-
wick), but not all, have legislation dealing specifically with
the education of the blind child. Ontario's Bill 82, the "Edu-
cation Amendment Act", provides every exceptional pupil in
the province, including those that are visually impaired, with
access to educational programs. Regardless of where the
child lives, the low vision child is entitled to have an educa-
tion based on sound educational philosophy. These programs
must be of the highest quality.
HILL [30] examined the quality of programming to visual-

ly impaired students attending public day school programs in
Canada. She found that on the whole, educational services in
Canada were of high quality. In some parts of Canada, edu-
cational services to the preschool and/or multihandicapped
visually impaired children were weak. There is, however,
always room for improvement. Teachers of the visually
impaired must continue to strive for excellence.
A wide continuum of services must be available for low

vision children, designed to meet their unique needs. The
various program alternatives, suggested by BRYANT [31]
range from self-contained classrooms/programs (i.e., residen-
tial schools; local day schools) to consultative programs (i.e.,
teacher-counselor/consultant support to regular classroom
teachers).
The services must be provided by well qualified teachers,

with specialized training in working with low vision students.
These teachers "need a knowledge base, skills, sensitivity and
appreciation of the unique learning styles and needs of visual-
ly handicapped students" [32].
527
The low vision student is different from the totally blind

student, and, consequently a unique curriculum is necessary.
Programs designed to meet the needs of the totally blind may
not be suitable for the partially sighted (e.g., orientation and
mobility programs). The program must meet the educational,
social, psychological and emotional needs of the low vision
students so that they will be prepared to "make their own
choices as to how and where they will live, work and recre-
ate" [32].
In Canada, teachers can obtain preservice training (diplo-

ma level) in working with the visually impaired at the Univer-
sity of British Columbia. The University of Western Ontario
offers a 3-summer Additional Qualifications program, for
teachers in Ontario that are employed to teach the visually
impaired. In a study of 104 teachers of the visually impaired
working in Canada completed in 1984, it was foUnd that only
46% had specific training in working with the visually
impaired at the diploma or degree level [30]. There is an
urgent need to expand the preservice and inservice training
options for teachers of the visually impaired in Canada.
BARRAGA [33] suggested that teacher training programs

must "adapt and change the content ••• to fit the roles per-
formed and the population served". Teachers of the visually
impaired are working with more and more students that have
significant amounts of residual vision. Consequently, teach-
ers must have specific training to work with low vision stu-
dents. She stated:
How much attention are we devoting in teacher

preparation courses to the 80 or 85 percent of
the school-age population that has usable
vision? Certainly not 80 to 85 percent of the
curriculum. Teachers complain that 60 to 75
percent of the curriculum relates only to blind
learners [33].
The increasing emphasis on the use of residual vision (i.e.

"vision training" or "vision stimulation"), with both "normal"
low vision children and multihandicapped visually impaired
students, must be addressed by teacher training programs.
I am aware of only one training program that deals exclu-

sively with services to low vision individuals - the Pennsylva-
nia College of Optometry's Master of Science Degree Pro-
528
gram in Vision Rehabilitation. This is a one year program and

is designed "to prepare professionals from a wide variety of
disciplines (e.g. Optometry, Special Education, Orientation
and Mobility, Rehabilitation Teaching, Rehabilitation Coun-
seling and related fields) to employ a team approach in work-
ing with low vision individuals" [34]. There is no such pro-
gram in Canada.
Low vision children have the right to be provided with all

the necessary equipment that they need to complete their
schooling and to train them for future employment, along
with the right to receive training in the proper use of such
equipment. The term "equipment" not only refers to optical
aids, but also to a wide variety of non-optical aids.
Many children need aids and appliances for both school

and home - both should be made available. The financial sta-
tus of the child's parents should not be a consideration. If the
parents are unable to pay for the necessary equipment, and it
is not supplied by the educational system, other sources of
financial support must be tapped (e.g., service clubs, founda-
tions, insurance companies, etc.). Too often, there are signif-
icant delays in obtaining the necessary equipment, due to lack
of funding.
The low vision child and the child's parents have a right
to educational, psychological and vocational counseling by
persons that are trained specifically to provide the necessary
assistance. The teacher of the visually impaired is one pro-
vider of counseling, however, often the teacher works in con-
junction with many other professionals (e.g., guidance counse-
lors). Teachers of the visually impaired should be cognizant
of their own limitations in the area of counseling, and refer
their students for further assistance, if appropriate (e.g., psy-
chiatric counseling, vocational counseling, etc.).
In the area of educational counseling, the teacher of the

visually impaired should be knowledgeable about the educa-
tional programs available to the student, in the child's home
community. She/he must help the student choose courses
that are appropriate for future career goals, must be able to
discuss with the student and the student's teachers any modi-
fications and/or adaptations that will be necessary, and must
be aware of resources that may assist the student (e.g., read-
er service, books on tape).
529
One of the greatest handicaps of being visually impaired

is society's lowered expectations of the individual. The
teacher of the visually impaired must provide counseling to
students, other teachers, parents, etc., that is realistic and
appropriate. Visually impaired students are expected to com-
pete alongside of their sighted peers - visually impaired stu-
dents should not be getting As because they are different -
they should be getting As (or Bs or Cs) because they deserve
them.
In the area of psychological counseling the teacher of the

visually impaired plays a significant role, in her work with
both the student and the student's parents. A great deal has
been written about the psychological impact of a visual
impairment [35-38]. The adjustment of a person to being par-
tially sighted is considered by some to be greater than the
adjustment to being totally blind. They are people "in limbo",
neither sighted nor blind, the "marginal man" [39] often des-
perately trying to appear "normal". TUTTLE stated, "unfor-
tunately, low vision provides the individual with the opportu-
nity to hide his disability, to play "as if" he were fully
sighted" [38].
The teacher of the low vision student can be a "sounding

board" for the child and the child's family (e.g., siblings). The
teacher can be instrumental in bringing together visually
impaired students so that they can discuss the adjustment
problems inherent with low vision. The low vision student,
particularly the adolescent student, needs to define and delin-
eate the personal issues intrinsic to being a young adult with
low vision. Along with his teacher, ROESSING [40] suggested
that the student should explore such issues as: how he differs
from peers because of his low vision, if at all; his feelings
about the visual impairment and/or physical handicap; if
being different equals being inferior or superior to others;
how sighted peers perceive him; the issue of driving a car; the
attitudes and opinions of his peers with low vision; the opin-
ions expressed by successful blind members of the community
about the trials and pleasures of being a low vision person;
dating, marriage, family, children; and finally, the work ethic
and visual loss.
The teacher of the visually impaired may also be a pro-

vider of vocational counseling to the low vision student. Gui-
dance counselors, employment counselors and rehabilitation
counsellors also playa vital role. Career education is an
530
"essential part of every student's education" [39]. SPUNGIN

stated:
I find that many visually handicapped young-

sters do not have an accurate picture of them-
selves in relation to their peers. Students tend
to think either that they can do absolutely any-
thing in a particular field, or that they can do
absolutely nothing. Many students seem to be
laboring with a serious lack of information
regarding their own specific strengths and limi-
tations. [39]
It is the role of the vision teacher or rehabilitation coun-

selor to help the student understand the effects of the vision
loss in terms of future vocational plans. Does the child's
vision loss affect whether or not the job can be done, with
modifications, without modifications, or not at all? The days
of the "stereotypic" jobs for the visually impaired (such as
chair caning, mop making and piano tuning) hopefully are
over. Many different opportunities exist in the 1980s. Tech-
nology for the visually impaired has opened a lot of doors
[41,42]. The low vision student has a right to be made voca-
tionally productive. The decision regarding a specific occu-
pation must be based on "the capacity of the individual to
function efficiently and effectively rather than upon any
measured acuity or labeled categorization" [43].
The employment prospects for the handicapped unfortu-

nately are bleak, particularly in this time of high unemploy-
ment. The partially sighted individual, given proper training,
which results in obtaining the necessary qualifications has the
right to employment. Along 'with the right to employment
are ancillary rights to adequate housing and means of trans-
portation.
KIRCHNER and PETERSON [44] examined employment in
the blind and visually impaired in the United States. They
found that there was widespread unemployment. Approxi-
mately 75% of the general adult population is employed,
whereas, less than one third of the blind and visually impaired
were working. In another study by KIRCHNER and
PETERSON [45] it was found that of those blind and visually
impaired persons that were employed, many were underem-
ployed {i.e. working fewer hours than desired or employed in
jobs that are not commensurate in status, income or chal-
lenge with their training skills and desires}.
531
There have been no comparable studies in Canada that I

am aware of. However, it is not unreasonable to expect to
find similar statistics. As advocates for the visually
impaired, it is our responsibility to promote the hiring of vis-
ually impaired individuals, if they are qualified for the job.
4. Summary
In summary, it is apparent that low vision children and their

parents have many rights, however, many of their basic rights
are not being met by those of us that are working with them.
They have medical rights, that must be met by the medical
profession (i.e., ophthalmologists, optometrists, low vision
technicians, etc.). They have educational rights, that must be
met by the education profession (i.e. educators, administra-
tors, counselors) and they have rehabilitative rights, that
must be met by rehabilitation agencies (i.e. rehabilitation
counselors, employment counselors).
There are many professionals who deal with low VISIon

children. Each of us must determine our unique role, and
strive to provide the best service available within our man-
date. We must be the advocates for those with low vision.
They are a unique group. They have unique needs. Our goal
should be to make the low vision individual function at his/her
highest possible level. However, none of us can "do it alone".
We must work together in an interdisciplinary fashion. We
must not let low vision children become part of the "forgot-
ten population".
5. References
1. Statistical Studies on the Blind Population of Canada

Registered with CNIB: 1984. (Canadian National Insti-
tute for the Blind, Toronto 1984) pp. 15-16.
2. A.W. Pratt: The functions of a low vision clinic and

the use of aids. In Insight in Sight: Proceedings of the
Fifth Canadian Interdisciplinary Conference on the
Visually Impaired Child, ed. by A.M. Sykanda, J .E. Jan,
532
S. J. Blockberger, B.K. Buchanan, M. Groenveld (CNIB,

Vancouver 1984) pp. 114-122.
3. S.M. Genensky: Acuity measurements - Do they indi-

cate how well a partially sighted person functions or
could function? Am. J. Optom. Physiol. Opt. 53, 809
(1976)
4. M.E. Woodruff: The visually "at risk" child. J. Am.

Optom. Assoc. 44, 130 (1973)
5. C. Pennock, J. Shapiro: A survey of visual screening

programmes in British Columbia schools. B.C. J. Spec.
Educ. 4, 377 (1980)
6. C. Pennock, J. Shapiro: Visual screening programmes

in British Columbia schools revisited. B.C. J. Spec.
Educ. 9, 285 (1985)
7. Save your vision week on now. The Daily Gleaner

(March 6 1985) p. 36
8. P.J. Cress, C.R. Spellman, T.J. DeBriere, A.C. Size-

more, J.K. Northam, J.L. Johnson: Vision screening
for persons with severe handicaps. J ASH 6(3), 41
(1981)
9. S. Appel, M. Steciw, M. Graboyes, K.S. Cote: Manag-

ing the child with special needs. J. Vision Rehab. 3, 2
(1985)
10. J. Allen, K. Fraser: Evaluation of visual capacity in

visually impaired and multihandicapped children.
Rehab. Optom. 1, 5 (1983)
11. K.B. Costello, P. Pinkney, W. Scheffers: Visual Func-

tioning Assessment Tool. (Stoelting Co., Chicago
1982)
12. S. Dougherty, D. Levy: Functional vision assessment

for the severely impaired visually handicapped. Paper
presented at the CEC 58th International Convention
(1980)
13. C.A. Dowell: The assessment of functional vision in

the severely multihandicapped child. Paper presented
at the CEC 58th International Convention (1980)
533
14. M. Efron, B. DuBoff: A Vision Guide for Teachers of

Deaf-Blind Children (Special Education Instructional
Materials Center, Winston-Salem 1975)
15. C. Gates, C.S. Berry: Revised Manual for Visual

Assessment Kit (Unpublished manuscript, University of
Northern Colorado 1979)
16. M.B. Langley: Functional Vision Inventory for the

Multiply and Severely Handicapped (Stoelting Co., Chi-
cago 1980)
17. B. Langley, R.F. Dubose: Functional vision screening

for severely handicapped children. N. Outlook Blind.
70, 346 (l976)
18. A.J. Smith, K.S. Cote: Look at Me: A Resource Manu-

al for the Development of Residual Vision in Multiply
Impaired Children. (Pennsylvania College of Optome-
try Press, Philadelphia 1982.)
19. H.W. Hofstetter: Unmet vision care needs. J. Vision

Rehab. 3, 16 (l985)
2.0. J. Ferraro: Training: A team approach. Rehab. Optom.

1, 20 (l983)
2.1. L. Davis: Community resources - why should we use

them? J. Am. Optom. Assoc. 47, 1445 (1976)
2.2.. G. Watson, R.T. Jose: A training sequence for low

vision patients. J. Am. Optom. Assoc. 47, 1407 (l976)
2.3. G.R. Friedman: The teacher/doctor coordinating form

- an instrument in rehabilitation of the partially sight-
ed child. J. Am. Optom. Assoc. 47,1418 (l976)
24. D.K. Kelleher: Orientation to low vision aids. J. Visu-

alImpairment & Blind. 73, 161 (1979)
25. D. Stet ten: Coping with blindness. N. Engl. J. Med.

305,458 (l981)
26. R. Brilliant: Magnification in low vision aids made

simple. J. Visual Impairment & Blind. 77,169 (l983)
534
27. P.B. Freeman: Managing the low vision patient. J.

Am. Optom. Assoc. 50, 1267 (1979)
28. R.A. Pagon: The role of genetic counseling in the pre-

vention of blindness. Sightsaving Rev. 49, 157 (1980)
29. H.E. Cross: Genetic counseling and blinding disorders.

In Blindness 1974-1975 (American Foundation of Work-
ers for the Blind, Washington D.C. 1975) pp. 29-41.
30. J .L. Hill: A study of factors influencing quality of

programming for visually impaired children attending
public day schools in Canada. Diss. Abstr. Int. A Hum.
45, 1365A (1984)
31. N.W. Bryant: The continuum of services for visually

impaired students. In Quality Services for Blind and
Visually Handicapped Learners: Statements of Posi-
tion, ed. by G.T. Scholl (ERIC Clearinghouse on Handi-
capped and Gifted Children, Reston 1984)
32. K.M. Huebner: Services for the blind and visually

impaired. In Quality Services for Blind and Visually
Handicapped Learners: Statements of Position, ed. by
G.T. Scholl (ERIC Clearinghouse on Handicapped and
Gifted Children, Reston 1984) pp. 8-9.
33. N.C. Barraga: Innovations in teacher education. J.

Visual Impairment & Blind. 75, 96 (1981)
34. Fact Sheet: Master of Science Program in Vision Reha-

bilitation (1983) (Available from A.J. Smith, Program
Director, Master of Science Program in Vision Reha-
bilitation, Pennsylvania College of Optometry, 1200
West Godfrey Avenue, Philadelphia PA 19141)
35. R.M. Lambert, M. West, K. Carlin: Psychology of

adjustment to visual deficiency: A conceptual model.
J. Visual Impairment & Blind. 75, 193 (1981)
36. H.M. Mehr, E.B. Mehr, C. Ault: Psychological aspects

of low vision rehabilitation. Am. J. Optom. Arch. Am.
Acad. Optom. 47, 605 (1970)
37. O. Overbury, D. Greig, M. West: The psychodynamics

of low vision: A preliminary study. J. Visual Impair-
ment & Blind. 76, 101 (1982)
535
38. D.W. Tuttle: Self-Esteem and Adjusting with Blind-

ness: The Process of Responding to Life's Demands
{Charles C. Thomas, Springfield 1984}
39. S.J. Spungin: Career development: The educational

context. In Yearbook of the Association for Education
and Rehabilitation of the Blind and Visually Impaired:
1983. {AERBVI, Alexandria 1983} pp. 18-29.
40. L.J. Roessing: Minimum Competencies for Visually

Impaired Students (Unpublished manuscript, Freemont,
U.S.D. 1980)
41. L.A. Scadden: Technology and the labor market,

implications for blind and visually impaired persons. In
Blindness 1982-1983 {American Association of Workers
of the Blind, Washington D.C. 1983} pp. 17-24
42. F. Schmidt: The effects of new technologies upon the

employment of blind and visually impaired people in
Canada. In Blindness 1982-1983 {American Association
of Workers of the Blind, Washington D.C. 1983} pp.
25-28
43. N.C. Barraga: Utilization of low vision in adults who

are severely visually handicapped. N. Outlook Blind.
70, 177 (l976)
44. C. Kirchner, R. Peterson: Employment: Selected char-

acteristics. J. Visual Impairment & Blind. 73, 239
(l979)
45. C. Kirchner, R. Peterson: Worktime, occupational sta-

tus, and annual earnings: An assessment of underem-
ployment. J. Visual Impairment & Blind. 74, 203 (1980)
Life Satisfaction of Low Vision Patients and
other Disability Groups: A Preliminary Study
Mary Santangelo, Olga Overbury, Rina Lang
1. Introduction
Measures of life satisfaction, happiness and morale have all

been extensively employed as indicators of an individual's
sense of well-being. These sUbjective measures tap the elu-
sive properties of what is referred to as quality of life, and as
such, are widely accepted as indicators of psychological
adjustment [1]. Among the factors associated with quality of
life, a strong positive correlation with health status has been
repeatedly reported, whether health was self-assessed [2,3] or
rated directly by physicians [4,5]. The degree of social inter-
action has also been found to correlate with well-being, in
that involvement in informal activities such as visiting with
friends and neighbors results in high levels of reported satis-
faction [2,6,7]. As well, lower socioeconomic status has been
reported to correspond with lower levels of life satisfaction
[2,8]. The demographic variables of age, sex, race, and
employment status show an inconsistent relationship to life
satisfaction [3].
As might be expected, poor health and physical disability

have been found to correlate negatively with life satisfaction
[2]. NEGRIN suggested that chronic illness or specific physi-
cal disability, when combined with other age related losses of
functioning, may create a situation of "multiple jeopardy" in
which the problems of adjustment are compounded further
[9]. The literature on adjustment to disability has identified
some of the common coping strategies adopted in response to
disability, such as denial, anger, compliance, and increased
helplessness [10,11]. This research has typically sought to
investigate the characteristics unique to each disabled popu-
lation, and has generally involved comparison of a disability
group with a normal population on varying criteria. However,
there is a lack of comparative data outlining the similarities
or differences among various disability groups in terms of
psychological adjustment and life satisfaction. This informa-
537
tion would prove useful in determining whether there exists a

common process of adjustment to different disabilities, or
whether adjustment and life satisfaction is specifically tied
to the type of disability.
This study constitutes a preliminary effort to obtain com-

parative data on three different disability groups in terms of
current adjustment levels, as reflected in measures of general
ilfe satisfaction. The diagnostic groups selected for this
study consisted of patients who were diagnosed as having
either low vision, renal failure or cardiovascular disorders.
These three groups are similar in that they represent "invisi-
ble disabilities" wherein the medical condition is not readily
discernable by casual observation [12]. However, they are
distinguished in terms of the specific nature of the disability
and the degree of consequent functional limitation imposed.
The specific objective of the study was to compare these

different disability groups on several measures in an attempt
to answer the following questions: How do low vision patients
compare to other disability groups in response to standard
measures of general life satisfaction? And since disabilities
are defined as limiting functional abilities, is there a rela-
tionship between reported levels of activity and life satisfac-
tion measures in these disability groups? How do these dis-
ability groups compare to normal peers on both life
satisfaction and activity measures? And finally, how does a
measure specifically designed to assess quality of life in disa-
bled individuals compare to standard life satisfaction meas-
ures in this sample of patients?
2. Methods
2.1 Subjects
A total of 27 men and women participated in the study.

Except for the seven people who comprised the control group,
all participants were patients at several downtown Montreal
hospitals.
The cardiac group consisted of seven people diagnosed as

having chronic stable angina pectoris or congestive heart fail-
ure. The dialysis group consisted of six hemodialysis patients
538
seen three times weekly in hospital. The low VISIon group

comprised seven individuals with the diagnosis of macular
degeneration. Best visual acuity ranged from 6/21 (20/70) to
6/120 (20/400) in these patients. The control group had no
major chronic illnesses.
2.2 Materials
The Life Satisfaction Index A is a widely used instrument

which requires that the respondent examine his or her present
life situation, compare it to the past, and project it to the
future [13]. Although standardized for use with a normal
healthy population, its demonstrated effectiveness with eld-
erly populations rendered it appropriate for the sample in this
study.
A recently developed instrument, the Assessment of Cur-

rent Community, Emotional and Social Satisfaction (ACCESS)
[14] questionnaire, was also employed. Designed specifically
for quality of life assessment in disabled populations, it incor-
porates both objective and subjective indices of life satisfac-
tion, and it samples directly the frequency of specific emo-
tions and activities.
An activities questionnaire, also used with the aged, was

employed to measure the range and frequency of activities
engaged in at both personal and social levels [15].
The Holmes Social Readjustment Scale was used to obtain

self reported levels of stress [16]. The respondent is required
to indicate which of a series of specific events he or she has
experienced during the past year.
2.3 Procedure
Informed consent was obtained from all study participants.

The assessment battery was administered in a semi-
structured interview following the testing of visual acuities.
539
2.4 Results
Subjects' scores were calculated for each dependent measure

and the data were analyzed separately using one-way analysis
of variance.
Analysis of Life Satisfaction Index scores yielded an

overall significant difference between groups, F{3,23) = 4.48,
p<0.05. Post-hoc analysis using the Scheffe test indicated
that the group means for dialysis patients (x = 8) and cardiac
patients (x = 8.5) were significantly lower than that of con-
trols (x = 14.7). However, the disability groups did not differ
significantly from one another. This indicates that using
standard life satisfaction measures, the disability groups were
not distinguishable from each other. Compared to controls,
however, both cardiac and dialysis patients reported less sat-
isfaction in their lives. Only low vision patients reported lev-
els of life satisfaction similar to controls.
For the purposes of this study, the ACCESS data were

analyzed in terms of the frequency of emotional responses
and particular activities in order to obtain one score for life
satisfaction.
The analysis on the ACCESS scores also yielded signifi-

cant findings overall F (3,23) = 3.96, p<0.05, but Scheffe post-
hoc analysis indicated that this was due to the lower scores in
the low vision group compared to the control group, F (3,23) =
10.33, p<O.Ol. Again, the disability groups did not differ sig-
nificantly from each other on this life satisfaction measure.
However, the significantly lower levels of life satisfaction in
low vision patients versus controls indicates a totally differ-
ent pattern of results from the Life Satisfaction Index meas-
ure. In fact, the results are reversed.
Analysis of the activity data revealed an overall signifi-

cant difference F (3,23) =4.66, p<0.05, which Scheffe testing
indicated was due to lower activity levels of the low vision
group (x = 50.2) compared to the control group (x = 67.8).
Low vision patients were thus functionally more impaired
than non-disabled peers, but not more impaired than either
the cardiac or dialysis patients.
The Holmes Social Readjustment data analysis did not

reveal significant findings, indicating that all four groups
reported similar levels of stress at the time of testing.
540
3. Discussion
This preliminary study comparing low vision patients to other

individuals suffering from disorders in renal and cardiovascu-
lar functioning failed to reveal differences in self-reported
levels of general life satisfaction, activity and stress. Signif-
icant differences did emerge when these disability groups
were compared to a non-disabled peer group.
Using standard measures of life satisfaction, low VlSIon

patients were the only ones to report levels of life satisfac-
tion comparable to that of healthy peers. The lower levels of
life satisfaction in cardiac and dialysis patients relative to
controls, may in part be due to their need for constant medi-
cal supervision, as well as the potentially life threatening
na ture of their illnesses.
These findings, however, were not reproduced when life

satisfaction was measured using an instrument specifically
designed for disability groups. On the ACCESS measure, low
vision patients were the only ones who were less satisfied
than healthy peers. Since this questionnaire directly assesses
the frequency of certain activities, of which almost one quar-
ter are particularly vision-dependent, it may be that this
instrument was especially sensitive to the functional limita-
tions experienced by the low vision patients in this sample.
The findings for level of activity are in accordance with this
interpretation, since low vision patients were again found to
be less active than non-disabled peers.
This preliminary study needs to be replicated with a larg-

er sample in order to validate these findings. Nonetheless, it
raises important issues which need to be considered in future
research. The relative importance of activity levels to the
low vision patient's sense of well-being and life satisfaction
needs to be more clearly delineated. As well the type of
activity which is more conducive to greater life satisfaction
in low vision patients as well as other disability groups needs
to be explored.
The different pattern of results obtained with the Life

Satisfaction Index and the ACCESS questionnaire indicate
that these instruments are not tapping the same aspects of
quality of life. It is likely that standard measure of life satis-
faction are simply not sensitive to the particular experiences
541
of disabled individuals. Further research is needed using

instruments such as the ACCESS questionnaire to provide
more information about the correlates of life satisfaction in
low vision and other disability groups.
4. References
1. J. Horely: Life satisfaction, happiness and morale:

Two problems with the use of subjective well-being
indicators. Gerontologist 24, 124 (1984)
2. J.N. Edwards, D.L. Klemmack: Correlates of life satis-

faction: A re-examination. J. Gerontal. 28, 497 (1973)
3. R. Larson; Thirty years of research on the sUbjective

well-being of older Americans. J. Gerontol. 33, 109
(1978)
4. F.C. Jeffers, C.R. Nichols: The relationship of activi-

ties and attitudes to physical well-being in older peo-
ple. J. Gerontal. 16, 67 (1961)
5. E. Palmore, C. Luikart: Health and social factors

related to life satisfaction. J. Health Soc. Behav. 13,
68 (1972)
6. B.W. Lemon, V.L. Bengston, J.A. Peterson: An explo-

ration of the activity theory of aging: Activity types
and life satisfaction among in-movers to a retirement
community. J. Gerontol. 27, 511 (1972)
7. K.J. Smith, A. Lipman: Constraint and life satisfac-

tion. J. Gerontal. 27, 77 (1972)
8. S.J. Cutler: Volunteer association participation and

life satisfaction: A cautionary research note. J. Ger-
ontal. 28, 96 (1973)
9. S. Negrin: Psychosocial aspects of aging and visual

impairment. In Understanding Low Vision, ed. by R.T.
Jose (American Foundation for the Blind, New York
1983) pp. 55-59.
542
10. F. Cohen, R.S. Lazarus: Coping with stresses of ill-

ness. In Health Psychology - A Handbook, ed. by G.C.
Stone, F. Cohen, N.E. Adler (Jossey-Bass, San Francis-
co 1979) pp. 217-254.
11. F.C. Shontz: The Psychological Aspects of Physical

fllness and Disability. (Macmillan, New York 1975)
12. D.R. Falvo, H. Allen, D.R. Maki: Psychological

aspects of invisible disability. Rehab. Lit. 43, 2 (1972)
13. B.L. Neugarten, R.J. Havighurst, S.S. Tobin: The

measurement of life satisfaction. J. Gerontol. 16, 134
(1961)
14. S.M. Shindell, G.L. Goodrich, M.E. Dunn: The Devel-

opment of a Life Satisfaction Scale Applicable for
People with Severe Disabilities. Study supported by
National Institute for Handicapped Research
(1984-1987)
15. R.D. Savage, L.B. Gaber, P.G. Britton, N. Bolton, A.

Cooper: Personality and Adjustment in the Aged.
(Academic, New York 1977)
16. T.H. Holmes, R.H. Rahe: The social readjustment rat-

ing scale. J. Psychosom. Res. 11, 213 (1967)
Sight Enhancement Services -
A Safety Net or a Spider's Web?
E.J. Herie, G. Grace
1. Sight Enhancement Services
This paper will review the development of the "human servi-

ces" aspect of sight enhancement, and the integral role of
human services in an effective and comprehensive sight
enhancement system.
First, what exactly is meant by human services? Briefly,

it is the organized capacity of people to care for one another
through organizations or social institutions - how we as a
society offer love, compassion and help to members of our
society and the world at large. With regard to sight enhance-
ment, human services are those rehabilitation, counseling,
educational and leisure services that allow individuals the
freedom to attain their potential and to participate fully in
society. Important as these human services are, they cannot
be offered in isolation from the diagnostic and remedial assis-
tance provided by vision health practitioners and specialists.
The use of the term "blind" refers to persons without any

vision at all, while at the same time, there is a close associa-
tion between blind persons and those whose vision is extreme-
ly poor, or is failing rapidly.
Early reference to care and compassion for blind persons

is found in the book of Leviticus: "Thou shalt not ••• put a
stumblingblock before the blind." In the 14th century
attempts were made to allow blind persons to read, but five
centuries were to pass before Louis Braille's embossed system
was to remove the barrier of illiteracy.
From this point Western societies were to witness the

evolution of organized human services for the blind and visu-
ally impaired. Braille's system was truly the key to organized
human services for the blind as they exist today. This evolu-
tion began with the development of schools for the blind. The
544
first school for the blind in Canada was established in 1861,

but over half a century passed before the CNIB as a service
organization received its charter in March of 1918. CNIB
proved to be a unique approach to human and rehabilitative
services given that it is a national service organization. Con-
trasted to other developed countries, CNIB represents a most
interesting approach, and one that is envied by other organi-
zations involved in human services in this country and else-
where.
The Canadian National Institute for the Blind - the CNIB

- is a complex rehabilitation organization providing a wide
range of human services. While the CNIB has only recently
become involved on an expanded scale with the new Sight
Enhancement Program for visually impaired individuals, its
goal has always been to help blind, visually impaired and blind
multi-sensory-deprived individuals find ways to lead produc-
tive and satisfying lives. The concern has always been with
the whole person, and not just his or her lack of vision.
As the needs and expectations of the blind and visually

impaired popUlation changed, and as society has become more
interested in the rights of disabled individuals, this concern
has taken a number of forms. The CNIB began by providing
rehabilitation services and training and worked to educate the
public while lobbying the government at all levels. This
approach paid off through legislation, including disability pen-
sions, public health measures, Income Tax Act exemptions,
and White Cane Acts. The CNm not only employed blind and
visually impaired persons, but built residences and service
centers. In 1968 the CNIB established a national vocational
guidance and training centre, a decision that in retrospect
was at variance with today's trend to community-based servi-
ces and programs.
Today the CNIB has only a few residences for elderly

blind persons remaining, as public and private sectors in Can-
ada have assumed responsibility for providing services in local
communities.
From its inception the CNIB has promoted consumer par-

ticipation in decision-making. Today the public has become
sensitized to the problems of the disabled, and organizations
of the disabled have achieved a stronger influence in
decision-making, planning and implementation of services
directed at them, and in numerous initiatives to secure gener-
al and specific human rights.
545
This trend will continue, along with decentralization of

service, and partnerships between government, service organ-
izations, and consumer groups in this field. The consumer -
the individual consumer - is going to have an increasing say,
not only in what service is provided, but in how, where and
when it is provided. To keep in touch with and respond to the
needs of this fast-changing market, service providers are
going to have to step up their sensitivity to and understanding
of the environment with increased penetration and analysis of
the "consumer" impact. Planned co-ordination of human and
sight enhancement services will become even more important
if service providers hope to deliver responsive, efficient and
useful service to our constituency.
Some of these trends are worrisome given the risk of

fragmentation and duplication and CNIB has become one of
the players in an ever-expanding network of professional
organizations. But CNIB, because of its concern with human
services and with its firm goal of helping individuals with
impaired vision achieve their full potential, must ensure that
individuals are receiving the service they need, and that the
complex network helps rather than hinders their attempts to
take their place in society.
Before addressing the issue of the sight enhancement sys-

tem or network in Canada, a brief description of CNIB's
expanded sight enhancement policy and program is required.
In 1985 CNIB served 46,000 individuals of whom approxi-

mately 10 per cent (or 4,636) were new registrants. Figure 1
depicts the dramatic increase in the number of registrants
from 1969 to 1985 - more than 68 per cent. This registered
group is comprised of individuals who fall within the 6/60
(20/200) or less category, often referred to as "legally blind".
In fact, 90 per cent do have a degree of residual vision.
Globally, the number of blind persons is expected to

increase from 42 million in 1978 to 100 million by the year
2000. The prevalence of blindness in Canada and other devel-
oped countries is approximately 222 cases per hundred thou-
sand people, compared to as many as 5,000 per hundred thou-
sand in developing countries. The prevalence of blindness for
developed countries supports our projection of a total of
55,000 people eligible for registration with CNIB. However,
if the scope of what is considered severe visual impairment is
broadened from 6/60 (20/200) or 10 per cent vision or less to
546
45,200 45,631
43,800 /
V
42,400 ~/
41,000 /
39,600 12 I
38,200
J,/
/
J6,800
I
[
J5,400
J4,OOO
j
32,600 J
&.. /
V
)1,200
~.V
'19,800
'1.8,400
-t LV
~7 ,000 ~
1969 1971 1973 1975 1977 1979 1981 1983 1984 1985
Figure 1
6/21 (20/70) or less, the number of potential registrants for

CNIB services increases by 195,000. Altogether, there may
be 250,000 people in Canada who might benefit from Sight
Enhancement Services, although by no stretch of the imagina-
tion would all these people require assistance, nor would
CNIB have the capacity to serve them.
It is for this group, those whose visual impairment is not

severe enough to be considered blindness as the term is usual-
ly defined, but whose sight is poor enough to create practical
or functional problems, emotional stress and family difficul-
ties, for whom CNIB has recently developed a new sight
enhancement policy and launched a new program, Sight
Enhancement Enterprise (SEE).
547
Through SEE, every person wishing to optimize his or her

remaining sight can be trained in the best use of this residual
vision and will have access to the most advanced technologic-
al products. Our first priority is and will continue to be those
with no vision or extremely poor vision, the deaf-blind and
the multi-handicapped. It is this group which the CNIB will
serve first and always.
Anyone with a vision problem can approach any of the 50

offices of CNIB across Canada for counseling, rehabilitation
services, or referral to other community services. The pro-
gram will be a co-operative effort involving health practi-
tioners and education and social service personnel from the
private and public sectors.
In short, SEE is about rehabilitation. It is designed to

reduce the impact of disabling and handicapping conditions,
with the objectives of enabling the disabled and the handi-
capped to achieve social integration and full participation.
The importance of such rehabilitation is well documented:
visual impairment, along with other disabilities, causes socie-
ty to incur high costs, both social and economic. Rehabilita-
tion and support programs can significantly reduce these
costs, and better still, prevention can avoid them altogether.
The bottom line, of course, goes back to human services and
the organized capacity of people to care for one another and
help each other to lead satisfying lives.
Two aspects of the SEE program are of special note.

First, everyone of the 50 CNIB offices will have a range of
technical aids on display, and 14 major display centers across
the country are now being developed. Individuals with vision
problems will have the opportunity to try new technical aids,
to borrow them and to receive training.
The second aspect is service to seniors. Today, nearly 60

per cent of the people registered with the CNIB are over 65
years of age, compared to about 45 per cent only 10 years
ago. Since 1971, when CNIB registered 1,100 new persons
over 65, the number of new registrants over 65 each year has
climbed 300 per cent to 3,300. That 3,300 represents 72 per
cent of the new registrants in 1985. It underlines the fact
that the population of older people, here and in the rest of
the world, is growing at a tremendous rate, and is going to
have considerable impact on health systems, including sight
enhancement.
548
Loss of vision usually occurs slowly, over time, and may

not be noticed by the individual, his family or his friends.
However, loss of vision can have a devastating effect on an
individual. Low vision aids have on occasion had impressive
results with older people who had been mislabeled "confused"
or "depressed" when the real problem was that they could no
longer see well.
Sight enhancement services for the elderly are particu-

larly important because they can mean the difference
between independence and institutionalization. The price we
have paid for industrialization in the developing world is the
erosion of the nuclear family. This is in stark contrast to
stronger family ties in the developing world. In Western soci-
ety we are more dependent on the safety net. Loss of vision
often forces an individual to give up his or her home and ulti-
mately to lose independence. And since independence is one
of the best measures of quality of life, the role of sight
enhancement services in maintaining independence is crucial.
The Sight Enhancement Enterprise Program, and CNIB

itself, are today just one part of the complex network of low
vision services in Canada portrayed in Fig. 2. At the centre,
of course, is the consumer, the person with poor or failing
vision. Next is the resource system -- optometry, ophthal-
mology, medical doctors, public health nurses, hospitals, uni-
versity clinics, vocational rehabilitation services, assistive
devices, research groups, consumer groups, volunteers and
other services and organizations such as the CNIB. Next is
the information system, which attempts to connect consum-
ers with the services available. This includes the media,
association newsletters, government publications, libraries,
consumer organizations, and so on.
It is a good system, it is a thorough system, and in Cana-

da there are a minimum of 45,000 points of entry to obtain
rehabilitative, medical and the entire range of sight enhance-
ment services. For North America as a whole, the evidence
suggests approximately half a million points of entry. There
exists enough entry points to the system to ensure access for
everyone in Canada or North America who needs assistance.
But the question is, does it? Does the system ensure access
to appropriate service? Unfortunately, the answer is no, not
always. It doesn't take much discussion with the consumers,
with the users of the system, the people who are supposed to
benefit, to discover problems with this very complex system.
549
Llbr&rlea
• Covern.tnt Publications
Figure 2. Low Vision Services Network
In other words, the safety net could be described as function-

ing more like the spider's web where you are caught in the
system rather than helped by it.
Too commonly, one of the first problems cited is lack of

money. It is difficult to calculate just how much money is
allotted for services for the visually impaired in Canada, but
if we knew, the amount would be staggering. More money is
not always the solution; rather, better planning and co-
ordination of comprehensive services can achieve surprising
results.
Another problem, and perhaps a more basic problem, con-

cerns co-ordination. In many cases, people are receiving con-
flicting information from a variety of sources. It is not
unusual for parents of a child with visual impairment to
receive advice from 15 different sources and to be left con-
fused and without clear direction. In other cases, people are
not receiving enough information, or what they are receiving
is fragmented. Or they get fed up with bureaucracy, with
being shunted from one place to another. The result is often
that they are trapped at a particularly part of the network,
550
labeled as blind or not blind, and not really receiving the most
beneficial assistance -- and CNIB must accept its share of
criticism in this regard. For these reasons it is imperative
that all of us work together to co-ordinate our efforts.
To return to that all-important human services aspect, it

is not unusual for consumers to complain that specialists in
low vision are authoritarian and unimaginative in their coun-
seling. Consumers say that low vision experts do not tell
them the things they really want to know. And perhaps worst
of all, many low-visioned people believe the experts have
focused their work so narrowly on the eyes themselves, that
they overlook in the person in whom the eyes reside.
The answer to many of these problems is planned co-

ordination of funds and services. In the absence of co-
ordinated services, this incredible system functions not as the
safety net it was meant to be, but as a spider's web, restrain-
ing and confusing, and preventing the individual from receiv-
ing equitable and beneficial service. With the necessary co-
ordination, basically adequate resources in the developed
world, and the potential for the establishment of adequate
resources in the developing countries, we can provide the kind
of safety net truly required.
Each discipline has a clear picture of what needs to be

done and how it should be done. The problem arises when
these disciplines become so narrowly focused on their own
specific areas that they are unable to work together with oth-
ers to present a clear direction to the consumer. Without the
ability to see beyond their own specialties and conditions,
valuable resources are wasted and effectiveness and ability to
truly help people with low vision to achieve their full poten-
tial is reduced. While we have sight, we sometimes lack
vision, the vision that will give us what we seek - an effec-
tive, efficient and caring sight enhancement system accessi-
ble to all.
A final word of caution - organizations of and for the

blind and visually impaired risk sliding under the umbrella of
the generic disability group. There is, in this worrisome
trend, an inherent danger that the problems that have been
identified for this group will be compounded if we water
down, or worse, lose the special knowledge and expertise that
is required for successful intervention.
551
2. Summary and Conclusion
This review of current and historical trends has been present-

ed as a guide for future planning and activities. Enough is
known of the future in respect of demographic, economic and
technological trends to project service requirements well into
the 21st century. It is clear that we must make our existing
resources work more efficiently and more quickly to serve
more people. We must allocate both human and financial sup-
port to make the principle of equity a reality throughout the
world. In this regard, we cannot overlook the dramatic and
excellent work currently underway to prevent blindness or
severe visual impairment, since as much as 80 per cent of
blindness is said to be preventable. Where prevention of
blindness is not possible, remedial programs must be either
developed or strengthened.
Relevant research and statistical data will help us all in

that essential process of planning and co-ordination. As this
information is studied and analyzed to determine its rele-
vance to the respective disciplines and organizations, it
should be done with a sense of excitement and renewal. This
is truly a unique opportunity to challenge and be challenged.
Our response must reflect our belief and resolve that the
impossible can be achieved through co-operation, planning
and co-ordination. We can reduce or eliminate barriers to
service, whether they be social, economic or organizational.
Too long have we accepted systems established to serve
humans, systems that, over time, have become inflexible and
insensitive, and that in their worst applications create
orphans rather than beneficiaries.
Much has been written about the incredible sight of the

eagle. The vision of a clam is a frightening parody of that
keen eyesight. The information reviewed in this paper sug-
gests that, in respect of sight enhancement systems, we have
a sky full of eagles with the clearest sight in their respective
areas. The challenge before us is to harness those eagles and
to focus their vision towards the future where the world can
be relatively free of blindness and severe visual impairment.
Astronauts, from their vantage point in space, have described
with awe the incredible beauty of our earth. Our view of that
world must be shared through a common vision of what can be
achieved through our collective contribution as individuals
with a common purpose.
552
3. References
1. B. Carroll: Eye Trumpets: A Consumer Guide to Low

Vision and Low Vision Aids {Low Vision Association of
Ontario, Toronto 1983}
2. I.R. Dickman: Making Life More Livable (American

Foundation for the Blind, New York 1983)
3. World Health Organization: Disability Prevention and

Rehabilitation, Technical Report, Series 668 (WHO,
Geneva 1981)
4. The Elderly in Canada. (Statistics Canada, Ottawa

1984)
5. S.M. Genensky: Data concerning the partially sighted

and the functionally blind. J. Visual Impairment &
Blind. 72, 177 (l978)
6. R.T. Jose: Understanding Low Vision (American

Foundation for the Blind, New York 1983)
7. A.S.M. Lim, B.R. Jones: World's Major Blinding Condi-

tions. In Vision, Vol. 1 (International Agency for the
Prevention of Blindness, Singapore 1982)
8. E.P. Scott, J.E. Jan, R.D. Freeman: Can't Your Child

See? (Pro-ed, Austin 1985)
9. B.N. Wattie, W. Harding LeRiche: All About Eyes: The

Work of Ophthalmologists in Ontario (The Ontario
Medical Foundation, Toronto 1977)
10. B.N. Wattie, W. Harding LeRiche, M. Langer: Vision

Care: A Survey of Optometrists in Ontario. (Faculty
of Medicine, University of Toronto, 1980)
Strengthening Low Vision Rehabilitation
Through the Accreditation Process
Alfred A. Rosenbloom
1. Introduction
This international symposium speaks to the dynamism of low

vision rehabilitation. During the past two decades, the scope
and depth of low vision practice - its methods, expanding
technology, management approaches, interdisciplinary care
and responsibilities, expansion of low vision services in clini-
cal settings within academic institutions and agencies - have
impacted dramatically.
Professional, ethical standards must keep pace with

changes in practice and service. Let me cite several issues:
1. With increased involvement in policy, planning and

research:
o What are the obligations assumed for scientific research

and objectivity?
o What rights of patients, their families, program popula-

tions, and institutions are to be protected?
o What responsibility must be assumed to assure against

misuse of the professional role and for acknowledging
the efforts of other professionals?
2. With the expansion of private practice including the

intrusion of for-profit corporate enterprises into vision care:
o Should there be limits on advertising or marketing of

professional services?
o What safeguards should be adopted for new diagnostic

and treatment methodologies for quality assurance?
554
o How can eye care providers make timely, appropriate

referrals for specialized low vision services beyond the
scope of their skills and know ledge?
3. With increased organized patient or consumer partici-

pation in service programs:
o What is the relative responsibility to individual/group

when both may be recipients of service?
o Under what conditions should the optometrist or

ophthalmologist be responsible for separating from
unethical or unprofessional arrangements?
4. With changing technology in practice:
o What are our responsibilities for scientific objectivity?
o What choices in treatment or management are to be

protected for the recipients of services?
o Who should represent the patient in relation to rights

that are attenuated by procedural requisites to service?
5. With the demographic imperative creating ever great-

er need for access to low vision care for elderly persons:
o How can third party programs become increasingly

responsive to this need?
The accreditation process with its responsibilities for

standards setting and enforcement is one approach to some of
these issues. This is the intent of my paper which draws on
my varied experiences as a member of both specialized and
regional accrediting bodies and as an optometric consultant
for over 30 years to a pioneer low vision facility in America,
the Chicago Lighthouse for the Blind.
More specifically, I have served 5 terms (24 years) on

optometry's educational and clinical accreditation bodies -
the American Optometric Association's Council on Optome-
tric Education and the AOA Council on Clinical Optometric
Care. I have also participated as an on-site evaluator on
regional accrediting bodies and on the National Accreditation
Council for Agencies serving the Blind and Visually Handi-
capped. In my view accrediting bodies serve a vital role as a
catalyst for constructive change and development.
555
The purpose of this paper is to consider several aspects of

accreditation as they relate to the promotion of quality
patient care and services in low vision rehabilitation pro-
grams in a clinical setting either within optometric institu-
tions or within agencies serving blind and visually impaired
persons.
This paper will consider 4 aspects of the accreditation

process:
o Accreditation defined
o The role of the National Accreditation Council (NAC)
o The Self Study as a central facet of the accreditation

process and
o The On-Site Team Report as another important compo-

nent of the total accreditation sequence
2. Accreditation Defined
A useful working definition of accreditation appears in

KELL's book entitled Self Study Processes [1]:
Accreditation is a voluntary, nongovernmental

process conducted by postsecondary institutions
or agencies to accomplish at least two things -
to attempt to hold one another accountable on a
periodic basis to live up to stated, appropriate
institutional or agency goals; and to assess the
extent to which the institution or agency meets
established standards. The major purposes of
the process are to foster improvement and to
identify institutions and programs that seem to
meet the agreed-upon standards.
SELDEN and PORTER [2] have identified additional uses

and values of accreditation:
Internal Uses:
556
o Identifying an institution or program as having met

established standards
o Encouraging the involvement of faculty or consultants

and staff in study and planning
o Stimulating self-improvement and thereby generally

enhancing quality
External Uses:
o Assisting in the identification of institutions and pro-

grams for the investment of funds
o Providing one basis for the determination of eligibility

for federal funds
o Serving as an instrument for the enforcement of social

policy
Professional Uses:
o Acting as one source of criteria for professional certifi-

cation and licensure
o Serving as a lever to gain increased support for a pro-

gram or programs
3. The Role of the National Accreditation Council (NAC)
The National Accreditation Council for Agencies Serving the

Blind and Visually Handicapped was founded 20 years ago, in
1966, by both blind and sighted leaders who were concerned
about the uneven quality of services available to blind and
visually handicapped Americans (National Accreditation
Council for Agencies Serving the Blind and Visually Handi-
capped, 15 West 65th Street, 9th Floor, New York, New York
10023). The NAC has a twofold commitment involving per-
sonal and professional goals -- to help blind and visually hand-
icapped people become more self reliant and capable of
achieving through education and training their fullest poten-
tial; and to help agencies and schools improve their manage-
ment and services through a comprehensive standards and
accreditation program.
557
In 1981, the National Accreditation Council completed a

two-year project to develop standards for evaluation of low
VISIon services. The published standards, developed by a
National Technical Committee of educator-administrators,
ophthalmologists, optometrists and consumer representatives,
serve as the criteria for evaluation and accreditation of low
vision programs.
The NAC standards are organized into 6 main areas. Let

us review the six areas and some of the major components in
each category.
3.1 Planning and Organization
o Philosophy of low vision care
o Goals and objectives of the service
o Board of Directors -- membership, qualifications, role
o Administrative organization including position descrip-

tions and nature of interactions
o Fiscal policy, controls and stability
o Pa tient population served
o Patient admission procedures including eligibility cri-

teria
o Marketing and fund raising activities
o Planning - short and long range participation in planning
3.2 Personnel
o Administrator of the service
o Clinical Director of Low Vision Service
o Low Vision Clinician(s)
o Low Vision Assistant

558
o Ophthalmological and Optometric Consultants and the

nature of their interaction
o Support staff and services
o Volunteers
3.3 Low Vision Evaluation
o Patient's background information (social, medical, trav-

el ability, life style, goals, etc)
o Clinical examination: basic and supplemental procedures
o Patient management
o Instruction and adaptive training
3.4 Continuing Services
o Provisions for other specialized instruction, training and

services:
o Orientation mobility
o Rehabilitation teaching
o Social work
o Special education
o Vocational services
o Workshop services
o Low vision re-evaluation and continuing care

559
3.5 Facilities and Equipment
o Accessibility and design of facility
o Space adequacy and utilization
o Nature and type of examination equipment and its main-

tenance
o Printed materials and aids for testing and instruction
o Loan of devices and materials
3.6 Program Evaluation
o Nature and scope of program evaluation to determine

the effectiveness and efficiency of the service
o Adequacy of available resources - finances, personnel,

facilities, materials and equipment
o Annual and long range plans
o Professional development including continuing education

opportunities
o Research and information dissemination
o Utilization of self study findings for: planning, budget-

ing internal and public report of program activities,
marketing the low vision service
It should be noted that unlike some accreditation stan-

dards which are stated quantitatively and tend to be applied
in a prescriptive way, NAC standards are stated in qualitative
terms. Many experienced low vision practitioners in this audi-
ence recognize that low vision care in a clinical setting and
practiced competently often reflects more individualized
behaviour and preferences especially as it relates to decisions
on the type and design of low vision devices. In my opinion
most low vision standards can be interpreted and evaluated
consistently while allowing for some individuality and vari-
ability in approach.
560
To date, there are 105 accredited organizations serving a

total of some 260,000 people. In many disciplines awareness
of the significance of accreditation has increased as federal
support for social programs and education has decreased.
Organizations are being asked today more than ever before to
justify their services and establish worthiness for support. In
this age of accountability and quality control, accreditation
offers a reliable means of ascertaining which agencies or
schools are providing sound services.
4. The Self Study as a Central Facet of the Accreditation

Process [3]
Prior to an on-site visitation by an evaluation team, an agen-

cyor school is required to conduct a comprehensive self-
study of its services and management. This pre-visitation
self-study involving organized participation by members of
the various segments of an agency or school has numerous
purposes and values.
The use of self-study can help agencies and programs

improve by clarifying goals; identifying problems; studying
goal achievement; reviewing and assessing programs, proce-
dures, and resources; and identifying and introducing needed
changes during and as a result of the self-study. Moreover,
the utilization of self-study can serve as a firm foundation
for and the basis of all planning efforts. Planning should be
based upon a clear sense of strengths and weaknesses. Honest
self-analysis provides the confidence for an agency or school
to project newly clarified goals and the means for their
attainment.
Most self-studies seek answers to the following questions:
1. What are the agency's goals? Are they clear, appro-

priate, and useful? Are they understood by profes-
sional and non-professional staff? Is there a consen-
sus on them?
2. Are the programs and services consistent with the

goals? What are the problems? How can they be
solved?
561
3. Are the resources (human, fiscal, and physical) avail-

able to carry out the programs and services? Will
they continue to be available? If yes, what steps
have been taken to assure continuity of service? If
no, what corrective measures are being taken?
4. Are the goals being achieved? How can evidence,

systematically gathered about the extent of achieve-
ment, be used to improve the agency or institution?
KELLS [4] uses a linear systems model (Fig. 1) of a program

self-study process. He identifies the external and internal
forces and their dynamic interplay in goal achievement and
problem solving. This schematic representation of the rela-
tionships of the various forces at work in the self-study pro-
cess includes: (1) the internal forces which include all individ-
uals involved in the functioning of the program; and (2) the
external forces which include the entire environment of the
institution. Many of these forces cannot be either controlled
or significantly influenced. They are also frequently unpred-
ictable; one can predict only that they will change over time.
This constant environmental flux makes it necessary that
self-study be a continuous, as well as a cyclical process.
In summary, the self-study process can be an integrating

force for the agency or school and its rehabilitation planning.
It can provide a focus for all planning efforts to improve the
facility and its program. Therefore, the self-study must be
managed, planned, organized, staffed, directed, studied and
controlled.
562
Linear Systems Model for Self-5tudy Process
FIGURE 1
Adapted from H.R. Kells, Self-Study Processes: A Guide for Postsecondary Institutions, Washington:
American Council on Education, 1980.
563
Diagram of the Comprehensive

Self·Study Approach
_I
R
I
~_----, E
~ OUTWMES
L-_--' E
S
FIGURE 2
From H.R. Kells, Self-Study Processes.' A Guide for Postsecondary Institutions, Washington: Ameri-
can Council on Education, t 980.
5. The On-Site Team Report as Another Important Com-

ponent of the Total Accreditation Sequence
Following review and acceptance of the self-study report, an

on-site team, whose members represent a variety of compe-
tencies and experiences, is appointed. This team has a
variety of tasks.
o Examine the organization's goals and objectives
o Evaluate achievements in relation to the goals and

objectives
o Assess strengths and plans to develop them further
o Assess limitations and plans to correct them

564
o Examine plans for future change and development in

relation to the overall long range plan
o Assess practices and procedures against professionally

accepted standards
The evaluation team will then submit a report that the

low vision rehabilitation service can use to reinforce its own
findings or, when there are differing conclusions, re-examine
itself in the light of an independent review by qualified,
experienced persons. Thus, the functions of the team can
actually confirm and enhance the benefits received from self-
study.
One might ask - are optometry school low V1S10n pro-

grams eligible for NAC accreditation? The answer is yes.
Only one low vision service in an American or Canadian
school of optometry has been accredited by NAC - the Wil-
liam Feinbloom Vision Rehabilitation Center in Philadelphia,
PA [5]. This is unique in that the Feinbloom Center is the
first organization not affiliated with a multiservice agency to
achieve NAC accreditation. There is also a one year low
vision residency and a two year regionally accredited Master
of Science program in Rehabilitation Optometry. There is at
least one qualitative difference in the. accreditation process
f or an agency low vision service and a low vision service
within a school of optometry. The teaching-supervision
dimension within an optometry school may compromise effi-
cient low vision patient care and management. Consequently,
the NAC evaluation must consider the quality of this instruc-
tional interaction and determine whether the self-study has
adequately considered this activity.
The New York Lighthouse Low Vision Service is another

NAC-accredited program with intensive education and train-
ing in low vision for students, optometrists, ophthalmologists
and low vision administrators. The American Optometric
Association's Council on Optometric Education established
standards for and has accredited seven Veterans Administra-
tion residencies in Rehabilitation Optometry or Low Vision.
Accreditation of a low vision rehabilitation service in a

clinical setting need not be limited to America or Canada.
The accreditation process can be applied internationally - but
as a voluntary, non-governmental activity, the agency or
institution must officially request an accreditation eligible
565
status from NAC and be prepared to meet the administrative,

staff, and financial responsibilities involved.
Let me urge my colleagues associated with low VISIon

clinics in American or Canadian optometric institutions to
consider the inherent value of the accreditation process and
make the commitment to self-study as an important first
step.
In the final analysis, accreditation is a voluntary process

that requires the co-operation of its many constituencies.
When it functions at its best, accreditation is successful in
helping organizations function more efficiently and effective-
ly. According to the former President of NAC, Dr. James D.
McComas, accreditation provides
an ongoing mechanism for improved agency

planning and program evaluation and develop-
ment; assists organizations in securing funds
from governmental bodies and corporate, foun-
da tion, and individual sources; and brings
together trustees, staff, consumers, and mem-
bers of the community to identify agency and
school strengths and weaknesses and set objec-
tives for the future.
Accreditation can be an effective means for achieving

our collective professional goal of excellence in the care and
rehabilitation of blind and visually impaired persons to a more
self reliant, useful life.
566
6. References
1. H.R. Kells: Self-Study Processes: A Guide for Postse-

condary Institutions. (American Council on Education,
2. W.K. Selden, H.V. Porter: Accreditation: Its Purposes

and Uses. (Council on Postsecondary Accreditation,
3. A.A. Rosenbloom: Relationship of the self-study pro-

cess to institutional effectiveness and accredita..tion.
Washington J. Optom. Educ. 7, 13 (1981)
4. Kells, Ope Cit.
5. P.D. Zion: The Feinbloom Center - A Leader in Ser-

vice. Houston Rehab. Optom. Fall 1984.
A Unique Model for a Resource/Rehabilitation Centre
for Consumers with Low Vision
Bill Shalinsky, Peter Shaw, Bill Carroll
1. Introduction
We are speaking today as representatives of the Low Vision

Association, a consumer self-help organization with approxi-
mately 3,000 members across the Province of Ontario. Our
Association came into existence in 1982 to encourage broader
recognition of the needs of low visioned consumers in Ontar-
io, to encourage the establishment of comprehensive low
vision services across the Province and to create a self-help
organization which would provide a consumer perspective in
these efforts.
The Resource/Rehabilitation Centre has emerged as an

important demonstration project, based on our assessment of
the current need:
2. The Need
In the Province of Ontario, there are approximately 90,000

citizens with some form of low vision. Although there are
many services available, they are usually provided on a piece-
meal, fragmented basis and are often inaccessible to those in
need by reason of distance, cost and the complexity of the
referral system. In addition, existing services are rarely con-
sumer controlled.
For these reasons, the Low Vision Association of Ontario

concluded that a new type of Resource/Rehabilitation Centre
f or low visioned consumers was necessary.
In the remainder of this paper, we will discuss the Centre

and its evolution under seven headings:
o The Essential Features of the Centre

568
o The Purposes of the Centre
o Some Terms
o The Objectives of the Centre
o Implementation -- How the Centre Will Work
o The Process by Which the Model Evolved, and
o Future Steps
3. The Essential Features of the Centre
The essential features of our Centre will hopefully correct

the weaknesses of the existing models.
a) It is consumer focused and consumer controlled: in

this way, we hope to avoid the development of an authoritari-
an and patronizing mentality which all too often exists in the
service system for low visioned persons.
b) It attempts to provide a more comprehensive approach

to services than existing clinics or centres. We will use exist-
ing community resources and services where possible, but in a
new and creative fashion so that the current fragmentation is
replaced by an integrated process of providing assistance.
In this way, our model also avoids the isolating of the low
visioned population in a closed service system which focuses
only on those with visual impairments. Instead, the model
will encourage consumers to make extensive use of communi-
ty resources and facilities, beyond the basic vision-related
services provided by the Centre.
c) The Centre exists outside the confines of "blindness"

agencies and organizations; it recognizes, instead, that the
needs of people with low but usable vision are distinctly dif-
ferent from the needs of those who are functionally blind.
569
4. The Purposes of the Centre
The following principles form the foundation on which our

model is built:
a) The Centre is based on the philosophy that the con-

sumer is the central focus of the program.
b) The Centre will serve as a co-ordinator of services for

low visioned coqsumers.
c) The Centre will maintain an independence which per-

mits it to recommend whatever course of action seems appro-
priate.
d) The Centre will adopt an advocacy stance, meaning

that it is prepared to go to bat for a consumer if this seems
appropriate.
e) The Centre and its staff must be accountable to the

consumer as the bottom line.
f) As part of its mission, the Centre will work to sensi-

tize the whole community to the requirements of the low
visioned and to establish positive and constructive attitudes
toward low vision on the part of the public and private agen-
cies and their staffs, as well as in the community at large.
g) The Centre and its staff must be alert to consumers

who have multiple needs.
h) The Centre does not provide funding or financial sup-

port, but will attempt to direct consumers to sources of fund-
ing or financing for assistive devices.
570
5. Some Terms
At this point, we want to clarify further what we mean by the

terms, Consumer, Consumer-led and Advocacy.
A Consumer is different from a "patient". A consumer is

actively involved in the total process of receiving service, in
assessing the service and in controlling the process so that it
meets his/her needs. By way of contrast, the term "patient"
conjures up the image of a helpless pawn in the hands of
experts - well meaning and competent experts to be sure, but
all too often, insensitive to the social and psychological needs
of their "patients". A consumer focus will avoid the latter.
A consumer-led Centre means that low visioned consum-

ers control the atmosphere in the Centre and the processes by
which staff provide service. In this way, we will make sure
that things don't become too bureaucratic or officious and
that staff remain sensitive to the total needs of consumers.
Advocacy means that an individual or organization will go

to bat for someone when that person needs outside help to
achieve some desired end. The Low Vision Association of
Ontario has practised advocacy on many occasions; for exam-
ple, when a young visually impaired woman was entitled to
and needed funding to purchase a sophisticated computer sys-
tem with large print capability in order to obtain a job. In
spite of the fact that the government Ministry responsible for
providing such funds agreed she needed the equipment, the
funds for this equipment were just not made avai,lable. We
interceded on her behalf and after eight months of struggle
were successful.
6. The Objectives of the Centre
The more specific objectives for the Resource/Rehabilitation

Centre follow:
o To provide a comprehensive low VlSlon resource/

rehabilitation service to low visioned consumers.
571
o To inform low visioned people of sight-enhancing vision

aids.
o To provide peer support and encouragement to low

visioned consumers.
o To help low visioned people to overcome the traumatic

effects of vision loss and isolation from family and
friends.
o To increase awareness of the resource/rehabilitation

Centre in the community in general - among the public,
the media and relevant professionals.
o To create public awareness of the requirements of low

visioned people.
o To provide current information and appropriate refer-

rals to resources and professionals in the low vision
field.
o To develop and manage regional consumer-run resource/

rehabilitation Centres in Toronto, Sudbury, London and
Ottawa.
o To record the expressed needs of low visioned consum-

ers.
7. Implementation - How the Centre will Work
This model stresses a small Centre where a core staff will

provide certain basic vision related services and will, to the
extent possible, use community-based resources and services
to meet other needs. Basic services to be provided at the
Centre include an initial assessment of the consumer's situ-
ation and needs, a core vision assessment by the low vision
clinician (optometrist) and rehabilitation work.
The Centre's focus is on the low visioned consumer - his/

her needs, social context or environment and his/her best
interests. A low visioned consumer is defined as anyone,
regardless of age, whose degree of vision impairment serious-
ly limits his/her ability to perform visually and who is likely
to benefit from sight-enhancing strategies.
572
One of the strengths of this model is its flexibility; it

provides only a core service and makes extensive use of com-
munity resources.
This model assumes the need for close contact with

appropriate health care specialties and community services in
the form of referrals from the Centre, referrals to the Cen-
tre and consultation.
The range of community resources is extensive: the

appropriate choices will be based on consumer needs and will
be made by the consumer with the help of Centre staff.
Examples might include: Canada Employment, Family Coun-
selling Agencies, existing low vision resources, Ontario's
Assistive Devices Program, and various Health Care facili-
ties.
Four Centres are proposed, so that services are available

in each of the four major regions of the Province - Central,
Northern, Southwestern, Eastern. These Centres will initially
be in Metro Toronto and in Sudbury, followed by others in
London and Ottawa.
Centres at the hub of rural areas (for example, the one in

Sudbury) might require a slightly different structure because
of the more extended geographical area; for example, in
these Centres, the core staff will have to identify appropriate
services in the consumer's home community and, when the
consumer requires it, help the consumer to take advantage of
the resources and services available in his/her locale. Addi-
tionally, it is suggested as a second stage that satellite or
associate centres may need to be established in more remote
areas of the province (for example in Thunder Bay which is
1004 miles North West of Sudbury).
Specifically, the recommended model will use a small

team of staff, including:
a) A resource co-ordinator
b) A low vision clinician - optometrist
c) A low vision rehabilitation person or specialist
d) A clerical person
573
Beyond their day-to-day activities these staff members

will serve as "community scanners" and "activists", identify-
ing appropriate services, identifying gaps in services and
actively trying to fill these gaps.
Brief job summaries for the core staff are as follows:
a) Resource Co-ordinator: This person will co-ordinate

staff, resources and budget, etc. In addition the resource
co-ordinator will provide the in-take service, collecting all
needed inf orma tion.
Then he/she will provide or oversee the provision of a

variety of services after the clinical exam.
These services might include discussion with the consum-

er, referral to a variety of community services, with careful
monitoring of the results and periodic assessment as to the
appropriate plans of action with the consumer. As already
noted, this person should also serve as a community scanner,
surveying and assessing community resources and services
which might be of use to consumers and assessing gaps in
available services.
b) Low Vision Clinician - Optometrist: In order to pro-

vide the consumer with appropriate therapy and decision
making information, an essential part of this model will be
the provision of a comprehensive low vision assessment. Low
vision assessments will be provided by optometrists with
expertise in low vision evaluation procedures and therapy
stra tegies. In addition to providing direct clinical services
the low vision clinician will supervise the rehabilitation pro-
cess.
c) Low Vision Rehabilitation Person/Specialist: This per-

son will assist the low vision clinician in the assessment of
the consumer's vision needs, in the selection of appropriate
aids and in instructing the consumer in the use of these aids.
In carrying out these duties the low vision rehabilitation per-
son may do the following things:
o Visit the consumer at work and at home during the

needs assessment
o Participate in office instruction sessions with the con-

sumer
574
o Visit the consumer at work and at home to assist the

consumer in using vision aids optimally in various set-
tings
o Assist the consumer in assessing the appropriateness of

various aids
o Maintain a loaner inventory for trial use by consumers.
d) Clerical Person: This person will assist the resource

co-ordinator to provide essential services by doing the follow-
ing:
o serve as receptionist
o provide day-to-day clerical service
o answer telephone
o order supplies and vision aids
o carry out basic bookkeeping
8. Evaluation Procedures
An innovative program requires careful evaluation procedures

to assess the operations and results of the Centre. We are
currently creating such an evaluation system for the Centres
to permit an assessment of their activities.
9. The Process by Which the Model Evolved
The present model was developed by a special committee of

the Low Vision Association of Ontario which met eight times
over a twelve month period. The committee itself was spe-
cially formed to represent both consumers and specialists in a
wide range of fields. The consumers brought a wealth of
experience and, in some cases, professional expertise. Over-
all, the committee, which came to be known as the Profes-
sional Advisory Committee, included the following profes-
575
sional specialties among others: Optometry, Ophthalmology,

Psychiatry, Nursing, Social Work Rehabilitation, and Educa-
tion.
During the twelve month period, the committee carried

out an extensive review of the literature on low vision clinics
and resource centres. Based on this information and the
experience of consumers, five possible models of resource/
rehabilitation centres were developed. The positive and neg-
ative features of each were then listed and evaluated. Out of
this process the model just outlined emerged.
The four other models were:
Model One: Community Based Intake and Service Deliv-

ery. In this model, intake managers would be located in many
community agencies; that is, one person would be designated
as "the low vision person" in each agency and would carry out
the tasks assigned to the Resource Co-ordinator which we
outlined earlier. This model was rejected because there
would be far too many people looking after the needs of the
low vision population, thus creating a fragmented, chaotic
situation. In addition, many agencies have age and other
restrictions, thus barring some people from service.
Model Two: Centralized Services. In this model, all servi-

ces are provided under one roof. In spite of its attractive
features, this model was rejected because it isolates the low
vision consumer within one service system; in addition, it pro-
vides dangerous opportunities for controlling the lives of low
visioned consumers as outlined by SCOTT [1].
Model Three: A Joint Venture. In this model, it was sug-

gested that a neutral Centre be created as a joint venture
between the Canadian National Institute for the Blind (CNIB)
and the Low Vision Association of Ontario (LVAO). This mod-
el was attractive because it would use the already existing
facilities of CNIB across the Province, and, in fact, across
the country. However it, too, was rejected because of major
differences in philosophy, rejection of a blindness agency by
many low visioned consumers and anticipated difficulties in
working cooperatively on a joint venture (based on present
and past experiences).
Model Four. A Government-Run Service. This is the way

low vision services are provided in Quebec and elsewhere in
576
the world. Whatever its merits, it seems so unlikely in Ontar-

io that it was not pursued.
Other models which represent existing clinics and centres

around the world were seen as unsatisfactory in various ways
and were also rejected in favour of our preferred choice.
10. Future Steps
What next? We are very pleased to report that the Low

Vision Association and the Optometric Institute of Toronto
(OIT) have just completed an agreement by which the OIT will
assume responsibility for providing the optometric services in
each Centre. This is an important step in making the Centres
a reality.
We had hoped to have two Centres operating by this time;

however, funding problems have delayed things. We estimate
that we need approximately $150,000 to equip each Centre.
We are actively pursuing this money and are ready to move as
soon as it is available. .
11. Conclusion
To conclude, the Resource/Rehabilitation Centre model this

paper describes is unique because it is 1) consumer focused
and 2) consumer controlled - both safeguards against an
authoritarian patronizing mentality. Fundamental to this
concept is peer support and peer encouragement.
The Centre will provide only core, vision-related services

and will use the community-based resources and services, of
the low visioned consumer's choosing, to meet his/her other
needs. This approach avoids isolating the low visioned popu-
lation in a closed service system. Combining its own core
services and community-based services the Centre will pro-
vide comprehensive service to low visioned consumers.
Existing outside the confines of a blindness agency, the

Centre recognizes that the needs of the low visioned are dis-
tinctly different from those of the functionally blind.
577
The Centre, which will be fiercely independent, will serve

as coordinator of services for low visioned people.
Finally, the Centre will advocate for - go to bat for -

consumers where it becomes clear that the consumer's own
efforts have failed him/her in obtaining the service he/she
requires.
12. References
1. R. Scott: The Making of Blind Men (Russell Sage

Foundation, New York 1969)

Low Vision Principle and Applications

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Low Vision Principle and Applications

Uploaded by

Copyright:

Available Formats

Low Vision

With 145 Illustrations

Library of Congress Cataloging in Publication Data

© 1987 by Springer-Verlag New York Inc.

ISBN-13: 978-1-4612-9152-7 e-ISBN-13: 978-1-4612-4780-7

The low vision clinic at the School of Optometry, Univer-

The School of Optometry has also become an important

The School of Optometry is proud of this record of

In July of 1983, the idea of hosting a low vision symposium

Excellent research and clinical papers were presented by

My special thanks go to Lynne Hanna without whose

The objective of the symposium was to provide an oppor-

Financial support for the symposium was provided in part by

Bausch and Lomb

1. Measurement of Vision Loss: Theory and Practice

Color Plates following page 10.

New and Improved Contrast Sensitivity Approaches

The Evaluation of the Reading Capability of Low

Threshold and Suprathreshold Deficits in Color

Contrast Sensitivity • •••••••••••••••••••••••••• 45

The Effect of Blur Upon Psychophysical Receptive

Quantifying the Magnitude of Visual Impairment

Spatial vs. Temporal Information in Suspected and

Low Vision Management in Selected Eye Diseases ••••• 96

The Role of X and Simple Cells in the Contrast

Differential Retinal Structural Damage Exhibited

The Visual Requirements of Mobility • •.••••••••••• 134

Visual Acuity Deficits and Chromatic Aberration in

Measurement of Central Fields Following Macular

Detection of Visual Field Defect Using Topographic

Preliminary Study of Topographic Visual Evoked

2. Prescribing Low Vision Aids

Prescribing Magnification: Strategies for Improving

The Amorphic Fresnel Prism Trioptical System •••••• 209

Sensorimotor Adaptation to Telescopic Spectacles • ••• 216

Paradoxical Cases of Visual Improvement Offered

Magnification Efficiency in the Low Vision Patient ••• 249

An Overview on the Use of a Low Magnification

3. Reading in Low Vision

Observations from the Psychology of Reading

Contrast Polarity Effects in Low Vision Reading • •••• 288

Effect of Magnification and Field of View on

Predicting Reading Performance in Low Vision

4. Low Vision Care

Visual Impairment and Disability: Enhancement and

The Clinical Profile of a Young Visually

Pathology Characteristics and Optical Correction of

Assessment of Vision of Deaf-Blind Persons: A

An Evaluation of Follow Up Systems in Two Low

Vision Examinations of Handicapped Children at the

Accommodation in the Visually Impaired Child •••••• 4ZS

Abnormal Arm Tone, Cigarette Smoking and Use of

The City Study - Preliminary Findings • •••••••••••• 448

The Silver Pages: Are They Easier to Read? •••••••• 463

A Hierarchy of Perceptual Training in Low Vision • ••• 471

Low Vision Performance as a Function of Task

1. Some Issues in Rehabilitation in Low Vision

National Long Term Care Facility Survey •••••••••• 502.

Rights of Low Vision Children and Their Parents • •••• 519

Life Satisfaction of Low Vision Patients and Other

Sight Enhancement Services - A Safety Net or a

Strengthening Low Vision Rehabilitation Through

A Unique Model for a Resource/Rehabilitation