Potenciales Evocados Aaitif M Husain

Illustrated Manual of
Illustrated Manual of Clinical Evoked Potentials

An Imprint of Springer Publishing
Clinical Evoked Potentials

AATIF M. HUSAIN, MD
Evoked potentials have been used for decades to assess neurologic function in outpatient
studies and are now routinely used in the operating room during surgery. Illustrated Manual
of Clinical Evoked Potentials is a modern, practical guide to performing these studies and
Clinical Evoked
interpreting the results. The book is uniquely organized as a singular resource that provides
the necessary background for understanding and conducting evoked potential studies. It
functions as a multi-purpose text, atlas, and reading session, with numerous examples of
studies and findings and discussion of key takeaways.
Divided into five chapters, the book opens with an introduction to the basics of data
acquisition and interpretation that lays the foundation for the modality-specific chapters that
Potentials
follow. The next group of chapters are in-depth reviews of visual, brainstem auditory, and
somatosensory evoked potentials. Each of these chapters lays out the specifics of the modality
and study protocol with examples to show how things should—and should not—be done.
Sample studies with discussions about how to interpret them highlight a particular aspect of
normalcy or pathology. Imaging correlates are provided to emphasize salient points and offer
perspective. The final chapter is an overview of the use of evoked potentials during surgery
with imaging and case discussions to introduce the reader to this very important application.
AATIF M. HUSAIN
Neurologists, clinical neurophysiologists, technologists, trainees, or anyone who has to
interpret evoked potential studies, even intermittently, will find this illustrated manual to be a
trusted companion in gaining proficiency and avoiding common pitfalls.
Key Features
n Detailed review of methodology of evoked potential studies

n Many examples of actual patient studies with imaging correlates
HUSAIN
n Interpretation of each evoked potential study presented in detail
n “Reading session”-like discussion of each example
n Special chapter on evoked potentials in the operating room
Recommended
Shelving Category:
Neurology
An Imprint of Springer Publishing

11 W. 42nd Street
New York, NY 10036
www.demosmedical.com 9 781933 864723
An Imprint of
Springer Publishing
Aatif M. Husain, MD
Department of Neurology
Duke University Medical Center
Durham, North Carolina
Neurodiagnostic Center
Veterans Affairs Medical Center
Durham, North Carolina
Visit our website at www.demosmedical.com
ISBN: 9781933864723
ebook ISBN: 9781617050107
Acquisitions Editor: Beth Barry

Compositor: diacriTech
Copyright © 2018 Springer Publishing Company.

Demos Medical Publishing is an imprint of Springer Publishing Company, LLC.
All rights reserved. This book is protected by copyright. No part of it may be reproduced, stored in a retrieval system, or transmitted
in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission
of the publisher.
Medicine is an ever-changing science. Research and clinical experience are continually expanding our knowledge, in particular our
understanding of proper treatment and drug therapy. The authors, editors, and publisher have made every effort to ensure that
all information in this book is in accordance with the state of knowledge at the time of production of the book. Nevertheless, the
authors, editors, and publisher are not responsible for errors or omissions or for any consequences from application of the informa-
tion in this book and make no warranty, expressed or implied, with respect to the contents of the publication. Every reader should
examine carefully the package inserts accompanying each drug and should carefully check whether the dosage schedules men-
tioned therein or the contraindications stated by the manufacturer differ from the statements made in this book. Such examination
is particularly important with drugs that are either rarely used or have been newly released on the market.
Library of Congress Cataloging-in-Publication Data

Names: Husain, Aatif M., author.
Title: Illustrated manual of clinical evoked potentials / Aatif M. Husain.
Description: New York: Demos Medical, [2018] | Includes bibliographical
references and index.
Identifiers: LCCN 2017013583| ISBN 9781933864723 | ISBN 9781617050107 (e-book)
Subjects: MESH: Evoked Potentials
Classification: LCC QP376.5 | NLM WL 102 | DDC 616.8/047547—dc23
LC record available at https://lccn.loc.gov/2017013583
Contact us to receive discount rates on bulk purchases.

We can also customize our books to meet your needs.
For more information please contact: sales@springerpub.com
Printed in the United States of America by Publishers’ Graphics.

17 18 19 20 21 / 5 4 3 2 1
To my loving sons, Aamer Aasim Husain and Aayaz Siraj Husain—you bring me more
joy than I could have imagined. Without you, life would be . . . black and white.
Contents
Preface ix
Share Illustrated Manual of Clinical Evoked Potentials
1. Basics of Evoked Potentials 1

2. Visual Evoked Potentials 35
3. Brainstem Auditory Evoked Potentials 113
4. Somatosensory Evoked Potentials 191
5. Evoked Potentials During Surgery 265
Index 311
vii
Preface
Is there really a need for a book on evoked poten- societies offer few courses on evoked potentials.
tials? Aren’t they a test of the past that is not used There are no contemporary books on this topic, and
anymore? most of the older ones are out of print and not avail-
Nothing could be further from the truth. Clinical able. The available books have few illustrations of
evoked potentials remain an integral part of actual evoked potential studies. This has led to very
neurophysiological assessment of patients with few resources from which to learn about evoked
neurologic and nonneurologic diseases. Most neu- potentials for neurologists, clinical neurophysiolo-
rophysiology laboratories continue to perform clin- gists, trainees, technologists, and allied health care
ical evoked potential studies, perhaps a little less professionals.
often than a few decades ago. Adults and children Illustrated Manual of Clinical Evoked Potentials will
with many different neurologic conditions continue help fill this void. It provides contemporary infor-
to need assessment with evoked potentials. New mation about clinical evoked potentials. This book
areas of utility are being discovered in the assess- is uniquely organized with the chapters having a
ment of coma and other conditions. Evoked poten- narrative section, sample evoked potential studies,
tials are used on a daily basis in the operating room; and a discussion about those studies. The prose of
an understanding of the basics of evoked potentials this book provides the necessary background for
is essential to effectively use them for monitoring interpreting clinical evoked potential studies. The
during surgery. Advanced and research applications use of many sample studies gives this book the feel
with event-related and cognitive evoked potentials of an atlas. The discussion provided with the exam-
are opening other avenues of use. A Google Scholar ples makes it feel like a “reading session.” Together
search of clinical evoked potentials in early January the three pieces make Illustrated Manual of Clinical
2017 revealed about 1,190 hits since the start of 2017 Evoked Potentials a single resource that can be used
and 30,200 hits since 2016. Comparatively, a search to teach and learn about this topic.
for clinical electroencephalography netted 878 hits I have divided this book into five chapters. The
since the start of 2017 and 26,100 hits since 2016; first chapter discusses the basics of acquisition and
clinical electromyography had 668 hits since the interpretation. This is an essential chapter to read
start of 2017 and 16,900 hits since 2016. before the others as it will lay the foundation for
What has changed is the education available further learning. The second, third, and fourth
about evoked potentials. Neurology residencies chapters review visual, brainstem auditory, and
and clinical neurophysiology fellowships concen- somatosensory evoked potentials, respectively.
trate on other aspects of clinical neurophysiology. Each of these chapters has an initial section that
Often training programs have no one with expertise discusses the basics of the modality with exam-
in evoked potentials, yet laboratories continue to ples to show how things should and should not be
perform these studies. National and international done. This is followed by examples from patients
ix
x ■ Preface
with discussions about how to interpret them. Each the physicians and technologists over the years,
example highlights a particular aspect of normalcy must be commended. Without their support, the
or pathology. Imaging correlates are provided to Duke University Evoked Potentials Laboratory
emphasize certain points and provide perspec- could not have become what it is. Of course, most
tive. The final chapter is an overview of the use of important are the patients that have visited the
evoked potentials during surgery. This section also Evoked Potentials Laboratory; it is they who teach
has examples, discussions, and imaging correla- us, and it is for them that we learn.
tions to introduce the reader to this very important Beth Barry at Demos Medical Publishing must be
use of evoked potentials. recognized and thanked. I am grateful that she not
Neurologists and clinical neurophysiologists only agreed with the need for this book, but pushed
will find this book very useful. Anyone who has to me to complete it. Without her reminders and
interpret evoked potential studies, even intermit- encouragement, I may not have been able to finish.
tently, will find that this book helps them become I would be remiss if I did not pause to remember
more confident and proficient in rendering inter- my parents, Mairaj and Suraiya Husain. Though
pretations and avoiding common pitfalls. Trainees they both have passed away, their memories grow
in these disciplines will find that they finally have stronger with each passing year. The number of
a book they can read cover-to-cover and feel like life lessons they taught me are too numerous to
they have a good understanding of this field. recount, and each day I am reminded of a different
Technologists and others involved in perform- one. They will always be my role models. I hope
ing and interpreting evoked potential studies will they are smiling.
find many useful pearls in this book. Professionals This book is an absolutely special one for me.
involved in neurophysiologic intraoperative mon- It is one whose every word is mine. At no point
itoring will find an excellent source of information could I call a coauthor or contributor and request
about evoked potentials that will have great appli- changes or demand completion of a chapter. As
cability to what they do in the operating room. editor of my own work, I often felt the author
Neurophysiology laboratory managers will find needed to completely rewrite things and sent the
this to be a source of reference for their technolo- writing back to him! All of this took much more
gists and providers. time than I expected; time that came at the expense
There are many people that I must acknowl- of my wonderful family. I spent many hours work-
edge, without whom this book would not have ing on this book instead of being with them. My
materialized. Foremost, I must recognize the two lovely wife, Sarwat Mohsin Husain, was always
people who introduced me to evoked potentials very understanding of my not spending time with
and mentored me for many decades, Dr. C. William her and always let me know that she and the boys
Erwin and Andrea Erwin. Their guidance and would be waiting for me when I was finished.
support has been invaluable. The Duke University I really missed my sons, Aamer and Aayaz, while
Evoked Potential Laboratory has been second to writing this book. There were too many basket-
none in the quality of work they do. I have been balls that I missed shooting, too many football
fortunate to work with the extremely talented passes I did not throw, and too many squash prac-
and dedicated technologists of that laboratory. It tices that I did not attend. Yet never did they show
is their hard work that I am sharing in this book. their disappointment; only their delight when we
Though I have learned extensively from all of finally did play together. Only because of Sarwat,
them, two must be named. Sharon Elliott and Aamer, and Aayaz’s love and support have I been
Emily Kale are two extremely talented individuals able to complete this labor of love.
that I have had the good fortune to learn from. The
laboratory’s administrators, who have supported —Aatif M. Husain, MD
1
Basics of Evoked Potentials
Evoked potentials are electrical potentials recorded with an auditory stimulus, most often a broadband
over various parts of the nervous system in click. Somatosensory evoked potentials (SEPs)
response to sensory or motor stimulation. Each stim- are obtained with electrical stimulation of periph-
ulation results in a low amplitude evoked response. eral nerves. Most often large mixed nerves are
In order to adequately visualize and measure these stimulated to obtain SEP. In the upper extremity,
responses, evoked potentials are usually aver- median and ulnar nerves are used most often, and
aged and amplified. When many evoked responses in the lower extremities tibial and peroneal nerves
have been averaged, they are referred to as an are used.
evoked potential. Each evoked potential has a series Each type of evoked potential consists of several
of waves or peaks in response to the stimulus. peaks occurring at latencies ranging from a few ms
to several hundred ms after stimulation. Peaks are
designated as short, middle, or long latency wave-
Types of Evoked Potentials forms depending on when they occur. In general,
Clinical evoked potentials performed in a clinical short latency waveforms are clinically more useful
laboratory most often involve stimulation of one as they are more easily reproducible, consistently
of the sensory systems and recording the evoked identified in normal subjects, and resistant to med-
potentials off peripheral nerves, spinal cord, or ication effects. Short latency AEPs occur within
brain. The motor pathways can also be stimulated 10 ms after stimulation and are called brainstem
with an electrical or magnetic stimulus; however, auditory evoked potentials (BAEPs). Short latency
neither of these is done routinely in an outpatient SEP occur up to 30 to 50 ms after stimulation and
environment. Electrical stimulation is too pain- are used clinically. VEP responses are long latency
ful in an awake patient, and magnetic stimulation potentials that occur at about 100 ms.
is mostly used in research applications. Evoked
potentials are named according to the neural path- Nomenclature
way that is stimulated.
The individual peaks of the evoked potentials that
are important in clinical evaluation are named
Modalities according to their latency after stimulation, polar-
Three types of evoked potentials are routinely used ity, or number in a sequence. VEP and SEP peaks
in clinical practice: visual, auditory, and somatosen- are named based on their polarity and typical
sory. Visual evoked potentials (VEPs) are obtained latency in a normal adult. Thus, the P100 wave-
with light flashes or a patterned stimulus like a form of the VEP is a positive potential that typically
checkerboard. These are called flash VEP (FVEP) occurs at a latency of 100 ms. The N34 waveform
and pattern reversal VEP (PRVEP), respectively. of the tibial nerve SEP is a negative peak that typ-
Auditory evoked potentials (AEPs) are obtained ically occurs at a latency of 34 ms. BAEP peaks are
1
2 ■ Illustrated Manual of Clinical Evoked Potentials
named according to the order in which they appear; potential whose amplitude is greatly dependent on
consequently the five BAEP peaks are called waves the location of the recording electrode (Figure 1.1).
I–V. Though there is agreement about waveform In contradistinction to near-field potentials are
nomenclature, there is no universal standard on far-field potentials. Far-field potentials are recorded
how waveforms should be displayed. Whereas for from an electrode that is distant to the site of the
most neurophysiologic studies negative potentials waveform generator. These potentials are biphasic,
are displayed with an upward deflection, this is not with the recording electrode seeing only a moving
always true for evoked potentials. Negativity may phase of depolarization through neural pathways.
be displayed as an upward or downward deflec- Far-field potentials are of low amplitude and do not
tion. Regardless of how it is displayed, it should be significantly change in amplitude and morphology
clearly noted on the study. with slight movement of the recording electrodes.
Certain peaks are designated as “obligate wave- The N34 waveform obtained after tibial nerve stim-
forms.” Obligate waveforms are so named because ulation is such a waveform (Figure 1.1). Most of the
their absence usually denotes an abnormality. Not BAEP waveforms (waves II–V) are also far-field
all waveforms of an evoked potential are consid- potentials. BAEP and SEP contain both near- and
ered obligate. In tibial nerve SEP, the lumbar poten- far-field potentials. VEP, on the other hand, consist
tial (LP) waveform is not considered obligate, so its of only near-field potentials.
absence is not necessarily considered abnormal. Evoked potentials can also be thought of as being
cortical, subcortical/spinal cord, and peripheral
nerve evoked potentials. Cortical evoked poten-
Waveform Generators tials are generated by synaptic activity in cortical
Evoked potentials are generated when impulses neurons and the thalamocortical projections. These
travel through the nervous system. The various potentials are usually near-field potentials. The VEP
components of an evoked potential are generated and the later components of the SEP are cortical
when there is synaptic transmission or when an evoked potentials. Subcortical/spinal cord evoked
impulse is traversing a fiber tract, particularly when potentials are produced by impulses moving through
that fiber tract changes direction (1). An example of fiber tracts and brainstem nuclei. Evoked potentials
the latter is when a somatosensory impulse ascends recorded from brainstem structures are far-field
from the dorsal horn to the dorsal column pathway. potentials as the recording electrodes are distant to
In the cervical region, this transition contributes the site of potential generation. Most of the BAEP
to the N13 waveform potential seen after median waveforms are subcortical, far-field potentials.
nerve stimulation. Spinal cord evoked potentials are recorded much
When evoked potentials are recorded close to closer to their site of generation and are considered
the generator site, they are referred to as near-field near-field potentials. The N13 waveform is a spi-
potentials. To record a near-field potential, the nal cord evoked potential. Peripheral nerve evoked
recording electrode is placed directly over or very potentials are recorded from electrodes directly
close to the peripheral nerve, spinal cord, or brain overlying these nerves and represent a wave of
structure contributing to the evoked potential. depolarization traveling through that nerve. These
These potentials are often triphasic, with a large near-field potentials are often of high amplitude,
negativity preceded and followed by a smaller such as the Erb’s point (EP) waveform potential.
positivity. The initial positivity occurs due to a Peripheral nerve evoked potentials are comparable
wave of depolarization approaching the record- to mixed nerve action potentials recorded in an elec-
ing electrode. The large negativity occurs when tromyography laboratory. Anesthetic medications
the depolarization passes beneath the recording affect peripheral nerve evoked potentials the least
electrode, and the final, small positivity occurs due and cortical evoked potentials the most.
to repolarization. Near-field potentials are of high Specific neural structures are often regarded as
amplitude and very sensitive to electrode place- the generators of particular evoked potential wave-
ment. Slight changes in the location of the record- forms. These associations have been made mostly
ing electrode can drastically change the amplitude from lesion and autopsy studies. Waveforms have
of the response. The P37 waveform obtained after progressively ascending neural structures assigned
tibial nerve stimulation is an example of a near-field to them. When a particular waveform is absent or
Chapter 1. Basics of Evoked Potentials ■ 3
Age: 14 years
Sex: Female
Stimulation rate: 5.7/s
Filters: 30–3,000 Hz
Scale: Amplitude = 0.5 mV/div; Latency = 8 ms/div (−↑)
Side: Right
Stimulation duration: 0.2 ms
Stimulation intensity: 18.8 mA
Number of repetitions: 2,000
Tibial nerve conduction velocity
(tibial nerve to T12S): 48 m/s
Absolute Latencies
Waveform Latency (ms) Amplitude (mV)
PF 8.30
LP 20.8 Interpeak Latencies
P31 27.5 Waveforms Latency (ms)
N34 30.8 PF–P37 26.1
P37 34.4 2.38 (P37–N45) LP–P37 13.6
Discussion: The P37 waveform is a near-field potential whose amplitude is greatly dependent on the location of the active recording
electrode. Notice that the P37 waveform is of high amplitude in the CPz–C5S derivation, but moving the electrode to CP3, CP4,
and FPz (top three channels) results in a dramatic change in the amplitude of this near-field potential. On the other hand, the N34
waveform is a low-amplitude, far-field potential. It is clearly visualized in the CPz–C5S derivation, and its amplitude and morphology
do not change with moving the active electrode to CP3, CP4, and FPz.
SEP, somatosensory evoked potential.
Figure 1.1 Tibial nerve SEP near-field and far-field potentials.

prolonged, the site of pathology is at or distal to its Parameters

generator. While clinically useful, this approach is too
simplistic and has some shortcomings. First, while Stimulation parameters include intensity, duration,
particular neural generators are thought to result rate, and repetitions. Varying these parameters may
in certain evoked potential waveforms, it should result in alteration of the evoked potential wave-
be recognized that a single nucleus or fiber tract form morphology and latency.
may contribute to more than one waveform. Each Intensity
waveform in turn may be composed of potentials
The lowest stimulation intensity that is needed to
from many different sites (2). Additionally, evoked
produce an evoked potential is known as the thresh-
potentials do not ascend in the nervous system like
old intensity. As the intensity increases the latency
an electrical signal through a cable, and evoked
decreases and amplitude increases. The latency
potential peaks do not simply represent different
and amplitude change is not linearly related to the
points on this cable. Evoked potentials may ascend
stimulation intensity, and the changes occur up to
through several different pathways and since many
a maximum. Beyond this limit, further increase in
different components contribute to each waveform,
stimulation intensity does not affect the evoked
the waveforms may appear out of order. An example
potentials. Central components of the evoked poten-
of this is sometimes seen with median nerve SEP.
tials reach maximum amplitude earlier than periph-
Usually the N13 waveform (thought to be generated
eral components. This occurs because of “central
in the cervical spinal cord) is seen at a shorter latency
amplification” of the stimulus (3). Central amplifi-
than the P14 waveform (thought to arise from
cation refers to the ability of a peripheral stimulus
the nucleus cuneatus). Sometimes, however, the
to activate more fibers in central pathways than in
P14 waveform has a shorter latency than the N13
peripheral nerves. Consequently, with gradually
waveform (Figure 1.2). This would be impossible if
increasing stimulus intensity, cortical and subcorti-
evoked potential waveforms were simply waypoints
cal evoked potentials will cease to become larger at
in an ascending pathway. In the author’s labora-
a lower intensity than peripheral evoked potential.
tory, the P13 waveform is tagged instead of the P14
This is discussed in more detail in Chapter 4.
waveform; the P13 waveform, the positive wave-
Intensity of stimulation is measured differently
form immediately preceding the P14, is often better
for different types of evoked potentials. For SEPs,
visualized and thought to have the same generator.
the intensity is milliampere (mA) of electrical stim-
ulation. BAEP stimulation intensity is measured by
Stimulation Methodology the decibel (dB) of the auditory clicks. Intensity of
stimulation of PRVEP depends on the contrast ratio
The type of stimulus used to obtain evoked poten- between the light and dark patterns.
tials depends on the type of evoked potential being
performed. A light flash or patterned stimulus is Duration
used for VEP, auditory clicks are used for BAEP, Duration of the stimulus is closely related to the
and electrical stimuli are used for SEP. Specific intensity. It refers to the time that the stimulus is
details of each stimulation type will be discussed in applied. Increasing the duration of the stimulus
later chapters, and here a summary of basic princi- has an effect similar to increasing intensity. Longer
ples is presented. duration stimuli will result in shorter latency and
higher amplitude of the evoked potential wave-
forms with the same caveats as noted above for
Electrodes
intensity.
Stimulating electrodes are used for SEP. BAEP stim-
ulation involves headphones (or ear inserts) while Rate
VEP are obtained with visual stimulation. Though The stimulus rate is the number of stimuli deliv-
both surface and needle electrodes can be used ered per s. It is often denoted in hertz (Hz), but
to deliver the electrical stimulus for SEP, surface this is incorrect, as Hz implies that the stimulus
electrodes are used most often in routine practice. is sinusoidal. The stimulus rate must be such that
Details of electrode properties are discussed in the the entire evoked potential waveform of interest
following sections. is recorded before the next stimulus is delivered.
Age: 56 years
Sex: Female
Scale: Amplitude = 0.5 mV/div; Latency = 3 ms/div (−)↑
Side: Right
Median nerve conduction velocity (median nerve
to EP waveform): 46 m/s
Absolute Latencies
EP 9.90 Interpeak Latencies
N13 13.9 Waveforms Latency (ms)
P13 13.2 EP–N20 9.90
N18 16.8 2.31 (P13–N18) EP–P13 3.30
N20 19.8 1.08 (N20–P22) P13–N20 6.60
Discussion: The N13 waveform is thought to arise from the cervical spinal cord and the P13 (and P14) waveform from the nucleus
cuneatus. In this example, the P13 waveform has a shorter latency than the N13 waveform even though its generator is rostral to
the generator of the N13 waveform. This suggests that individual waveforms cannot be considered to arise from single generators.
In reality, each waveform receives contributions from several generators and each generator contributes to multiple waveforms.
Figure 1.2 Median nerve SEP waveform latency variations.

Thus, the rate is dependent on the latency of the produce a reliable evoked potential. The smaller the
evoked potential waveforms. Short latency wave- amplitude of the evoked potential and higher the
forms require a shorter time window of recording amplitude of noise (low signal-to-noise ratio),
and can be acquired using faster stimulation rates. the more repetitions will be needed. The number of
Long latency potentials require a longer time win- repetitions needed is reduced if the signal-to-noise
dow of recording and need a slower stimulation ratio is improved. To some extent this can be done
rate. For BAEP waveforms, a recording time (or by increasing stimulus intensity and duration and
sweep) of 10 to 15 ms can be used with a stimu- reducing noise. Higher number of repetitions leads
lation rate of 50/s. On the other hand, VEP wave- to better evoked potential waveform morphology
forms often require a sweep of 200 ms, and so must and reproducibility (Figures 1.4A–E). The number
be obtained with a slow rate of about 1 to 2/s. If of repetitions that are recommended for each type
the stimulation rate is too fast, the evoked potential of evoked potentials modality will be discussed in
from one stimulus starts before the evoked poten- the respective chapters.
tial from the previous one has been completed. This
Replication
results in steady state evoked potentials. Steady
state evoked potentials are not routinely used clini- Replication should not be confused with repeti-
cally. The stimulation rate should not be an integer tions. Whenever an evoked potential is obtained,
of 60, as alternating current (AC) in power lines in it must be reproduced at least once to ensure that
the United States is generated at 60 Hz (it is 50 Hz the waveforms recorded are reproducible. This is
in some parts of the world). If the stimulation rate known as replication of the evoked potential. When
is an integer of 60, differential amplifiers may be evoked potentials are not very reproducible, more
unable to eliminate the electrical artifact from the than one replication may be necessary to confirm
evoked potentials recording. the presence of low amplitude waveforms. Most
In general, with slower rates of stimulation, the modern evoked potential machines acquire two
evoked potential waveform morphology is better, repetitions simultaneously; that is, alternate stimuli
latency shorter, and amplitude higher. Faster rates are used to create two different averages.
stress neural transmission, particularly through
synapses, resulting in longer latencies and lower Recording Methodology
amplitudes. However, the slower the rate, the lon-
The methods used for recording evoked poten-
ger it takes to acquire the data which may compro-
tials vary to some degree depending on the evoked
mise patient cooperation. The optimal stimulation
potential type. However, there are many aspects
rate must be a compromise between evoked poten-
of recording that are similar, such as recording
tial morphology and speed of acquisition. In some
electrodes, averaging, and postacquisition signal
laboratories, evoked potentials will be obtained
processing.
with a “fast” rate initially. If the evoked potential is
normal at this rate, the test is complete. However,
if it is abnormal, the evoked potential will be Electrodes
repeated at a slower rate. Occasionally, an evoked Different types of electrodes can be used to record
potential is normal with the slow rate, but abnor- evoked potentials. Any electrode used should
mal with fast rate. How such a test is interpreted have certain properties that will allow reliable data
is controversial. Some would interpret an evoked acquisition.
potential as normal if it is normal at any stimulation
rate. On the other hand, some would interpret an Properties
evoked potential abnormal if it is abnormal at any Recording electrodes must be able to conduct neu-
stimulation rate. The author’s practice is the former ral signals without distortion. They must be made
(Figures 1.3A and B). of material that resists polarization and does not
interact with skin or other human tissue with which
Repetitions it makes contact. Electrodes coated with gold, plat-
The number of repetitions refers to the number inum, or silver coated with silver chloride are most
of evoked responses that must be averaged to often used. Periodically the electrode’s resistance
Age: 49 years
Sex: Male
Ear inserts (add 0.9 ms): Yes
Side: Right
Click polarity: Rarefaction
Stimulation intensity: 70 dBnHL
Masking intensity: 40 dBnHL
Absolute Latencies
(+↑) I N/A N/A (I–In)
III 4.88
V 6.84 0.27 (V–Vn)
Vc 6.86
N/A (V/I ratio)
Interpeak Latencies
Waveforms Latency (ms)
I–Vc N/A
(A) I–III N/A
Stimulation rate: 51.1/s III–Vc 1.98
Scale: Amplitude = 0.05 mV/div; Latency = 1 ms/div
Absolute Latencies
I 2.40 0.07 (I–In)
(+↑)
III 4.66
V 6.64 0.35 (V–Vn)
Vc 6.72
5.0 (V/I ratio)
Interpeak Latencies
I–Vc 4.32
(B) I–III 2.26
III–Vc 2.06
Discussion: Faster stimulation rates can cause degradation in waveform morphology, loss of amplitude, and prolongation of
latency. A BAEP obtained with a fast stimulation rate of 51.1/s is shown (A). Notice that the wave I is difficult to identify, despite
the sensitivity being set at 0.05 mV/div. In the same patient, when the rate is slowed to 11.1/s (B), the wave I is clearly seen despite
the sensitivity being lower (0.1 mV/div). Additionally, the wave V latency is shorter with the slower rate. It is the author’s practice to
consider a study normal if it is normal at any stimulation rate.
BAEP, brainstem auditory evoked potential.
Figure 1.3 (A and B) Effect of stimulation rate on BAEP waveforms.

Age: 43 years (B)

Sex: Male Number of repetitions: 10
Filters: A = 1–100 Hz; B = 10–100 Hz; C = 1–30 Hz
Preauricular−preauricular distance: 38 cm
Scale: Amplitude = 2 mV/div; Latency = 20 ms/div
Eye: Right
Visual angle: 30′
Visual acuity: 20/20
Corrective lenses: Yes
(A) (C)
Number of repetitions: 1 Number of repetitions: 50
Figure 1.4 (A–E) Effect of number of repetitions on VEP waveforms. (continued )

(D) (E)
Number of repetitions: 100 Number of repetitions: 200
Discussion: This PRVEP shows the utility of increasing repetitions. In Figure A, only one repetition is shown. Figures B, C, D,
and E have 10, 50, 100, and 200 repetitions, respectively. With 1 and 10 repetitions, the P100 waveform cannot be identified. With
50 repetitions, the P100 waveform can be identified, but has poor morphology and reproducibility. As the number of repetitions is
increased to 100, a much more clearly defined P100 waveform is seen, and this changes only slightly with 200 repetitions. When
the evoked potential waveform ceases to change with additional repetitions, averaging is adequate.
PRVEP, pattern reversal visual evoked potential; VEP, visual evoked potential.
Figure 1.4 (A–E) (continued )
must be checked. This is the opposition to direct (EEG) cup electrodes are the type of surface
current flow and is measured with an ohmmeter. electrodes used most often. These electrodes are 5 to
Electrode resistance should be no more than a few 10 mm in diameter. Before electrodes are applied,
ohms, and if it is very high, a breach in the integ- the skin is prepared with a mild abrasive to remove
rity of the electrode should be suspected. Electrode oils and dead skin. The electrode is attached with
impedance is the opposition to AC flow. Rather collodion or paste depending on the indication of
than measuring the integrity of the electrode, the study. Collodion application takes longer and
impedance evaluates the connection of the electrode should be performed in a room with adequate ven-
to skin or soft tissue and is measured by an imped- tilation due to the smell of the chemical. It stays
ance meter. Electrode impedance should be below attached to the patient for longer period of time.
5,000 Ω (4). Higher impedances degrade the ability After the electrode has been attached, the cup is
of the amplifier to average low amplitude signals. filled with conductive jelly to ensure adequate
contact between the electrode and skin. Acetone is
Types used to dissolve the collodion and remove the elec-
Surface and needle electrodes can be used to record trode. Conductive electrode paste can also be used
evoked potential signals. Electroencephalogram to attach electrodes to skin. This method is much
quicker, however it is not as secure as collodion and Receptacles in the input board also limit the amount
electrodes can be dislodged more easily. of current that can flow from the machine to the
Needle electrodes can also be used to record patient, so patients cannot receive electrical shocks
evoked potentials. Because they are more inva- from the machine.
sive than surface electrodes, needle electrodes are
not typically used for outpatient evoked poten- Selector switches
tial studies. They can be applied more quickly Selector switches allow pairing two electrode
than surface electrodes as skin preparation is not inputs into the amplifier. This allows the amplifier
necessary. This is useful when evoked poten- to compare the signal between the two electrodes.
tials are performed in the operating room (OR). It is with selector switches that recording montages
Needle electrodes have high impedance which can are created.
improve signal quality when the amplifier’s input
impedance is also high. Differential amplifier
The basic principle of a differential amplifier is that
Placement it magnifies and displays the difference between
Electrode placement depends on the type of evoked two inputs. Each differential amplifier has two main
potential being recorded. VEP recording electrodes inputs called Input 1 and Input 2; previously these
are placed on the scalp, whereas BAEP electrodes were also called Grid (G) 1 and G2. Because Input 1
are placed on the scalp and mastoids. SEP elec- is often close to the site generating the potential of
trodes are placed on not only the scalp, but also interest, it is also referred to as the Active electrode,
the neck, spine, and limbs. The exact placement whereas Input 2 is referred to as the Reference
is determined by locating anatomical landmarks. electrode.
Electrodes placed close to each other are used for The output of a differential amplifier is the dif-
recording near-field potentials, while electrodes ference between Input 1 (G1, active) and Input 2
placed at a distance are better at demonstrating far- (G2, reference) electrodes. Even though one elec-
field potentials. trode is called “active” and the other “reference,”
both contribute to the ultimate waveform. The dis-
Amplifiers play convention used determines if the waveforms
deflect upward or downward. At times, more than
The amplifier is the backbone of any evoked poten-
one electrode can be linked to create one reference
tial recording device. A few basic properties of
(Input 2, G2) electrode. An example of this would
these amplifiers are presented.
be linking the two ears as a single reference.
Basics Because the differential amplifier displays the
Amplifiers have the task of taking signals that are difference between two inputs, signals that are
of very low amplitude and isolating them from similar in both inputs will be subtracted. This is
background noise. The signals are then amplified known as rejection of the common mode signal. An
to a degree that can be measured and appropriate example of signal that is often rejected because it
calculations are performed. Amplifiers consist of an is in common mode is ECG artifact. Since the ECG
input board, selector switches, differential ampli- artifact is seen in both electrodes, it is subtracted or
fier, and filters. rejected. However, if the ECG artifact is of different
amplitudes in the two electrodes, only some of it
Input board will be rejected, and a low amplitude artifact may
The input board, sometimes also called the jack be seen in the output. The ability of a differential
box or head box, is a small box that is placed near amplifier to reject common mode signal is measured
the patient. Electrodes attached to the patient are by the common mode rejection ratio (CMRR). The
plugged into receptacles in this box. The input sig- CMRR is defined as the “ratio of amplifier output
nals are amplified by the input board before they produced by a signal applied differentially over
are transmitted to the main unit of the evoked the amplifier output produced by the same signal
potential machine. This allows the biologic sig- when applied in common mode” (1). Contemporary
nal to be amplified before any exogenous noise is evoked potential machines should have a CMRR of
introduced as the signal travels through cables. at least 10,000:1 (4).
Averaging are not sine waves. It is not possible to determine

Neurophysiologic signals such as evoked poten- the exact frequency of evoked potential waveforms,
tials that have amplitude that is lower than the and most evoked potentials contain waveforms of
accompanying noise can only be visualized with many different frequencies. Consequently, filter
averaging (5). Averaging involves adding succes- settings are intentionally set with a broad bandpass
sive responses and dividing the sum by the num- so as not to eliminate any waveform of interest.
ber of responses. Since the evoked potential is Filters can also distort the shape of waveforms caus-
time locked to the stimulus and always occurs at ing them to appear slightly earlier or later. This is
the same latency while noise is random, averag- called phase shift; progressively greater LFF cause a
ing allows reduction of noise and emergence of the phase advance (waveforms appear earlier) and pro-
evoked potential signal. Averaging involves digi- gressively lower HFF cause a phase lag (waveforms
tization of the analog signal in an analog-to-digital appear later) (Figures 1.5A–C).
(A/D) converter. The digital signal is averaged and Another special type of filter affects only a very
then converted back to analog form for display. narrow frequency band. This is known as a notch
The number of responses that need to be averaged filter. Most often this is a 60 Hz filter in the US as
to resolve an evoked potential waveform depends the electrical AC is at this frequency. In other parts
on the amplitude of the waveform of interest versus of the world, a 50-Hz band filter is available.
the amplitude of noise. This relationship is called Digital filters are used after the signal has been
the signal to noise ratio. As discussed earlier, the digitized. Several types of digital filters are available
lower the signal to noise ratio, the more responses and one that is used occasionally in clinical evoked
that need to be averaged. If the signal to noise ratio potentials is called “smoothing” (1). Smoothing
is increased, either by increasing the amplitude of uses a computer algorithm to average three to five
the signal or reducing the noise, fewer responses consecutive data points to help eliminate noise and
need to be averaged. Averaging improves the signal “smooth” out waveforms. Smoothing can be used
to noise ratio by a factor equal to the square root of more than once on a waveform. However, it should
the number of repetitions averaged. It is a common be used sparingly as it can affect waveform mor-
misperception that averaging increases the ampli- phology (Figures 1.6A–C). Digital filtering, how-
tude of the signal, which is incorrect. It reduces the ever, does not cause phase shifts.
amplitude of the noise.
Artifact rejection
Filters Artifact rejection is another technique used to help
Evoked potentials contain waveforms of a lim- resolve low-amplitude signals such as evoked
ited frequency range. Filters are used to eliminate potentials. Amplifier input voltage is set so that
unwanted waveforms that are below and above the signals of excessively high voltage that are unlikely
frequencies of interest. By eliminating low and high to be of interest are automatically rejected and
frequency activity that is not of interest, fewer repe- not included in the average. This prevents the
titions are needed to resolve low amplitude evoked average from being contaminated from these high-
potential waveforms. Analog filters are applied to amplitude signals. The limits of artifact rejection
the signal before it is digitized. should be set so that signals of interest are not
Analog filters are divided into low- and high- affected. Artifact rejection allows a reduction of
frequency filters (LFFs and HFFs). LFFs reduce the number of responses that need to be averaged.
the amplitude of low frequency activity, and HFFs However, if the artifact rejection level is set too low,
similarly affect high frequencies. Filters are fur- too many responses will be rejected, increasing the
ther named based on the frequency of sine waves time and responses needed to average.
whose amplitude they reduce by a fixed percent-
age, usually 30%. Thus, a 1 Hz LFF will reduce the
amplitude of 1 Hz activity by 30%, and amplitude
Interpretation
of activity slower than 1 Hz will be reduced more Interpretation of evoked potentials involves a
than 30%. The activity range between the LFF and detailed evaluation of the waveforms. This includes
HFF is referred to as the bandpass. Properties of fil- determining the presence, morphology, latency,
ters are based on sine waves, and biological signals amplitude, and several other features of waveforms.
Age: 43 years
Sex: Male
Filters: A = 1–100 Hz; B = 10–100 Hz; C = 1–30 Hz
Eye: Right
Number of repetitions: 200
Visual angle: 30′
Absolute Latency
Derivation
MO–MF N75 (ms) P100 (ms) N145 (ms)
A 84.1 112.0 158.3
B 81.3 95.6 128.0
C 88.7 114.5 160.1
(A)
Amplitude
Derivation N75–P100 P100–N145
MO–MF (mV) (mV) Mean (mV)
A 4.20 8.21 6.21
B 2.20 5.10 3.65
C 5.86 8.34 7.10
(B)
(C)
Discussion: Filters can affect the shape and latency of evoked potential waveforms. In Figure A, a PRVEP with usual filter setting of
1 to 100 Hz is shown. Figure B shows same waveform obtained with the LFF increased to 10 Hz (10–100 Hz). Notice that this
reduces the amplitude and shortens the latency of the P100 waveform. Figure C shows the same waveform with the HFF reduced
to 30 Hz (1–30 Hz). Because the P100 waveform is a low-frequency waveform, this does not reduce the amplitude but increases
the latency. Thus, increasing the LFF causes a phase advance and decreasing the HFF causes a phase delay of the evoked potential
waveforms.
HFF, high-frequency filter; LFF, low-frequency filter; PRVEP, pattern reversal visual evoked potential; VEP, visual evoked potential.
Figure 1.5 (A–C) Effect of filter changes on VEP waveforms.

Age: 32 years
Sex: Female
Scale: Amplitude = 0.1 mV/div; Latency = 1.5 ms/div
Side: Left
Click polarity: Alternating
Absolute Latencies
I 2.43 0.19 (I–In)
III 4.77
V 6.69 0.44 (V–Vn)
Vc 6.75
2.3 (V/I ratio)
Interpeak Latencies
I–Vc 4.32
I–III 2.34
III–Vc 1.98
(A)
Absolute Latencies
I 2.43 0.19 (I–In)
III 4.77
V 6.69 0.42 (V–Vn)
Vc 6.69
2.2 (V/I ratio)
Interpeak Latencies
I–Vc 4.26
I–III 2.34
III–Vc 1.92
(B)
Figure 1.6 (A–C) Effect of smoothing filter on BAEP waveforms. (continued )

Absolute Latencies
I 2.28 0.20 (I–In)
III 4.77
V 6.66 0.39 (V–Vn)
Vc 6.69
2.0 (V/I ratio)
Interpeak Latencies
I–Vc 4.41
I–III 2.49
III–Vc 1.92
(C)
Discussion: Excessive smoothing can distort evoked potential waveforms. Figure A shows a BAEP waveform which has not
undergone smoothing. Figure B is the same waveform with smoothing done once. Figure C shows the waveform with smoothing
performed 10 times. Notice the change in absolute and interpeak latencies. Smoothing does make the waveforms look “cleaner,”
but that is at the expense of accuracy.
Figure 1.6 (A-C) (continued )
The patient’s evoked potential is compared to evaluating the waveforms for each type of evoked
normative data to determine if an abnormality is potential, which will be discussed in the respec-
present. tive chapters.
A few basic principles apply to identification of
all waveforms. Responses that are difficult to ascer-
Peak Identification tain because of low amplitude may be improved
The various evoked potential types used in by increasing the stimulation intensity or duration.
clinical practice have a typical series of peaks This is particularly true for SEP. On the other hand,
that occur at expected latencies. This allows for occasionally the stimulation intensity can be too
nomenclature of the peaks, as discussed earlier. high and extra waveforms are seen that make iden-
Peaks may be labeled according to their polar- tification of waveforms of interest difficult. This
ity and expected latency (i.e., N20 waveform is a happens when very high stimulation intensities
negative peak occurring at approximately 20 ms) are used for obtaining BAEP. This results in extra
or their number in a series (i.e., wave III is the waveforms, which makes identifying waves I to V
third BAEP waveform). Peak identification is easy difficult. Reducing the stimulation intensity can
when a normal response is present. However, eliminate the extra waveforms, making interpreta-
in cases of abnormalities, peak latencies may be tion easier.
highly variable, there may be excessive noise At times instead of a single peak, two smaller
making evoked potential waveforms difficult to peaks may appear. This is known as a bi-fed wave-
discern, and morphology of the waveform may be form. When a bi-fed waveform is seen, it may be
distorted. There are specific recommendations for suggestive of pathology or may be an artifact of
the recording method. An example of the latter Absolute latency

can be seen when the P100 waveform is obtained The absolute latency, also referred to as the peak
with a midoccipital to mid frontal (MO-MF) der- latency, is measured from the start of the stimu-
ivation. Contributions from the N105 waveform lus to peak of the waveform of interest. The onset
recorded from the MF electrode and the P100 latency is the time from stimulus onset to the start
waveform recorded from the MO electrode are of the waveform. This latter latency measure is not
added together to create the bi-fed waveform. routinely used as the onset of a waveform is much
This bi-fed waveform, also called W-shaped harder to determine than the peak (Figure 1.8).
waveform, can be resolved by changing the ref- The absolute latency of a waveform is the time
erence electrode to Ai-Ac (linked ears). When the an impulse takes to travel from the point of stimu-
bi-fed waveform cannot be resolved with mon- lation to the generator of the waveform of interest.
tage changes, a decision needs to be made how Because the stimulus is applied to end organ, abso-
latency and amplitude measurements will be lute latency measures conduction in the peripheral
obtained. Latency can be measured to the first and central nervous system (or in the case of VEP,
peak, to the second peak, or to the average of the the retina). A lesion of the peripheral nervous sys-
two peaks. Similarly, amplitude can be measured tem, such as demyelination of a peripheral nerve,
of the first or second peak. Regardless of which can cause the absolute latency to be prolonged.
method is used, it should be consistent with labo-
ratory protocol. Interpeak latency
Excessive noise can also make peak identification Interpeak latency (IPL) is the latency between two
difficult. An attempt should be made to reduce the peaks of the same evoked potential. It represents
noise by having the patient relax or sleep. The noise the time an impulse takes to travel between the
can be measured by averaging without stimulating. presumed generators of the waveforms. IPLs are
If a flat baseline is not obtained with this method useful in eliminating the peripheral nervous sys-
after a standard number of repetitions, the evoked tem contribution to the latency. If the IPL between
potentials may be difficult to resolve. The number two waveforms that are both in the central nervous
of repetitions obtained can be increased; however system is determined, only central nervous system
the signal to noise ratio increases by a factor of conduction is measured. This is particularly help-
the square root of the number of repetitions. This ful in SEP and BAEP. In SEP, the IPL between the
makes increasingly higher number of repetitions of upper and lower spinal cord (for upper and lower
diminishing utility. extremity SEP) and brain can determine central con-
Modern evoked potential equipment automati- duction time (CCT). IPL measurement is preferred
cally tags waveforms based on a preprogrammed to absolute latency measurements as they are less
algorithm. This works reasonably well when the variable and not influenced by peripheral nervous
response is normal. However, when the response system pathology (Figure 1.8).
is abnormal, the automated peak labeling may
be incorrect. Similarly, evoked potential technol- Amplitude
ogists label waveforms they think correspond to
As with latency measurements, amplitude of wave-
waveforms of interest. Experienced technologists
forms can be measured in several ways. Amplitude
are often correct, but even they may be inaccurate
abnormalities are not as reliable as latency abnor-
with more complex studies (Figure 1.7). It behooves
malities. Evoked potentials are seldom interpreted
the interpreter to double check that the waveforms
as abnormal if only amplitude abnormalities exist.
are appropriately tagged.
Peak to peak amplitude
Peak to peak amplitude is the most commonly
Latency used amplitude measurement. It is a measure of
Latency measurement is the most important the voltage difference between successive peaks of
measure of evoked potential waveforms. Two main opposite polarity. The amplitude of the ascending
types of latencies can be determined. limb, descending limb, or the average of both is
Age: 18 years
Sex: Male
(−)↑
Side: Right
Absolute Latencies
LP N/A
P21 (31) 32.0
N24 (34) 37.1 Interpeak Latencies
P27 (37) 43.9 0.18 (P27–N35) Waveforms Latency (ms)
N35 (45) 50.4 LP–P27 N/A
Discussion: Before looking at waveform latencies, the interpreting physician must first evaluate the waveforms and confirm that
they have been properly tagged. This is a peroneal nerve SEP, and absolute latencies of peroneal SEP are 10 ms less than tibial nerve
SEP. The waveforms are correspondingly called P21, N24, and P27. In some laboratories, for simplicity, the nomenclature for tibial
nerve SEP is retained for peroneal nerve SEP, that is, the previously noted waveforms are called P31 waveform, N34 waveform, and
P37 waveform. A PF waveform is not recorded as that is the site of stimulation. All waveforms have been incorrectly tagged. The
waveform tagged P21 is actually the P27. Notice that this waveform is seen not only in the CPz–C5S and CP4–C5S derivations, it
is also seen in the CPz–FPz derivation. It is not seen in the FPz–C5S and CP3–C5S derivations. If this waveform was a subcortical
P21, it should not have been seen in the CPz–FPz derivation (it would have been in “common mode” in both electrodes), and it
should have been seen in the FPz–C5S and CP3–C5S derivations. The waveform marked with the arrow is the P21, and it follows
the properties of the subcortical response discussed earlier. If this was not noticed and the absolute latencies relied upon as pro-
vided, the interpretation for this study would have been that it was abnormal due to prolongation of the P27 waveform. In fact, the
interpretation should be that it is a normal study since the latency of the P27 waveform is actually 32.0 ms (upper limit of normal
for P27 waveform is 37.0 ms).
Figure 1.7 Automated tagging of peroneal nerve SEP waveforms.

Age: 30 years
Sex: Male
Stimulation rate: 5.7/s (−↑)
Side: Left
Median nerve conduction velocity (median
nerve to EP waveform): 52 m/s
Absolute Latencies
P13 12.9 EP–N20 8.94
N18 16.1 1.78 (P13–N18) EP–P13 3.60
N20 18.2 1.52 (N20–P22) P13–N20 5.34
Discussion: Waveform latencies can be determined either to their onset or peak. Peak latencies are used more often in evoked
potentials as onset may be difficult to identify. In this example, the N20 waveform onset latency be either where the light or dark
arrow is placed. The peak latency is where the N20 waveform is tagged. More important than peak (or absolute) latencies are IPL.
IPLs are more accurate than absolute latencies in determining central conduction.
IPL, interpeak latency; SEP, somatosensory evoked potential.
Figure 1.8 Latency measurements of SEP waveforms.
determined (Figure 1.9). Whichever method is used, period of the evoked potential (Figure 1.9). This
it is important to be consistent within a laboratory measurement may be difficult to determine as the
and the normative data must have been obtained baseline may not be horizontal.
with the same method.
Amplitude ratio
Baseline to peak amplitude The ratio of the peak to peak amplitude of two wave-
The baseline to peak amplitude is the voltage from forms of the same evoked potential can be determined
the baseline to the peak of the waveform of inter- and is thought to be a more sensitive amplitude mea-
est. The baseline is the immediate post stimulus surement than peak to peak amplitude of a single
Age: 39 years
Sex: Female
Stimulation rate: 3.9/s ↓(+)
Filters: 1–100 Hz
Eye: Right
Visual angle: 30′
Corrective lenses: No
Absolute Latency Amplitude

Derivation N75 (ms) P100 (ms) N105 (ms) N145 (ms) N75–P100 P100–N145
MO–MF 80.0 102.0 128.0 Derivation (mV) (mV) Mean (mV)
MO–AU 80.0 101.0 126.0 MO–MF 12.2 16.1 14.2
Oz–AU 80.0 101.0 126.0 MO–AU 10.5 14.6 12.6
MF–AU 106.0 Oz–AU 9.68 13.0 11.3
LO–MF 82.5 101.0 125.0 LO–MF 4.95 8.55 6.8
RO–MF 82.5 103.0 127.0 RO–MF 10.7 14.2 12.5
P100 Amplitude Ratio

Location Ratio
OS–OD 1.2
LO–RO 1.8
Discussion: Amplitude of a waveform can be measured from baseline to peak or from peak to peak. The latter is used more often in
evoked potentials since identifying peaks is more reliable than identifying the baseline. In the example shown, the baseline to peak
(narrow line) P100 waveform amplitude is not as reliable as the N75–P100 waveform peak to peak amplitude (broad line). Peak to
peak amplitudes can be measured for the descending or ascending limb of the peak, or it can be an average of the two. The above
example displays amplitudes of the descending and ascending limb, but it is the average that is used in the author’s laboratory.
Amplitude ratios are used in PRVEP to compare parasagittal P100 waveforms (those obtained from LO–MF and RO–MF derivations)
and the P100 waveforms obtained after left and right eye stimulation. In the above example, the parasagittal waveform amplitude
ratio is within normal limits, as is left and right eye P100 waveform amplitude ratio (PRVEP to left eye stimulation not shown).
Figure 1.9 Amplitude measurements of VEP waveforms.

waveform. This is commonly used in BAEPs where Interindividual variability

wave V/I amplitude ratio is measured. Amplitude Evoked potentials can vary considerably between
ratios can also be used to compare the same wave- individuals. Sex and body or head size, which are
form obtained after stimulation of the left and right related, can have a significant effect on evoked
side or recorded from different locations. This is potential waveform latency. Age can also have an
done with PRVEP where the P100 waveform can effect on evoked potentials. The biggest change
be compared after left and right sided stimula- is from infancy to maturity, but changes can also
tion. The P100 waveform recorded from parasag- occur with old age. Anatomical variability of neu-
ittal electrodes can also be compared (Figure 1.9). ral structures may also cause morphological differ-
ences in the evoked potentials. An example of this
Other Measurements is the P37 cortical waveform obtained after tibial
It is also important to review other measures that nerve stimulation. Because the leg somatosensory
are assessed in the evoked potentials laboratory cortex is near the vertex, the P37 waveform is typi-
which are not directly related to the evoked poten- cally best recorded contralateral to the side of wave-
tial. These include measures such as visual acu- form generation (ipsilateral to side of stimulation
ity for PRVEP, hearing thresholds for BAEP, and since dorsal column pathway fibers decussate in
peripheral nerve conduction velocities for SEP. the medulla). However, occasionally the P37 wave-
These measures help in interpreting the evoked form is better recorded over the vertex or ipsilateral
potential since they can provide clues as to whether to site of waveform generation.
pathology exists in the end organ being stimulated
or the peripheral nervous system. Impaired visual
Normal Evoked Potentials
acuity can lead to P100 waveform latency prolon-
gation, hearing loss can cause prolonged absolute When interpreting an evoked potential study, var-
latency of wave I of the BAEP, and slowed conduc- ious aspects of the data are compared to norma-
tion velocity can affect SEP waveforms. tive values. Details of what to evaluate for each
modality will be presented in the respective chap-
Variability ters. Normative data are obtained by performing
Evoked potentials obtained with similar methodol- evoked potentials on a group of individuals who
ogy may appear different within the same patient do not have neurologic disease. It is important to
when recorded at different times, and between include both sexes and the spectrum of ages that
patients the variability may be even more. There are will be evaluated. At the extremes of ages, it is use-
many reasons for this variability. ful to have separate norms. The data are evaluated
to determine if it is normally distributed. If it is,
Intraindividual variability the mean values for latency and amplitude of the
Many factors within a patient may affect an evoked waveforms of interest have to be calculated. The
potential. Foremost is the attention level. Cortical upper limit of normal is then defined as the mean
responses are of higher amplitude when the patient is value plus 2.5 or 3 standard deviations above the
awake than if they are drowsy or asleep. If the patient mean. At least 20 individuals should be tested to
is awake when the study is obtained on one side and determine the mean latency and amplitude values.
falls asleep when the other side is acquired, there If the data are not normally distributed, the per-
may be considerable variability of the evoked poten- centile method is used, which indicates the prob-
tial (6). This is seen most remarkably in tibial nerve ability that the evoked potential from the control
SEP (Figures 1.10A and B). If the background activity group will be similar to the acquired evoked poten-
or noise increases, the evoked potential may become tial. In the latter method, about 100 subjects must
more difficult to visualize as the signal to noise ratio be studied to obtain normative data. Ideally, all
is lower. A drop in blood pressure may make the evoked potential laboratories should obtain their
amplitude of the evoked potential lower, which also own normative data by using methodology that
reduces the signal to noise ratio. This is seldom seen is similar to that which will be used for perform-
in an outpatient setting, but may occur when evoked ing clinical studies. However, obtaining normative
potentials are performed in an intensive care unit data is difficult, especially for small laboratories.
and OR. When there are lesions of the cerebral cortex, In these instances, it is acceptable to use published
evoked potentials may be more variable. normative data if clinical studies are done using
Age: 43 years
Sex: Male
(−↑)
(A)
Side: Left
Tibial nerve conduction velocity (tibial nerve to T12S): 41 m/s
Absolute Latencies
PF 9.30
LP 23.9
P31 32.9
N34 35.9
P37 40.3 1.52 (P37–N45)
Interpeak Latencies
PF–P37 31.0
LP–P37 16.4
Figure 1.10 (A and B) Effect of attention on tibial nerve SEP waveforms. (continued )
(B)
Side: Right
(−↑) Number of repetitions: 3,000
Tibial nerve conduction velocity (tibial nerve to T12S):
41 m/s
Absolute Latencies
PF 9.10
LP 24.1
P31 32.9
N34 34.9
P37 40.8 2.75 (P37–N45)
Interpeak Latencies
PF–P37 31.7
LP–P37 16.7
Discussion: State change of the patient, such as waking up or falling asleep, can affect evoked potential waveforms, particularly
tibial nerve SEP. In this example, right tibial nerve SEP (B) was performed first when the patient was fully awake. When left tibial
nerve SEP stimulation was started (A), the patient was becoming drowsy. Thus, the left tibial nerve SEP was obtained when the
patient was not as alert as he was for the right tibial nerve SEP. Notice that the cortical P37 waveform is of higher amplitude with
right sided stimulation, when the patient was fully awake. The absolute and interpeak latencies are normal, but the P37 waveform
amplitude is asymmetric. This finding alone does not make this study abnormal.
Figure 1.10 (A and B) (continued )

the same methodology as the reference laboratory. the neural pathways from the point of stimulation to
Additionally, a few studies on normal individuals the neural generator of the waveform. The slowing
should be done using the reference laboratory’s is often caused by a lesion along the pathway. When
methodology to confirm that normal results are recording evoked potentials, the slowing of inter-
obtained. est is that which affects the central nervous system.
In clinical evoked potentials interpretation, the However, when absolute latencies are evaluated,
most important variable to analyze is the latency. peripheral nerve conduction is also measured. For
For each type of evoked potential, a series of IPL and example, a delayed N20 waveform absolute latency
absolute latencies should be compared to normative of the median nerve SEP may be due to demyelinating
data. In BAEP and SEP, IPLs are more important than peripheral neuropathy or a brainstem lesion affecting
absolute latencies, but for VEP the absolute latencies the medial lemniscus (Figure 1.11). Similarly, prolon-
are evaluated. Even if latencies are within normal gation of the P100 waveform of the PRVEP may be
limits, they should be compared from side to side due to ocular problems or an optic chiasm lesion.
to determine if an asymmetry exists. Asymmetric IPL assessments are much more accurate for eval-
latencies may also suggest an abnormality. uating the central nervous system. These measures
The amplitude of waveforms should also be deter- allow assessment of neural pathways between gen-
mined. As with latency, there are some waveforms erators of two waveforms. For example, for median
for each type of evoked potential for which ampli- nerve SEPs, the waves EP–N20 (EP to N20 wave-
tude is routinely determined. Most laboratories do form) IPL assesses neural pathways between the
not use peak to peak or baseline to peak amplitude brachial plexus and the contralateral somatosensory
to determine normalcy of waveform. When ampli- cortex and the waves P14 to N20 IPL provides an
tude is used, it is often the amplitude ratio that is assessment of the fibers between the nucleus cunea-
evaluated. The amplitude ratio may be of different tus and the contralateral somatosensory cortex.
waveforms of the same evoked potential (wave V/I Diseases of the peripheral nervous system are much
ratio of the BAEP), the same waveform obtained less likely to affect IPL (Figure 1.11). Similarly, BAEP
after contralateral stimulation (ratio of P100 wave- waves I to V IPL allows assessment of the auditory
form amplitude obtained after left and right eye pathway after removing the contributions of the
stimulation), or the same waveform obtained with peripheral hearing apparatus. It is not unusual to see
different recording derivations (ratio of the left a prolonged wave V absolute latency with a normal
and right parasagittal P100 waveform amplitude). wave I to V IPL. This suggests a lesion at or before
Morphology and topography of the waveform the cochlea and not of the central nervous system.
are generally not used in determining whether an The VEP is unique in that only the P100 waveform
evoked potential study is normal. As noted earlier, is measured, so IPL cannot be determined.
both of these are highly variable between individu- Absolute latencies and IPLs are compared to nor-
als. Often changing montages and recording from mative data to see if they exceed the upper limit
different electrodes to compensate for variable of normal latency measures. Even if these mea-
topography will result in more “normal” appearing surements do not exceed the upper limits of nor-
waveforms. mal, it is useful to compare latencies from side to
side. If greater than maximal allowable asymmetry
Abnormal Evoked Potentials is noted, the side with slower conduction may be
As noted earlier, the most significant evoked poten- abnormal (Figures 1.12A and B).
tial abnormality is latency prolongation. Amplitude
Amplitude reduction
and shape of the waveform should also be evalu-
ated, but these are seldom used as the sole criteria A reduction in the amplitude of evoked potential
for abnormality. The abnormalities seen in evoked peaks is most often accompanied by latency pro-
potentials can be used to localize the site of the longation. The lower amplitude can be explained
potential lesion. by temporal dispersion if conduction is slowed.
However, at times amplitude reduction is noted
Latency prolongation without latency prolongation (Figures 1.13A
Prolongation of the latency of a waveform implies and B). Whereas this may imply axonal loss in
that there is slowing of the conduction velocity of the pathway being tested, it may also be due to
Age: 60 years
Sex: Male
Side: Left
Median nerve conduction velocity (median
nerve to EP waveform): 39 m/s
Absolute Latencies
EP 12.8
N13 16.3
P13 17.3
N18 20.2 1.09 (P13–N18)
N20 22.7 1.07 (N20–P22)
Interpeak Latencies
EP–N20 9.90
EP–P13 4.50
P13–N20 5.40
Discussion: Absolute latencies are a measure of both peripheral and central conduction velocity. IPLs are a more accurate method
of evaluating central nervous system conduction. In this example, absolute latencies of all waveforms, including the EP and N20
waveforms, are prolonged. However, when the waves EP–N20 IPL and other IPLs are evaluated, it is evident that the absolute
latency prolongation is due to a peripheral nervous system process. This is further supported by what appears to be a slowing of
median nerve conduction velocity. It is cautioned that supramaximal stimulation is not applied to the peripheral nerve in the evoked
potentials laboratory, and so the conduction velocity should be interpreted with the understanding that all nerve fibers may not
have been stimulated.
Figure 1.11 Latency prolongations of median nerve SEP waveforms.

Age: 41 years
Sex: Female
Scale: Amplitude = 0.5 mV/div; Latency = 8 ms/div ↑(−)
(A)
Side: Left
Absolute Latencies
PF 9.9
LP 24.2
P31 29.9
N34 34.0
P37 38.6 1.45 (P37–N45)
Interpeak Latencies
PF–P37 28.7
LP–P37 14.4
Figure 1.12 (A and B) Latency asymmetries of tibial nerve SEP waveforms. (continued )
(B)
Side: Right
↑(−) Number of repetitions: 1,500
Tibial nerve conduction velocity (tibial
nerve to T12S): 43 m/s
Absolute Latencies
PF 9.4
LP 24.4
P31 31.2
N34 36.0
P37 43.0 1.49 (P37–N45)
Interpeak Latencies
PF–P37 33.6
LP–P37 18.6
Discussion: Absolute and interpeak latencies are compared to normative data to determine if pathology exists. Additionally, laten-
cies should be compared from side to side in the same patient. At times, latencies can be within normal limits, but a significant
difference is present between the two sides. If this latency difference is beyond the upper limit of normal, even if the latencies them-
selves are within normal limits, the study is considered abnormal. In this example, left (A) and right (B) tibial nerve SEP are shown.
The waves LP–P37 IPL are asymmetric (14.4 ms vs. 18.6 ms), with a shorter latency after left-sided stimulation. Both IPLs, however,
are within normal limits. Because of this asymmetry, this study is considered abnormal.

Age: 56 years
Sex: Female
(A)
Side: Left
Median nerve conduction velocity (median nerve to EP waveform): 60 m/s
Absolute Latencies
EP 9.72
N13 14.0
P13 13.4
N18 17.4 1.04 (P13–N18)
N20 20.6 0.57 (N20–P22)
Interpeak Latencies
EP–N20 10.9
EP–P13 3.72
P13–N20 7.14
Figure 1.13 (A and B) Low amplitudes of median nerve SEP waveforms. (continued )
(B)
Side: Right
Median nerve conduction velocity (median nerve to EP
waveform):
Absolute Latencies
EP 10.0
N13 13.6
P13 15.1
N18 17.7 1.33 (P13–N18)
N20 20.9 1.41 (N20–P22)
Interpeak Latencies
EP–N20 10.9
EP–P13 5.04
P13–N20 5.88
Discussion: Low amplitude of evoked potential waveforms is seldom considered an abnormal finding unless latencies are also pro-
longed. Many patient-related factors (i.e., sleep), technical factors, and pathology can cause an amplitude reduction. In this median
nerve SEP example, the N20 (and EP) waveform has a lower amplitude after stimulation of the left median nerve (A) compared to
the right median nerve (B). Despite the amplitude asymmetry, this is considered a normal study. The latency of the N20 waveform
is within normal limits.

Age: 49 years
Sex: Male
Stimulation rate: 3.9/s (+↓)
Filters: 1–100 Hz
(A)
Eye: Right
Visual angle: 30′

MO–AU 78.0 118.0 155.0 MO–MF 5.97 7.26 6.6
Oz–AU 78.0 118.0 155.0 MO–AU 4.11 6.61 5.36
MF–AU 136 Oz–AU 3.61 5.72 4.67
LO–MF 78.0 129.0 160.0 LO–MF 6.91 8.23 7.60
RO–MF 78.0 131.0 161.0 RO–MF 7.93 8.39 8.20
P100 Waveform Amplitude Ratio

Location Ratio
OS–OD 1.30
LO–RO 1.08
Figure 1.14 (A and B) W-shaped VEP waveform. (continued )

(B)
Eye: Right
Visual angle: 60′

MO–AU 73.0 109.0 139.0 MO–MF 7.68 5.62 6.70
Oz–AU 73.0 110.0 139.0 MO–AU 6.46 4.48 5.47
MF–AU 111.0 Oz–AU 4.95 2.60 3.78
LO–MF 73.5 117.0 144.0 LO–MF 5.90 3.95 5.00
RO–MF 71.0 112.0 146.0 RO–MF 5.35 1.21 3.30

Location Ratio
OS–OD 1.14
LO–RO 1.51
Discussion: The P100 waveform can have an abnormal morphology, and the most common such abnormality is a bi-fed peak,
called a W-shaped P100 waveform. This may occur because of technical reasons, patient physiology, or visual pathway problems.
In example (A), a W-shaped P100 waveform is noted after right eye stimulation. This patient also had reduced visual acuity (20/50).
The PRVEP was repeated with 60′ check size (B). With the larger check size, the P100 waveform morphology improved, and there
is no longer a W-shaped waveform. In this case, the W-shaped waveform occurred due to reduced visual acuity.

technical reasons or evoked potential variability. between the superior olivary complex and inferior
If reduced amplitude is the only significant finding colliculus (Figure 1.16). Different evoked potentials
in an evoked potential study, it is seldom desig- can be used jointly for localization determination as
nated as abnormal. Most evoked potential labora- well. An example of this would be if a median SEP
tories do not have normative data for amplitude. is normal but the tibial nerve SEP shows LP–P37
Amplitude ratios are used more often than abso- IPL prolongation. The most likely site of abnormal-
lute amplitude values. They are less susceptible ity would be above the cauda equina and below the
to technical issues and normal evoked potential mid cervical level.
variability.
Abnormal waveforms Report
The shape of the waveform is not used in isolation
As with any other test, the report of the evoked
as a determinant of abnormality. This can be very
potential study is a crucial part of the entire test.
variable depending on not only technical factors
A poorly written or inaccurate report calls into
but also patient factors, such as alertness and ana-
question not only the technical quality of the study
tomical variations. One type of abnormal wave-
but also the validity of interpretation. Every report
form seen commonly is a bi-fed peak. This is most
should start with a “History” section. This should
commonly seen with a PRVEP P100 waveform,
briefly describe the symptoms relevant to the
where it is referred to as a “W-shaped” waveform
evoked potential study being performed. The ques-
(Figures 1.14A and B). Even this is frequently due
tion being asked by the referring physician should
to technical reasons because of how it is recorded
be noted. A list of medications, especially those
(as discussed earlier), but sometimes it may be due
affecting the nervous system, should be listed. It
to pathology.
is worthwhile for the interpreting physician to be
Absence of waveforms aware of the reimbursable indications for evoked
Complete loss of all waveforms of an evoked potential studies in their area.
potential may be suggestive of severe pathology The second paragraph of the report should
in the pathway being tested. Before this deter- be the “Report” section. In this section, a brief
mination is made, it is important to confirm description of the technique used to obtain the
that technical problems did not cause this find- data is described. Common items mentioned in
ing. Rechecking all components of the evoked this section include which structures were stim-
potential machine, including the amplifiers and ulated, whether unilateral stimulation was used
electrodes is very important. If all waveforms and the stimulation rate. Presence of reproducible
except the first are absent, pathology involving waveforms is noted. The significant normal and
the pathway being tested is much more likely abnormal parameters (mostly latencies) are listed.
(Figure 1.15). This is because presence of the first All latency and amplitude measurements do not
waveform documents delivery of adequate stimu- need to be noted. When describing an abnormal-
lation to the neural structures. ity, it is best to note the side that was stimulated
to produce that abnormality. For example, instead
Localization of saying, “A right sided abnormality was noted,”
The site of abnormality can be localized in the evoked it is more descriptive and accurate to say, “An
potentials by determining the pattern of abnormali- abnormality was noted after right sided stimula-
ties. When IPLs are prolonged, the lesion is between tion.” This takes into account that an abnormal-
these two waveforms. IPL prolongations over pro- ity seen with right-sided stimulation may imply
gressively smaller regions further clarify the site a left hemispheric lesion. Toward the end of the
of abnormality. For example, prolongation of the “Report” section, it is important to describe other
waves I–V IPL implies a lesion between the vestib- important measures noted in the evoked potential
ulocochlear nerve and the inferior colliculus. In the laboratory. This includes visual acuity for PRVEP,
same patient, if the wave III–V IPL is prolonged but hearing thresholds for BAEP, and peripheral
the waves I–III IPL is not, the lesion is most likely nerve conduction velocity for SEP.
Age: 5 years
Sex: Male
Scale: Amplitude = 0.05 mV/div; Latency = 1.5 ms/div (+↑)
Side: Left
Absolute Latencies
I 1.61 1.10 (I–In) Interpeak Latencies
III N/A Waveforms Latency (ms)
V N/A N/A (V–Vn) I–Vc N/A
Vc N/A I–III N/A
N/A (V/I ratio) III–Vc N/A
Discussion: Absence of all waveforms of an evoked potential can suggest severe pathology of the pathways involved; however, it
is possible that technical problems caused these findings. When the first waveform is present and all others absent, the interpre-
tation can be much more certain that nervous system pathology is accounting for the findings. In this example, only the cochlear
microphonic is seen. This indicates that the cochlea was successfully stimulated. Absence of all other waves suggests that there is a
severe lesion in the auditory pathway proximal to the cochlea.
Figure 1.15 Absence of BAEP waveforms.
The next paragraph is the “Interpretation” and this section is redundant and should be avoided.
should state whether the test is normal or abnormal. Clinical correlation of the evoked potential find-
If it is abnormal, the abnormalities are listed in order ings as they relate to the available history is the
of significance. Calling a study “borderline” should job of the interpreting physician. Correlating the
be avoided. The final paragraph is the “clinical evoked potentials findings to the entire history of
correlation.” In many ways, this is the most import- the patient must be done by the referring physi-
ant part of the report. In this section, the interpret- cian. This latter “clinical correlation” is implied in
ing physician should determine the localization of all tests and does not need to be explicitly restated
the abnormality if present and answer the question in the “clinical correlation” section of the report.
posed by the referring physician. In the author’s Examples of reports of normal VEP, BAEP, and SEP
opinion, noting “clinical correlation required” in studies are presented (Figures 1.17A–C).
Age: 3 years
Sex: Male
Scale: Amplitude = 0.1 mV/div; Latency = 1.5 ms/div
Side: Left
Click polarity: Condensation
(+↑)
Absolute Latencies
I 2.55 0.53 (I–In)
III 4.89
V 8.04 0.18 (V–Vn)
Vc 8.16
0.40 (V/I ratio)
Interpeak Latencies
I–Vc 5.61
I–III 2.34
III–Vc 3.27
Discussion: IPL prolongation helps localize the site of abnormality in evoked potentials. A prolonged IPL suggests a lesion between
the generators of the waveforms between which the latency is being measured. In this example, the waves I–Vc IPL is prolonged,
suggesting that the lesion is in the auditory pathway between the vestibulocochlear nerve (generator for wave I) and the inferior
colliculus (generator for wave V). Further localization is possible with waves I–III and III–Vc IPLs. The waves I–III IPL is within normal
limits, but the waves III–Vc IPL is prolonged. This suggests that the lesion is likely between the superior olivary complex (generator
for wave III) and the inferior colliculus.
BAEP, brainstem auditory evoked potential; IPL interpeak latency.
Figure 1.16 Prolongation of IPLs of BAEP waveforms.

(A)
Visual Evoked Potential sample report
HISTORY: This is a XX-year-old patient who is undergoing a visual evoked potential study to evaluate for visual pathway dysfunction. Current
medications include: XXXXXX
REPORT: Pattern reversal visual evoked responses were obtained using a XX′ check size visual arc following independent stimulation of the left
and right eyes at a rate of X/s. This resulted in high signal to noise ratio with good reproducibility of the waveforms. The P100 waveform absolute
latencies were normal at XX ms and XX ms following independent stimulation of left and right eyes, respectively. There was no significant side-to-
side amplitude asymmetry. Parasagittal P100 waveform derivations did not show any significant latency shifts or amplitude asymmetries. The visual
acuity was XX and XX of the left and right eyes, respectively.
INTERPRETATION: This is a normal visual evoked response study.
CLINICAL CORRELATION: This study does not demonstrate an abnormality in the visual pathways.
(B)
Brainstem Auditory Evoked Potential sample report
HISTORY: This is a XX-year-old patient who is undergoing a BAEP study to evaluate for brainstem dysfunction. Current medications include:
REPORT: Brainstem auditory evoked potentials were obtained following independent monaural stimulation with XX dBnHL XXXXX clicks delivered
at a rate of XXX/s. All obligate waveforms were obtained with good reproducibility. The waves I–Vc IPLs were normal at XX ms and XX ms following
stimulation of the left and right ears, respectively. Other IPLs and the amplitudes of the obligate waveforms were normal. A behavioral audiogram
was normal.
INTERPRETATION: This is a normal BAEP study following independent stimulation of both ears.
CLINICAL CORRELATION: This study does not demonstrate abnormality in central conduction involving the brainstem auditory pathways.
(C)
Somatosensory Evoked Potential sample report
HISTORY: This is a XX-year-old patient who is undergoing an SEP study to evaluate for dorsal column dysfunction. Current medications
include:
REPORT: SEPs were obtained in the upper extremities following independent stimulation of bilateral median nerves at a rate of XX/s. All obligate
waveforms were obtained with good reproducibility. The waves EP–N20 IPLs were normal at XX ms and XX ms following independent stimulation
of the left and right median nerves, respectively. The amplitude of the various components and other IPLs were normal. The median nerve con-
duction velocity was XX m/s.
SEPs were obtained in the lower extremities following independent stimulation of bilateral tibial nerves at a rate of XX/s. All obligate waveforms
were obtained with good reproducibility. The waves LP–P37 IPLs were normal at XX ms and XX ms following independent stimulation of the left
and right tibial nerves, respectively. The amplitude of the various components and other IPLs were normal. The tibial nerve conduction velocity
was XX m/s.
INTERPRETATION: This is a normal somatosensory evoked potential study following independent stimulation of bilateral median and tibial nerves.
CLINICAL CORRELATION: There is no evidence of conduction abnormalities along the peripheral nerves stimulated, dorsal columns, or medial
lemniscal pathways.
Discussion: These are sample PRVEP (A), BAEP (B), and median and tibial nerve SEP (C) reports of normal studies. All absolute and interpeak
latencies do not need to be listed; only the most significant ones should be noted. Reports for abnormal studies should be appropriately
amended.
BAEP, brainstem auditory evoked potential; IPL, interpeak latency; SEP, somatosensory evoked potential; VEP, visual evoked potential.
Figure 1.17 Sample VEP, BAEP, and SEP reports.

are available to enhance the waveforms. Guidelines

Conclusions are available to help provide consistency in acqui-
Clinical evoked potentials are performed in many sition and interpretation. Reports must be accurate
neurophysiology laboratories, with VEP, BAEP, and successfully relay the important findings of the
and SEP being the ones done most often. Since these study. Like any neurophysiologic study, evoked
potentials are of very low amplitude, attention to potentials should be repeated if additional data at a
detail during acquisition is critical. Many techniques different time point would be helpful.
References
1. Misulis KE. Spehlmann's Evoked Potential Primer. 4. ACNS. Guideline 9A: guidelines on evoked poten-
2nd ed. Boston, MA: Butterworth-Heinemann; 1994. tials. J Clin Neurophysiol. 2006;23:125-137.
2. Chiappa KH. Principles of evoked potentials. 5. Chiappa KH, Ropper AH. Evoked potentials in
In: Chiappa KH, ed. Evoked Potentials in Clinical clinical medicine (first of two parts). N Engl J Med.
Medicine. 3rd ed. Philadelphia, PA: Lippincott- 1982;306:1140-1150.
Raven Publishers; 1997:1-30. 6. Yamada T, Kameyama S, Fuchigami Y, et al. Changes
3. Eisen A, Purves S, Hoirch M. Central nervous of short latency somatosensory evoked potential
system amplification: its potential in the diagnosis in sleep. Electroencephalogr Clin Neurophysiol.
of early multiple sclerosis. Neurology. 1982;32: 1988;70:126-136. Figure 1.4 (A–E) Visual evoked
359-364. potentials (pattern reversal)
2
Visual Evoked Potentials
Visual evoked potentials (VEPs) are obtained with maintain focus on a checkerboard pattern. These
stimulation of the visual pathway. VEPs are of rela- are discussed in more detail later in this chapter.
tively high amplitude when compared to other types PRVEP can be further classified based on how
of evoked potentials, and consequently are easier to much of the visual field is stimulated. A full-field
record. Changes in the character of the stimulus can stimulus is when the temporal and nasal halves of
affect the response, and so the stimulus parameters the visual field of one eye are stimulated. A VEP
must be kept constant. This chapter discusses basics obtained with such a stimulus is called a full-field
of VEP classification, anatomy, methodology, inter- PRVEP. Alternatively, the nasal or temporal half of
pretation, special pediatric considerations, and clin- one or both visual fields can be stimulated. This is
ical applications. VEP examples with neuroimaging called a monocular or binocular hemifield PRVEP.
correlations will follow. Hemifield PRVEP is very time consuming and nor-
mative data are not very precise. Consequently,
these are seldom used clinically.
Classification
VEPs are classified based on the method of stim-
ulation used to elicit responses. The stimulation
Anatomy
method used most commonly in clinical practice is Visual perception starts with light entering the eye
a checkerboard pattern. The light and dark squares, from the cornea, passing through the pupil and
or checks, alternate to provide the stimulus. This lens to the retina. The lens inverts the light, so that
is known as pattern reversal visual evoked poten- images from the nasal aspect of the visual field are
tial (PRVEP). Pattern shift VEPs are very similar to reflected on to the temporal aspect of the retina,
PRVEP and differ only in the manner in which the images from the superior visual field are reflected
light and dark checks switch positions. PRVEP has on the inferior retina, and vice versa. The retina,
tightly defined normative data. The patient must which is the start of the visual pathway, contains
be alert and focus on the checkerboard pattern to photoreceptors known as cones and rods. Cones
obtain reliable waveforms. detect color vision in bright light and are concen-
A strobe light can also be used to elicit a VEP. trated in the center of the retina. Rods mediate
This diffuse light response is known as a flash vision in dim light and are concentrated around the
visual evoked potential (FVEP). The normative periphery of the retina. The centermost part of the
data for FVEP are not clearly defined, and they are retina is the macula, and the center of the macula
interpreted simply as present or absent. A FVEP is is the fovea. Cones are of highest concentration in
not sensitive in detecting visual pathway abnormal- the fovea, and it is responsible for the highest visual
ities, and consequently is used only when PRVEP acuity. The fovea represents the central 3° of vision.
cannot be used. The principle use of FVEP is in Cones and rods project to bipolar cells, which,
very young children or older patients who cannot in turn, project to ganglion cells. The axons of the
35
ganglion cells make up the optic nerve. The fovea Because of the mesial representation of the visual cor-
contributes a disproportionately large number of tex, the VEP waveforms are often better seen over the
axons to the optic nerve and, consequently, has contralateral hemisphere, that is, the VEP generated
a proportionately higher cortical representation. from the left visual cortex is often best recorded with
Optic nerve fibers are somatotopically arranged in electrodes over the right occipital (RO) lobe.
the optic nerve as it extends backward from the eye. Because of the partial decussation of the visual
The mesial-most fibers are from the nasal half of the pathway, the right visual field (i.e., fibers arising
retina, while fibers in the lateral aspect are from the from the nasal retina of the right eye and temporal
temporal half of the retina. retina of the left eye) projects to the left visual cor-
Both optic nerves progress posteriorly and meet tex. The VEP from the left visual cortex is best seen
at the optic chiasm. Fibers from the nasal half of the over the RO lobe. Conversely, the left visual field
retina cross to the contralateral side and join with (i.e., fibers arising from the nasal retina of the left
fibers from the temporal half of the opposite retina. eye and temporal retina of the right eye) projects to
Together the contralateral nasal and ipsilateral tem- the right visual cortex, and the VEP is best seen over
poral retina fibers form the optic tract. Optic tract the left occipital (LO) lobe.
travels to the lateral geniculate body and synapses.
From the lateral geniculate body the optic radiation
extends to the posterior aspect of the occipital lobe Methodology
known as the striate cortex, which is the primary The stimulation methodology can greatly affect
visual area (often referred to as the visual cortex). the VEP response, and so these parameters must
A small group of fibers from the contralateral nasal be carefully monitored. Once normative data are
retina project to the temporal horn of the lateral obtained, the stimulation parameters should not be
ventricle before proceeding back to the occipital changed. If normative data from another laboratory
cortex; this projection is known as the Myer’s loop are used, the stimulation parameters must be kept
(1). The visual pathway is illustrated in Figure 2.1. the same as the reference laboratory’s.
The visual cortex extends from the lateral to the
mesial surface of the occipital lobe. Almost one-third
of this area represents the fibers from the fovea.
Patient Variables
The patient undergoing a VEP study should be
seated comfortably in a chair at a standard distance
from the monitor that will display the stimulus.
Visual acuity should be determined. If the patient
has a refractive error, this should be corrected
with appropriate lenses (glasses or contact lens).
Optic Nerve Impaired visual acuity can result in prolonged laten-
cies of waveforms (Figure 2.2A and B). If full-field
PRVEP is being done, the nonstimulated eye is cov-
Optic ered with an eye patch. With the uncovered eye, the
Chiasm
patient is asked to focus at the center of the screen
displaying the checkerboard pattern. It is import-
ant for the patient to stay alert during the test, and
Optic Tract the technologist should look at the patient’s eyes
to ensure that he or she is looking at the stimulus.
If the patient shifts focus away from the stimulus,
LGB LGB
the PRVEP can be adversely affected. A patient can
intentionally not focus on the checkerboard pattern
Optic Radiation
even while looking at it, which also will adversely
Occipital Occipital affect the response (Figure 2.3A and B).
Cortex Cortex
If hemifield stimulation is used, one-half of the
image should be eliminated so that only one hemi-
Figure 2.1 Diagrammatic representation of the visual field is stimulated. Alternatively, the patient can
pathway arising from the retinae LGB (lateral geniculate body). be asked to look at the corner of the image, which
Chapter 2. Visual Evoked Potentials ■ 37
will also result in stimulation of a hemifield. With the eye, thus increasing the luminance. Careful
hemifield stimulation, it is even more critical that inspection of the pupils should be undertaken
the patient remain alert and attend to the stimulus. before stimulation to ensure that it is not dilated.
Ambient light can also affect luminance of the stim-
Stimulus Parameters ulus. The room in which the test is being performed
Various stimulus parameters affect PRVEP. It is should be dimly lit to minimize the effects of ambi-
important to realize the effect that changing these ent light.
parameters can have on the response. Display monitor
Stimulation rate Traditionally cathode ray tube (CRT) monitors
VEP stimulation rate is usually between 1 and 4/s. have been used to display the checkerboard pat-
Care must be taken to avoid a stimulation rate that tern. Most laboratories have obtained their norma-
is an integer of 60. Slower frequencies take much tive data with these monitors. With technological
longer to produce a reliable response, and patients advances, CRT monitors are being replaced with
may lose attention and not focus on the stimulus. liquid crystal display (LCD) monitors. However,
Faster stimulation rates result in one VEP response even LCD monitors with the fastest refresh rate
overlapping the one before it, resulting in a steady (2 ms) cause a delay in the P100 waveform latency
state VEP. Steady state VEPs are not used clinically. and may lead to erroneous interpretation (Figure
2.5A–D). Slow refresh rate (30 ms) LCD moni-
Luminance and contrast tors cause an extreme latency prolongation (2).
Luminance refers to the intensity of the light stim- If LCD monitors are used, normative data must
ulus. As the luminance decreases, the VEP latency be obtained using the monitor that will be used
increases. Amplitude is affected less, but can during clinical testing.
decrease with decreasing luminance. The lumi-
nance can be adjusted with the brightness controls Visual angle
of the display monitor. Because of the effect lumi- The visual angle is the measure of the size of image
nance can have on VEP waveforms, it should be created by one check of a checkerboard pattern on
locked at a particular value (at which normative the retina. It is a function of the size of the check and
data were obtained). the distance of the check from the eye. The visual
Related to luminance is the contrast between the angle (β) is expressed as a/b, where a is the size of the
light and dark checks of the checkerboard pattern. check and b is the distance from the eye. The size of
Higher degrees of contrast result in shorter VEP checks used in PRVEP is expressed in minutes of arc
latencies. The brightness controls of the display (1° = 60 minutes or 60′). Checks of smaller size stim-
monitor can also affect contrast the same way it can ulate the fovea, whereas larger check sizes stimulate
affect luminance. more of the peripheral retina. Because of this, in general
A photometer is used to measure luminance. For smaller check sizes produce VEP of shorter latency.
checkerboard patterns, the luminance of the light However, VEP latency and amplitude are also very
and dark checks can be measured independently. dependent on the ability of the retina to clearly distin-
Once the luminance is measured, the contrast guish between the light and dark checks. If the checks
between the light and dark checks is expressed in become so small that this distinction is compromised,
percentage according to the following formula: the VEP latency increases and amplitude decreases.
(Lmax − Lmin)/(Lmax + Lmin) × 100 (L = luminance as Thus, the optimal check size is the smallest one whose
measured by a photometer). The VEP latency and margins can be clearly distinguished. Most often this
amplitude are not affected much beyond a contrast is a 30-minute check size, but for some patients with
of 40% (Figure 2.4A and B). In most clinical stud- good visual acuity it may be 15 minutes. Others who
ies, the contrast is kept between 95% and 99%. The have a refractive error may need a check size of 60
luminance of a light flash is difficult to measure, minutes. In the author’s laboratory, a study is usu-
so it is measured by increasing the light flash fre- ally started with a 30-minute check size. If the study
quency above the fusion frequency. At that point, shows slight prolongation of latency (less than 8 ms
the light is on constantly. above the upper limit of normal), it is repeated with
Pupillary size also affects luminance. Dilatation a 15-minute or 60-minute check size depending on
of the pupil increases the amount of light entering whether the visual acuity is normal or reduced. If the
Age: 42 years
Sex: Male
Filters: 1–100 Hz (+↓)
Preauricular–preauricular distance: 38 cm
Scale: Amplitude = 2 μV/div; Latency = 25 ms/div
Eye: Right
Visual angle: 30′
(A)

N75 P100 N105 N75–P100 P100–N145
Derivation (ms) (ms) (ms) N145 (ms) Derivation (μV) (μV) Mean (μV)
MO–MF 86.0 106.0 139.0 MO–MF 4.40 5.86 5.13
MO–AU 85.5 108.0 139.0
MO–AU 2.61 3.19 2.90
Oz–AU 85.5 108.0 137.0
Oz–AU 2.91 3.83 3.37
MF–AU 104.0
Figure 2.2 (A and B) Effects of changing visual acuity with corrective lenses. (continued )
(+↑)
(B)

MO–MF 82.5 109.0 134.0 Derivation (μV) (μV) Mean (μV)
MO–AU 85.0 109.0 136.0 MO–MF 4.94 3.28 4.11
Oz–AU 85.0 109.0 140.0 MO–AU 2.82 1.40 2.11
MF–AU 109.0 Oz–AU 3.42 3.06 3.24
Discussion: Visual acuity should be tested before PRVEP are performed. If the patient wears corrective lenses, they should be used
during the test. In the above example, a PRVEP is shown with (A) and without (B) corrective lenses. Without corrective lenses, the
visual acuity changes from 20/20 to 20/50. The MO–MF derivation is used for latency and amplitude measurements. Notice that the
P100 waveform has a longer latency and lower amplitude when the PRVEP is performed without corrective lenses. This effect would
be more exaggerated with a smaller check size (15′) and less prominent with larger check size (60′).
PRVEP, pattern reversal visual evoked potential.

Age: 42 years
Sex: Male
Eye: Right
Visual angle: 30′
(A)

MO–AU 85.5 108.0 139.0 MO–MF 4.40 5.86 5.13
Oz–AU 85.5 108.0 137.0 MO–AU 2.61 3.19 2.90
MF–AU 104.0 Oz–AU 2.91 3.83 3.37
Figure 2.3 (A and B) Effects of not focusing on stimulus. (continued )

(+↓)
(B)

MO–AU N/A N/A N/A MO–MF 1.98 1.62 1.80
Oz–AU N/A N/A N/A MO–AU N/A N/A N/A
MF–AU 111.0 Oz–AU N/A N/A N/A
Discussion: When obtaining PRVEP, it is very important that the technologist makes sure that the patient is viewing the monitor on
which the pattern is displayed. If the patient looks away, the VEP will not be obtained. Even when a patient appears to be looking at
the monitor, he or she may not focus on it. In this example, PRVEP with the patient focusing (A) and not focusing (B) on the pattern
are shown. When the patient does not focus on the pattern, the VEP is difficult to identify, and the latency of the P100 waveform
is longer and amplitude much smaller.

Age: 42 years
Sex: Male
Eye: Left
Visual angle: 30′
(A)

MO–AU 85.5 105.0 125.0 MO–MF 4.43 4.87 4.65
Oz–AU 85.5 105.0 125.0 MO–AU 2.20 2.42 2.31
MF–AU 103.0 Oz–AU 2.79 3.15 2.97
Figure 2.4 (A and B) Effects of changing contrast of the checkerboard pattern. (continued )
(B)
Amplitude
Absolute Latency N75–P100 P100–N145
Derivation N75 (ms) P100 (ms) N105 (ms) N145 (ms) Derivation (μV) (μV) Mean (μV)
MO–MF 91.0 122.0 160.0 MO–MF 3.30 5.05 4.18
MO–AU 91.0 122.0 160.0 MO–AU 2.44 3.07 2.76
Oz–AU 91.0 122.0 160.0
Oz–AU 2.12 3.19 2.66
MF–AU 119.0
Discussion: Luminance of the checkerboard pattern on the monitor can have a significant effect on the PRVEP. The contrast
between the dark and light squares should be kept between 95% and 99%. In this example, PRVEP with contrast at 98% (A) and
25% (B) are shown. The MO–MF derivation is used for latency and amplitude measurements. The P100 waveform has a longer
latency, lower amplitude, and poor morphology with the reduced luminance.

Age: 45 years
Sex: Female
Filters: 1–100 Hz
Eye: Left
Visual angle: 30′
(A) (B)
P100 Latency
Figure Type of Monitor (MO–MF), ms
A CRT 109
B LCD 2 ms refresh rate 114
C LCD 8 ms refresh rate 118
D LCD 30 ms refresh rate 158
Figure 2.5 (A–D) Effects of changing the checkerboard display monitor. (continued )
(C) (D)
Discussion: Changing the PRVEP stimulating monitor can affect the P100 waveform latency. In these examples, a CRT monitor (A)
and three LCD monitors were used. The LCD monitors had refresh rates of 2 ms (B), 8 ms (C), and 30 ms (D). Generally, LCD mon-
itors with faster refresh rates are better at showing fast-moving pictures. The upper limit of normal for the P100 waveform latency
for a head size of 33 cm is 110 ms in the author’s laboratory. Even a good LCD monitor with a 2 ms refresh rate increases the P100
waveform latency to the extent that the study would be considered abnormal (114 ms) when using available normative data. When
an LCD monitor with an 8 ms refresh rate is used, the latency of the P100 waveform further prolongs to 118 ms. The LCD monitor
with a refresh rate of 30 ms caused a remarkable prolongation of latency to 158 ms. If stimulating monitors are changed, normative
data must be obtained again or reconfirmed. This is particularly true if a CRT monitor is replaced with an LCD monitor.
CRT, cathode ray tube; LCD, liquid crystal display; PRVEP, pattern reversal visual evoked potential.
Figure 2.5 (A–D) (continued )
abnormality persists, the VEP should be repeated stimulation. During full-field stimulation, the
again with the third check size (Figure 2.6A and B). patient is asked to look at the center of the screen dis-
playing the checkerboard pattern. The fovea, which
Field of stimulation is most sensitive to small check sizes, contributes
PRVEPs are usually obtained with stimulation of the most to the VEP with full-field stimulation, with
each eye independently. This is known as full-field the peripheral parts of the retina contributing only
Age: 39 years
Sex: Male
Stimulation Rate: 3.9/s
Filters: 1–100 Hz
Eye: Right
(A)
Visual angle: 30′

MO–AU 103.0 124.0 147.0 MO–MF 3.65 5.34 4.50
Oz–AU 103.0 122.0 147.0 MO–AU 1.51 3.23 2.37
MF–AU 122.0 Oz–AU 1.96 3.67 2.82
Figure 2.6 (A and B) Effects of changing checkerboard check size. (continued )

(B)
Visual angle: 60′
Discussion: When the P100 waveform latency is slightly Absolute Latency

prolonged (<10 ms above upper limit of normal in the
Derivation N75 (ms) P100 (ms) N105 (ms) N145 (ms)
author’s laboratory), a larger (or smaller) check size should
be tried to determine if the prolongation is due to impaired MO–MF 73.0 112.0 153.0
visual acuity. In this example, the initial study (A) was MO–AU 73.0 111.0 149.0
obtained with a 30′ check size, and the subsequent study
(B) obtained with a 60′ check size. The MO–MF derivation Oz–AU 74.0 108.0 149.0
is used for latency and amplitude measurements. With the MF–AU 117.0
30′ check size stimulation, the P100 waveform latency is
prolonged at 123.0 ms (upper limit of normal is 117.0 ms
for a preauricular to preauricular distance of 39 cm). Since Amplitude
this prolongation is less than 10 ms above the upper limit
N75–P100 P100–N145
of normal and the visual acuity is 20/50, PRVEP is repeated
with a 60’ check size. The P100 waveform latency normal-
Derivation (μV) (μV) Mean (μV)
izes to 112.0 ms with the larger check size. If the patient’s MO–MF 3.99 6.84 5.42
visual acuity was 20/20 or 20/15, a 15-minute check size
MO–AU 2.99 4.99 3.99
should have been attempted. The P100 waveform latency
is shortest when the smallest check size that can be clearly Oz–AU 3.07 5.14 4.11
seen is used. If the check size becomes so small that the
transition between the light and dark squares is blurred, the
P100 waveform latency starts to increase.

slightly to the overall amplitude. A field of stimula- Square” method (Figure 2.7). The following
tion of 6° to 12° is sufficient for full-field PRVEP as electrodes are placed to record full-field VEP:
that will adequately stimulate the fovea.
Hemifield stimulation can also be used to obtain MO: Midoccipital; 5 cm above the inion
PRVEP. With hemifield stimulation, the temporal or MF: Midfrontal; 12 cm above the nasion
nasal half of the retina is stimulated. This type of stim- LO: Left occipital; 5 cm lateral (left) to MO
ulation is done by having the patient look at the corner RO: Right occipital; 5 cm lateral (right) to MO
of the checkerboard pattern or displaying the pattern A1/A2: Left and right ear or mastoid
in only one-half of the screen and having the patient Ground: Vertex
look at the center of the screen. The peripheral parts
of the retina are stimulated, and since these areas Alternatively, recording electrodes can be placed
are not sensitive to small check sizes, checks of 30 or according to the International 10/20 electrode place-
60 minutes are most often used. Additionally, since ment system in which the midline posterior electrode
the peripheral part of the retina is stimulated, the is placed at Oz, lateral posterior electrodes are placed
field of stimulation is at least 16°. The topography of at O1 and O2, and the midline frontal electrode is
the recorded VEP is different in hemifield versus full- placed at Fz. In this system, the Fz electrode is approx-
field stimulation, and this will be discussed later. imately 11 cm above the nasion and the O1 and O2
electrodes are not as laterally placed as LO and RO.
Recording Parameters Occasionally further lateral electrodes are also
Various recording parameters can affect VEP. placed. These are called LT and RT (left and right
Among the most important is electrode placement temporal) electrodes and are placed 10 cm lateral to
as it is not standardized between laboratories. MO. They are most useful in recording hemifield VEP.
The Queen Square method is preferred as it
Electrode placement results in better VEP morphology (Figure 2.8).
Electrodes are placed along the posterior and ante- There are exceptions, however, and sometimes
rior head regions to record VEP (3). At least one the P100 waveform is seen better in the Oz deriva-
midline posterior, two lateral posterior, and one tion than the MO derivation. When this occurs, the
midline anterior electrode is necessary. Placement parasagittal electrodes should be moved to the Oz
of electrodes is usually according to the “Queen plane rather than keeping them in the MO plane.
Vertex
(CZ) Vertex (CZ) Vertex (CZ)
MF MF
12cm up from
12cm up from inion nasion on midline
along midline
50% of measurement from
50% of measurement from nasion to inion
50% of measurement
nasion to inion from pre-auricular to pre-auricular
50% of measurement from
pre-auricular to pre-auricular
Parasagital plane Parasagital plane 5 cm outward along Parasagital plane Parasagital plane
parasagital plane from MO
LO RO Nasion
MO 5cm Nasion 5cm up from
MO
RO pre-auricular
5 cm outward along 5 cm outward along Preauricular
parasagital plane from parasagital plane from MO 5 cm up from inion
MO 5cm up from inion along midline
Inion A2 A2
Inion
Ears (A1 and A2)
Midfrontal (MF)
Midoccipital (MO)
Lateral occipital
(LO and RO)
Measurement
landmarks
Figure 2.7 Diagrammatic representation of scalp electrodes according to the Queen Square method.
Source: Image courtesy Ms. Emily Kale.
Age: 34 years
Sex: Male
Filters: 1–100 Hz
Preauricular–preauricular distance: 36
Eye: Right
Visual angle: 30′

MO–AU 82.5 107.4 142.6 MO–MF 19.9 21.8 20.85
Oz–AU 81.5 107.4 142.5 MO–AU 19.9 19.5 19.70
MF–AU N/A Oz–AU 7.5 8.5 8.00
Discussion: The Queen Square and the International 10/20 systems can be used for placing electrodes to record PRVEP. The MO
electrode (Queen Square method) is placed slightly above the Oz electrode. The MO electrode is often better for recording the P100
waveform than the Oz electrode. This is demonstrated in this example. Note that the P100 waveform is of much higher amplitude
in the MO derivation than in the Oz derivation.
Figure 2.8 Effects of different recording electrode placements.

Recording montage Amplifier settings

A minimum of four channel montages is recom- Low and high frequency filter settings used to
mended for recording VEP (3). Additional channels record VEP are 1 to 100 Hz with a 12 dB/octave roll
may be necessary to resolve waveforms of unusual off slope. The analysis time is 250 ms, however if
morphology. The usual montage used is: responses are markedly delayed, an analysis time of
500 ms may be necessary. Usually 100 to 200 repeti-
Channel 1: LO–MF tions are needed to obtain a reproducible response.
Channel 2: MO–MF More repetitions may be needed in the case of a low
Channel 3: RO–MF amplitude response and for hemifield recordings.
Channel 4: MF–AU The response must be replicated at least once.
The polarity convention for VEP is not standard-
Additional channels may be necessary when a ized. Some laboratories display positivity upward,
“W-shaped” P100 waveform is noted. In this situa- while others display positivity downward. Since
tion, it is helpful to reference LO, MO, and RO–AU. there is no standardized way to display VEP, the
When low amplitude P100 waveform is noted, an tracing should have clear markings indicating
Oz, inion, and midparietal electrode referenced which polarity convention was used.
to MF may result in a larger amplitude response.
This happens when the P100 waveform peak is dis- Interpretation
placed downward or upward (Figure 2.9). Finally,
LT–MF and RT–MF channels may be needed when Interpretation of VEP involves identifying wave-
recording hemifield VEP. forms that are expected to be present and determin
In the author’s laboratory, a few extra channels ing their latency. It is also important to compare
are added to clarify the various waveforms and latencies between left and right eyes to evaluate for
compensate for variability within patients. These asymmetry.
extra channels are:
Expected Waveforms
Channel 5: MO–AU The most prominent waveform of the PRVEP is
Channel 6: Oz–AU a positive peak best seen in the midline occipital
Channel 7: LT–MF region occurring between 90 and 115 ms. This is
Channel 8: RT–MF known as the P100 waveform and is thought to arise
from the occipital (visual) cortex, though the exact
Table 2.1 shows a typical montage with the wave- location within the occipital cortex is not clear. It
forms seen in each derivation. is often preceded and followed by smaller negative
Table 2.1 The Channels of a VEP and the Waveforms Seen in Those Channels
Montage Waveform(s)
MO–MF N75, P100, N145; highest amplitude
MO–AU N75, P100, N145
Oz–AU N75, P100, N145; may be higher in amplitude depending on vector projection
MF–AU N105
LT–MF N75, P100, N145; from contralateral visual cortex, low amplitude
LO–MF N75, P100, N145; from contralateral visual cortex, moderate amplitude
RO–MF N75, P100, N145; from contralateral visual cortex, moderate amplitude
RT–MF N75, P100, N145; from contralateral visual cortex, low amplitude
VEP, visual evoked potential.

Age: 34 years
Sex: Female
Eye: Left
Visual angle: 30′

MO–AU 82.5 99.5 128.0 MO–MF 4.77 7.98 6.38
Oz–AU 79.0 99.5 128.0 MO–AU 3.55 5.53 4.54
MF–AU 104.0 Oz–AU 5.29 8.70 7.00
Discussion: As discussed previously, usually the MO electrode is optimally placed to record the highest amplitude P100 waveform.
However, in some patients the P100 waveform vector may be caudally or rostrally displaced, making the P100 waveform appear
best in the Oz (or inion) or Pz electrodes, respectively. In this example, the P100 waveform is of higher amplitude in the Oz–AU
derivation than in the MO–AU derivation, suggesting that the P100 waveform vector is caudally displaced. This is a normal variant.
Figure 2.9 Variant topography of the P100 waveform.

peaks known as the N75 and N145 waveforms. Parameters Measured

A negative waveform is also recorded in the mid-
line frontal electrodes at about 105 ms, known as Latency and amplitude are important parameters to
the N105 waveform (it is also referred to as the measure for PRVEP. Topography and morphology
N100) (Figure 2.10A and B). The site or origin of the of the P100 waveforms should also be analyzed. Of
N75, N145, and N105 waveforms is uncertain, but these parameters, latency measures are the most
they are all thought to arise from the cortex (4–6). important.
The likely generator sites of these waveforms are Latency
listed in Table 2.2.
The most important measurement in PRVEP is the
Since monocular stimulation results in equal
absolute latency of the P100 waveform. With mon-
excitation of left and right visual pathways, both
ocular stimulation, this is measured at the MO site,
the left and right visual cortices are symmetrically
usually in the MO–MF derivation. In the author’s
activated and produce a P100 waveform with max-
laboratory, normative data are dependent on head
imal amplitude in the midline. The P100 waveform
size as measured by the preauricular to preauric-
is best seen in the MO–MF montage. This derivation
ular (P–P) distance (Table 2.3). In addition to the
takes advantage of the positivity (P100 waveform)
absolute latency, the P100 waveform latency differ-
in the active electrode and small negativity (N105
ence measured at the MO site after left and right
waveform) in the reference electrode. Alternatively,
eye stimulation should also be determined. A sig-
MO–AU derivation may have slightly smaller
nificant interocular latency difference (≥8 ms in
amplitude, but will show the true amplitude of
the author’s laboratory) is abnormal, even if the
the P100 waveform without contribution from the
absolute latency of the longer side is within normal
N105 waveform (Figure 2.10A and B). Lesions in
limits.
the visual pathway may cause the P100 waveform
to be prolonged and to be better seen over the left Amplitude
or right hemisphere. Amplitude measurements of the P100 waveform are
When only one visual cortex is activated, either also evaluated but are less important than latency
with hemifield stimulation or due to pathology, the measurements. After monocular stimulation, the
P100 waveform is better visualized over the con- P100 waveform amplitude is measured at the MO
tralateral side. For example, if only the left visual site, usually in the MO–MF derivation. The ampli-
cortex is activated with hemifield stimulation, the tude is measured from baseline to peak of the P100
P100 waveform is best seen over the RO and RT waveform or from peak to peak of the N75 to P100
electrodes. This is referred to as paradoxical later- waveforms or of the P100 to N145 waveforms. The
alization and is discussed in more detail later in the P100 waveform amplitude ratio is determined after
chapter. left and right eye stimulation. This is the interocu-
At times, more than one check size will be used lar amplitude ratio, and it should be less than 2.0
to obtain a VEP. This is because an abnormality was (Figure 2.10A and B).
noted with one check size, and a larger or smaller
check size is attempted to confirm that the abnor- Topography
mality was not due to a refractive error (Figure 2.6A Topography of the P100 waveform is evaluated by
and B). If the VEP parameters normalize with any determining its amplitude in the parasagittal occip-
check size, the author’s practice is to consider the ital electrodes. The amplitude of the P100 waveform
study normal. is measured at the LO and RO sites after stimulation
Table 2.2 VEP Waveforms and Their Neural Generators

N75 Striate cortex
P100 Dorsal extrastriate cortex
N145 Ventral extrastriate cortex
N105 Prefrontal cortex
VEP, visual evoked potential.

Table 2.3 PRVEP Normative Data*

P–P (cm) Latency (ms)†
30 106.5
31 107.5
32 109.0
33 110.0
34 111.0
35 112.0
36 113.5
37 114.5
38 116.0
39 117.0
40 118.5
P–P, preauricular to preauricular; PRVEP, pattern reversal visual evoked potential.

* Data from Evoked Potentials Laboratory, Duke University Medical Center, Durham, NC.
† Latency values least square fit + 3 SEM.
Stimulation rate ≈ 4/s.
Interocular latency asymmetry <8 ms.
Interocular amplitude asymmetry (larger/smaller) <2.0.
Parasagittal amplitude asymmetry (larger/smaller) <2.5.
of each eye. This is known as the parasagittal or waveform. The most reliable measure of abnormal-
interhemispheric amplitude ratio, and it should be ity is latency prolongation as this is least affected by
less than 2.5 (Figure 2.10A and B). The latency of the technical and patient cooperation factors. Latency
P100 waveform can also be determined at the LO and amplitude variability of the N75, N105, and N145
and RO sites. They, too, should be symmetric, with waveforms are not considered clinically significant.
the latency difference between the two parasagittal Figure 2.11A and B illustrates the expected monoc-
derivations being less than 8 ms. ular full-field VEP findings depending on the site of
lesion. Chiasmatic and postchiasmatic lesions should
Morphology be confirmed with hemifield stimulation when pos-
The typical P100 waveform morphology consists of sible (discussed as follows). If hemifield stimulation
a single large positive waveform (P100) preceded cannot be performed, these abnormalities should be
and followed by smaller negative waveforms (N75 interpreted with caution. Examples of various abnor-
and N145). This typical morphology is best seen malities are presented later in the chapter.
with a MO–AU derivation. The MO–MF results in
higher amplitude of the P100 waveform; however, Latency
the morphology may get distorted due to contribu- Latency abnormalities consist of prolongation of
tions from the N105 waveform. the P100 waveform latency or an abnormally large
interocular P100 waveform latency difference. The
side with the longer latency is abnormal. These
Abnormalities abnormalities are indicative of visual pathway dys-
PRVEP abnormalities may occur in latency, ampli- function. The following can be used as an interpreta-
tude, topography, and morphology of the P100 tion guide to P100 waveform latency prolongation.
Age: 54 years
Sex: Male
Preauricular–preauricular distance:
(A) P100 Waveform Amplitude Ratio

Eye: Left
Number of repetitions: 200 Location Ratio
Visual angle: 30′ OS–OD 1.10
Corrective lenses: Yes LO–RO 1.38

MO–AU 77.5 98.5 135.0 MO–MF 8.32 11.30 9.81
Oz–AU 77.5 98.0 135.0 MO–AU 6.74 8.80 7.77
MF–AU 102.0 Oz–AU 5.70 7.33 6.52
LO–MF 77.0 96.0 96.5 LO–MF 4.68 4.91 4.79
RO–MF 77.0 96.5 119.0 RO–MF 5.26 7.93 6.59
Figure 2.10 (A and B) Expected waveforms. (continued )

(B)
Eye: Right
(+↓) Visual angle: 30′
Absolute Latency
N105
Derivation N75 (ms) P100 (ms) (ms) N145 (ms)
MO–MF 78.5 100.0 134.0
MO–AU 78.0 101.0 134.0
Oz–AU 78.0 101.0 134.0
MF–AU 105.0
LO–MF 77.5 99.5 98.5
RO–MF 76.5 98.5 135.0
Amplitude
N75–P100 P100–N145
MO–MF 8.22 9.68 8.95
MO–AU 6.23 6.28 6.26
Discussion: When evaluating a PRVEP, several aspects of the
Oz–AU 5.24 5.01 5.13
waveforms are considered. The typical waveforms that are
expected are the N75, P100, N145, and N105 waveforms. LO–MF 4.72 5.52 5.12
Of these, the P100 waveform is the most important, and
the latency and amplitude of this waveform is measured and RO–MF 4.70 6.65 5.67
morphology considered. The topography of the P100 waveform
(symmetry in the parasagittal derivations) is also evaluated. In
the example shown, PRVEP after left (A) and right (B) eye stim- P100 Waveform Amplitude Ratio
ulation is shown. Notice that multiple derivations are shown
for latency and amplitude measurement. This is done to obtain Location Ratio
the best possible morphology of the VEP. In this example, the OS–OD 1.10
MO–AU derivation shows a higher amplitude response than
LO–RO 1.11
the Oz–AU derivation. The MO–MF derivation shows an even
higher amplitude as this derivation takes advantage of the neg-
ativity (N105) at the MF electrode. Latency measurements are
taken from the best response. The P100 waveform amplitudes between left and right eye stimulation are symmetric, as shown in
the OS–OD amplitude ratio. The parasagittal derivations (LO, RO) show symmetric responses after left and right eye stimulation. The
LO–RO amplitude ratios after stimulation of each eye are also symmetric.

Left Right
Monocular Monocular
full field full field
A
B
Occipital cortex
Left Right
LT LO MO RO RT
(A) (B)
LE, left eye; LT, left temporal; LO, left occipital; MO, mid occipital; RO, right occipital; RE, right eye; RT, right temporal;
V, higher amplitude P100 waveform; v, lower amplitude P100 waveform.
Figure 2.11 (A and B) A cartoon of the visual pathway is shown, indicating the pathways stimulated when monocular full
fields are activated (A). For example, when the left eye is stimulated (black), the nasal and temporal halves of the retina are
activated. The impulses travel in the left optic nerve. The nasal fibers decussate to the contralateral side in the optic chiasm, while
the temporal fibers stay ipsilateral. Both sets of fibers eventually proceed to the respective occipital cortices. The P100 waveform
generated by stimulating one eye will be best seen in the MO electrode, and smaller responses will be seen in the parasagit-
tal electrodes. The electrodes in which the P100 waveform will be seen with each eye stimulation are shown in the table (B).
Abnormalities suspected on monocular full-field stimulation should be confirmed with hemifield stimulation.
Monocular latency prolongation: Prechiasmatic (b) abnormally low amplitude, and (c) abnormally
lesion; if ocular and retinal disorders excluded, high interocular amplitude ratio. Amplitude can be
lesion in optic nerve. affected easily by technical factors and patient com-
Binocular latency prolongation without parasag- pliance, and these issues must be excluded before
ittal asymmetry: Bilateral visual pathway lesions, amplitude abnormalities are ascribed to an abnor-
likely symmetric bilateral prechiasmatic or bilateral mality of the visual pathway. The following can be
postchiasmatic. used as an interpretation guide to P100 waveform
Binocular latency prolongation with parasagit- amplitude abnormalities.
tal asymmetry: Chiasmatic or postchiasmatic, less Monocular absent waveform: Prechiasmatic
likely prechiasmatic; topographic analysis and lesion, but need to rule out technical issues; if ocular
hemifield stimulation can help further localize. and retinal disorders excluded, lesion in optic nerve.
Monocular low amplitude (abnormal interocular
Amplitude amplitude ratio) without parasagittal asymmetry
Amplitude abnormalities of the P100 waveform with normal latency: Prechiasmatic lesion; retinal
can be of three types: (a) absence of response, or ocular problem likely.
Monocular low amplitude with parasagittal 1. If the “W-shaped” waveform is seen in the
asymmetry: Possible chiasmatic or postchiasmatic; MO–MF derivation, it is possible that it is due
topographic analysis and hemifield stimulation can to a greater than usual difference in latency of
help further localize. the P100 and N105 waveforms. In such cases
Binocular low amplitude with parasagittal asym- a MO–AU derivation may resolve the abnor-
metry: Chiasmatic or postchiasmatic, less likely mality by removing the contribution of the
prechiasmatic; topographic analysis and hemifield N105 waveform.
stimulation can help further localize. 2. The “W-shaped” waveform may be due to
Binocular low amplitude without parasagittal contribution of the peripheral retina, which
asymmetry: Unclear clinical significance. activates slower conducting fibers of the visual
pathway. Use of a smaller check size may
Topography resolve the bifid peak by selectively activating
Topographic abnormalities of the P100 waveform only the fovea region.
consist of an abnormal interhemispheric (LO– 3. Due to retinal disease, a “W-shaped” wave-
RO) amplitude ratio or latency difference. In the form may occur due to abnormal activation
author’s laboratory, an interhemispheric amplitude of parts of the retina. In these cases, hemifield
ratio ≥2.5 is considered abnormal. Parasagittal P100 stimulation may be necessary.
waveform latencies are considered asymmetric if
the difference between them is ≥8 ms. An abnor-
mal P100 waveform interhemispheric amplitude Hemifield VEP
ratio or latency difference usually indicates a chi-
Hemifield VEPs are those responses obtained after
asmatic or postchiasmatic visual pathway abnor-
stimulation of one-half of the visual field. Nasal
mality. If the side of the lower amplitude or longer
and temporal halves of the visual fields can be
latency is consistently lateralized to one side with
stimulated separately to more selectively activate
stimulation of either eye, a lesion of the contralat-
corresponding parts of the retina. Details of stim-
eral postchiasmatic visual pathway is suspected
ulation and the waveforms that are recorded have
(recall paradoxical lateralization of the P100 wave-
been discussed earlier. A minimum P100 waveform
form with hemifield stimulation). This is known as
amplitude of 5 μV with full-field stimulation is
an uncrossed asymmetry. On the other hand, if the
needed to be able to perform hemifield stimulation.
side of lower P100 waveform amplitude or longer
P100 waveforms of smaller amplitude cannot be
latency switches sides with stimulation of the two
obtained with good reproducibility with hemifield
eye, a lesion of the optic chiasm is suspected. Such
stimulation (7).
an abnormality is referred to as a crossed asymmetry.
Whenever abnormal P100 waveform topography is
seen with full-field stimulation and chiasmatic or Indications
postchiasmatic lesion suspected, it must be con- Hemifield stimulation is used to localize chiasmal
firmed with hemifield stimulation. and postchiasmal visual pathway abnormalities.
Because abnormalities in these locations can be
Morphology visualized with neuroimaging, this type of testing
Morphologic abnormalities of the P100 wave- is seldom used. Hemifield VEPs are most often
form without latency and amplitude abnormali- used when the interhemispheric P100 waveform
ties are unlikely to represent clinically significant amplitude ratio or latency difference is abnormal
findings. Abnormal morphology can, however, after full-field stimulation.
make determination of the waveform latency dif- Hemifield stimulation can be done by stimulating
ficult. Consequently, improving the morphology is the temporal hemifield of one eye and nasal hemi-
important to better measure the various parameters field of the other eye simultaneously (left and right
of the P100 waveform. The commonest morpho- binocular hemifield stimulation) or the temporal and
logic abnormality is the “W-shaped” P100 wave- nasal hemifields of each eye can be stimulated inde-
form. Neither peak of the “W” should be assumed pendently (monocular hemifield stimulation). Both
to be the true P100 waveform. Several steps can be types require extensive patient cooperation, but mon-
taken to improve the morphology. ocular hemifield stimulation is more challenging.
The decision to perform binocular or monocular waveform must also not be confused with the N105
stimulation is based on the findings of interhemi- waveform recorded from the MF electrode after
spheric P100 waveform asymmetry (parasagittal P100 full-field stimulation.
waveforms, as recorded from LO and RO electrodes) Parameters measured
after full-field stimulation. If the asymmetry of the
The P100 waveform measurements typically per-
P100 waveforms between LO and RO derivations is
formed with full-field stimulation are not done
uncrossed (see under Topography in the Abnormalities
with hemifield stimulation. The P100 waveform is
section) after stimulation of both eyes, binocular hemi-
interpreted as being present or absent in various
fields can be performed. However, if the asymmetry is
derivations (LO–MF, RO–MF, LT–MF, RT–MF).
crossed, monocular hemifield stimulation is necessary.
This is because the latency and amplitude are very
Interpretation variable with hemifield stimulation, making norms
very difficult to define. Topography and waveform
Many of the principles of interpretation discussed
morphology of the P100 waveform obtained with
for full-field stimulation also apply for hemifield
hemifield stimulation are even more variable and a
stimulation. However, responses obtained with
detailed analysis is not performed. In the author’s
hemifield stimulation are smaller in amplitude and
laboratory, the P100 waveform of hemifield stimu-
less consistent than full-field stimulation and addi-
lation is interpreted as being present or absent.
tional caution must be exercised in interpretation.
If latency and amplitude measurements are to
Expected waveforms be performed on the P100 waveform obtained with
hemifield stimulation, the same criteria should be
Temporal hemifield stimulation causes activation
used as with full-field stimulation. Measurement
of the nasal retina and the contralateral visual cor-
of the P100 waveform latency to all hemifield stim-
tex, while nasal hemifield stimulation activates
ulations should be made. Additionally, the P100
the temporal retina and the ipsilateral visual cor-
waveform latency of the same hemifield after left
tex. When only one visual cortex is activated, the
and right eye stimulation should be compared. This
P100 waveform is paradoxically seen better over
is known as the interocular hemifield latency differ-
the side opposite from where it is generated, that is,
ence. The P100 waveform latency of temporal and
left visual cortex activation results in a better P100
nasal hemifield stimulation of the same eye should
waveform over the RO and RT electrodes. This
also be compared. This is called monocular hemi-
occurs because the P100 waveform arises from deep
field latency difference.
within the occipital cortex, and its vector projects to
Measurement of the P100 waveform amplitude
the contralateral hemisphere. Along with the P100
with hemifield stimulation is done in the same
waveform, the N75 and N145 waveforms are also
manner as with full-field stimulation, that is, base-
seen as with full-field stimulation. All three peaks
line to peak or peak to peak. The ratio of the P100
are of lower amplitude than their counterparts seen
waveform amplitude of the same hemifield after
with full-field stimulation. It should be remem-
left and right eye stimulation is determined and
bered, however, that whereas paradoxical lateral-
is called the interocular hemifield amplitude ratio.
ization of the P100 waveform is most common with
Additionally, the ratio of the P100 waveform ampli-
hemifield stimulation, it can be highly variable and
tude of the temporal and nasal hemifields of the
in some cases, may be better seen over the stimu-
same eye is determined and is called the monocu-
lated hemisphere.
lar hemifield amplitude ratio. Because of the vari-
In addition to the N75, P100, and N145 wave-
ability of these responses, normative data have not
forms that are best seen contralateral to the acti-
been established for these measurements in most
vated visual cortex, three additional waveforms
laboratories, including the author’s.
are seen ipsilateral to the side of activation. These
are the P75, N105, and P135 waveforms, which are Abnormalities
best seen in the lateral occipital and temporal leads. When interpreting PRVEP with hemifield stimula-
There is considerable overlap of these waveforms tion, it should be recognized that many variables are
with the N75, P100, and the N145 waveforms, espe- being analyzed. This increases the risk of overint-
cially over the MO lead, and the two sets of wave- erpretation. To guard against this, as noted earlier,
forms must be clearly differentiated. This N105 many laboratories interpret the studies as simply the
P100 waveform being present or absent in the lateral lesion. It should be remembered that whereas par-
derivations (LO–MF, RO–MF, LT–MF, RT–MF). If adoxical lateralization of the P100 waveform after
latency and amplitude are measured, very stringent hemifield stimulation is most common, the P100
criteria for abnormality should be used. Additionally, waveform can sometimes be best seen ipsilateral to
findings should be internally consistent. If a latency the side of the activated visual cortex, making inter-
prolongation is noted with one hemifield stimu- pretation more difficult.
lation, a similar abnormality should be noted with
the appropriate hemifield stimulation of the con-
tralateral eye. In the examples shown in this book,
Flash VEP
the PRVEPs with hemifield stimulation are inter- FVEPs are obtained with a diffuse light flash. They
preted as P100 waveform being present or absent. are much less sensitive for detecting abnormalities
Figure 2.12A and B illustrate the expected hemi- of the visual pathways than PRVEP. There are two
field VEP findings depending on the site of the main reasons why FVEPs are used: (a) in patients
lesion. Only if hemifield stimulation confirms unable to cooperate with maintaining fixation on a
interhemispheric amplitude ratio abnormalities or checkerboard pattern (i.e., very young or cognitively
latency prolongations seen with monocular full- impaired patients), and (b) individuals with such
field stimulation, is the study is considered abnor- poor refraction that they cannot distinguish between
mal and suggestive of chiasmal or postchiasmal the light and dark checks of the checkerboard
Left Right
Left Right Left Right
hemifield hemifield hemifield hemifield
A
B
Occipital cortex
Left Right
LT LO MO RO RT

(A) (B)
LE, left eye; LHF, left hemifield; LT, left temporal; LO, left occipital; MO, mid occipital; RO, right occipital; RE, right eye; RHF, right
hemifield; RT, right temporal; V, higher amplitude P100 waveform; v, lower amplitude P100 waveform.
Figure 2.12 (A and B) A cartoon of the visual pathway is shown, indicating the pathways stimulated when various hemifields are
activated (A). For example, when the left hemifield of the left eye is stimulated (black), the nasal retina is activated. The impulses
travel through the left optic nerve and chiasm to the contralateral side and the right optic tract. These fibers eventually relay in the
RO cortex. The P100 waveform generated by stimulating this pathway will be best seen in the LO and LT electrodes (paradoxical
lateralization). The electrodes in which the P100 waveform will be seen with each hemifield stimulation are shown in the table (B).
pattern. In the latter group of patients, a PRVEP on level of alertness. This makes interpretation
is attempted with 60-minute check size, and if a based on latency and amplitude very difficult. Most
response cannot be recorded, FVEPs are attempted. laboratories interpret a present FVEP as normal and
an absent response as abnormal. If a robust response
Methodology is noted, latency and amplitude can be compared
The flash stimulus consists of a light without color after left and right eye stimulation. Asymmetries
or pattern that is generated from light emitting of latency and amplitude between eyes should be
diode (LED) goggles or lamp. Goggles are preferred interpreted with extreme caution.
as they are placed directly over the patient’s eyes. When a FVEP is present, the only conclusion that
The eyes can be open or closed during stimulation. can be made is that at least some part of the visual
The field of stimulation is very large, and changes pathway between the retina and visual cortices is
in the patient’s direction of gaze have no effect on intact. A present FVEP certainly does not equate to
the response. Thus, a response can be obtained even normal visual acuity. Consequently, FVEP provides
in patients with limited ability to participate with only a very rudimentary assessment of the visual
testing. If an LED lamp is used for stimulation, it pathway. If technical problems can be excluded, an
should be placed 35 to 45 cm in front of the patient. absent FVEP implies lack of functional integrity of
The patient’s eyes should be open and gaze directed the visual pathways between the retina and occipi-
toward the lamp. Ambient light should be kept low, tal cortices (9, 10). In this situation, an electroretino-
but enough to allow observation of the patient by gram (ERG) can be helpful in determining if the
the technologist. Stimulation is best performed with abnormality is in the retina or more proximally in
one eye at a time. The nonstimulated eye should the visual pathway. If an ERG is present, the lesion
be covered to exclude all light to avoid uninten- is likely in the visual pathway and if it is absent, the
tional stimulation. The rate of stimulation should be problem is most likely ocular. A detailed discussion
approximately 1 to 2/s, avoiding integers of 60. Even of ERG is beyond the scope of this text.
slower rates may be necessary in very young infants.
FVEPs are best recorded from a midline occipi- Pediatric Considerations
tal electrode. A single channel with a MO-reference
derivation is sufficient to record the VEP. The refer- VEP is a useful test for an evaluation of the entire
ence is often AU. In some laboratories, if a response visual system in children. Whereas older children
is not noted, other derivation, such as LO-reference, can cooperate with visual examination, the young
RO-reference, and Vertex-reference are used. In the child may not be able to do so. In this situation, the
author’s laboratory, only a single derivation is used VEP may provide clues regarding the child’s vision.
for recording FVEP. Amplifier settings are the same In addition to appreciating the technical and inter-
as for PRVEP. Each response must be replicated at pretation differences between children and adults,
least once. maturational changes that occur in the VEP as the
child ages must also be considered.
Interpretation
There are up to seven peaks seen in FVEP in the first Maturational Issues
250 ms after stimulation. The first three peaks are Maturational changes in PRVEP and FVEP occur at
best recorded over the midline occipital electrode, different rates but follow similar trends. More data
while later peaks are better seen over the vertex are available for FVEP as these can be obtained more
electrode. Typically, the first major positive peak easily in the very young child since patient cooper-
occurs between 50 to 100 ms after stimulation, and is ation is not needed. FVEP can first be recorded as a
referred to as the P1 or P100 waveform (Figure 2.13). large negative peak at about 300 ms at age 22 weeks
Preceding this peak is a smaller negative peak, conceptional age (CA). A few weeks later, a positive
called the N1 waveform. After the P1 waveform peak can be seen occurring before the large negative
is another, larger negative peak (N2) occurring peak. This positive peak shortens in latency to about
between 100 to 250 ms after stimulation (8). Peaks 200 ms by term. By 44 weeks CA, this positive peak
occurring thereafter are less reliably present. becomes bifid, with the earlier peak occurring at
All peaks of the FVEP are highly variable within about 100 ms. By 6 months of age, the positive wave-
and between individuals and also very dependent form at 100 ms, known as the P1 or P100 waveform,
Age: 1 year Visual acuity: N/A

Sex: Male Corrective lenses: N/A
Eye: Both Filters: 1–100 Hz
Stimulation rate: 2.1/s Preauricular–preauricular distance: 33 cm
Number of repetitions: 200 Scale: Amplitude = 2 μV/div; Latency = 25 ms/div
(+↓)
P1 Waveform
OS Present
OD Present
Discussion: The first major positive peak obtained with FVEP typically occurs at a latency of about 100 ms. In this example, the P1
waveform occurs at a latency of about 110 ms after stimulation of the left and right eyes. Left eye stimulation is displayed above
right eye stimulation. Because of the variability of the response with FVEP, if the P1 waveform is present, the study is considered
normal. It is because of this that latencies of the P1 waveforms are not given above; rather simply “Present” is noted.
FVEP, flash visual evoked potential; VEP, visual evoked potential.
Figure 2.13 Normal flash VEP.
becomes the dominant waveform and will change saline jelly and nonirritating tape with minimum
little with further maturation (11). PRVEP follow a abrading. At least the midline occipital (MO) and
similar maturation pattern, and a dominant P100 reference electrodes (MF or AU) are placed, and if
waveform is achieved by 6 to 8 months of age (12). possible, lateral occipital (LO and RO) electrodes are
also applied. The MO–MF derivation is preferred.
FVEP can be obtained at birth and thereafter as
Methodology patient cooperation is not necessary. PRVEP has
In the older child, PRVEP can be obtained with the been obtained as early as 4 to 6 weeks after term and
same methodology as in adults. In younger children can reliably be obtained by 6 months of age. The size
and neonates, electrodes should be applied with of checks used to elicit the PRVEP depends on the
age of the child; less than 2 months—120 minutes, the site of the lesion in the visual pathways, the
2 to 3 months—60 minutes, 3 to 5 months—30 abnormalities are generally not specific to a particu-
minutes, and 15 minutes thereafter. If a response lar disease state (with a few exceptions). In this sec-
is not obtained, the check size should be doubled. tion an overview of the types of abnormalities that
Maintaining the child’s attention on the checker- occur will be discussed. Examples of VEP in various
board pattern can be challenging and having a car- conditions will be presented later in the chapter.
toon playing on the screen superimposed on the
checks may help. Binocular stimulation is used for Prechiasmatic Disorders
children less than 6 months of age, and the stim-
ulation rate is no faster than 1/s. Other amplifier PRVEPs are used most often in detecting lesions in
parameters are the same as for adults (13, 14). the prechiasmatic visual pathway.
Optic neuritis/multiple sclerosis
Interpretation VEPs are most useful in detecting optic nerve
Many of the interpretation principles used in adults lesions. Acute demyelinating conditions, such as
can also be used for children. As with adults, optic neuritis, result in loss of the VEP when tested
the latency and amplitude of the P1 waveform during the period of vision loss. As vision returns,
obtained with FVEP is very variable. Consequently, the amplitude of the P100 waveform gradually
most laboratories interpret a present waveform as increases, but the latency remains prolonged. After
normal and an absent one as abnormal. A normal optic neuritis has recovered and vision returns to
FVEP implies that at least some visual stimuli are near normal, as may be seen in patients with multi-
reaching the occipital cortices from the retina. This, ple sclerosis, the P100 waveform is of normal ampli-
however, does not necessarily mean that visual per- tude but greatly prolonged latency. This relatively
ception is occurring. On the other hand, an absent normal morphology of the P100 waveform with
FVEP implies no visual stimuli are reaching the long latency is very suggestive of a demyelinating
occipital cortices and visual perception is unlikely. process affecting the optic nerve (15).
When the FVEP is absent, an ERG may be helpful in
Tumors
further localizing the pathology to the eye or visual
pathway, as discussed previously. Tumors can also affect the optic nerve. The most
In pediatric PRVEP, the latency of the P100 wave- common abnormality is a loss of amplitude and
form is the main parameter measured. Clinical a mild prolongation of latency. Latency prolon-
significance of amplitude, topography, and mor- gations similar to those seen with demyelinating
phology changes is uncertain. Interpretation of processes are rare with tumors. With more severe
PRVEP is dependent on the child’s age. Ideally each involvement of the optic nerve with a tumor, the
laboratory should establish its own normative data, VEP may be completely abolished (16).
but this can be very difficult for pediatric evoked Neuropathies
potentials. Published normative data for the latency Various types of neuropathies can also affect the
of the P100 waveform is available, but if these data VEP. Anterior ischemic optic neuropathy (AION)
are used, the same methodology should be used as in affects the VEP in a manner similar to tumors; the
the reference laboratory. Normative data are based amplitude is affected more than the latency (17).
on not only age but also check size and whether bin- The morphology of the VEP may also be distorted
ocular or monocular stimulation was used. with AION. Other types of neuropathy that are pre-
dominantly axonal affect the VEP the same way as
Clinical Correlates AION. Demyelinating neuropathies cause a more
significant prolongation of the latency.
VEPs are usually performed for detection of neuro
logic problems, but both neurologic and ophthal- Metabolic disorders
mologic disorders can affect the response. Various Several metabolic disorders, especially the leuko-
ophthalmologic tests can detect disorders of the eye dystrophies, can affect the VEP. In most cases FVEP
with greater accuracy. Among neurologic disorders, has been studied as patient cooperation with PRVEP
VEPs are most often used for detection of prechias- is often compromised. FVEP may be normal early in
matic lesions. Though VEP can be used to determine the disease course, but in more advanced cases of
diseases such as Krabbe disease, there is loss of the latency may or may not be prolonged. An “uncrossed
FVEP (18, 19). Whereas these changes may be due to asymmetry” of the parasagittal P100 waveform with
bilateral prechiasmatic lesions, they are more often monocular full-field stimulation is noted with these
due to bilateral postchiasmatic lesions. Other met- lesions. Such findings should be confirmed with bin-
abolic disease can similarly affect the VEP (20, 21). ocular hemifield PRVEP.
Neuropathies and metabolic disorders that affect
both optic nerves cause prolongation of the P100 Disorders with symmetric lesions
waveform after monocular stimulation of both Neurodegenerative conditions like Alzheimer’s dis-
eyes. Usually there is no parasagittal asymmetry of ease and Parkinson’s disease and metabolic diseases
the P100 waveform in these situations, suggesting a such as leukodystrophies can also affect the post-
bilateral prechiasmatic localization. However, such chiasmatic visual pathway in a bilateral symmetric
a pattern may also be seen with bilateral postchias- manner (23). While these conditions predominantly
matic lesions. decrease the amplitude and may also prolong the
latency of the P100 waveform, the abnormalities are
Hysterical blindness usually not lateralized. Consequently, the topog-
Patients with hysterical blindness often have nor- raphy of the P100 waveform may not be affected
mal PRVEP (22). However, these patients, and those and only a mild prolongation of latency noted
that are malingering, can intentionally not focus on with monocular stimulation. Bilateral, symmet-
the checkerboard pattern and cause amplitude and ric involvement of both hemispheres accounts for
latency changes. An alert technologist can often these findings. This may lead to erroneous localiza-
detect the lack of cooperation by such patients. tion of the lesion to bilateral prechiasmatic.
Chiasmatic Disorders Ophthalmologic Disorders

Lesions in the vicinity of the sella turcica can cause VEPs are not routinely used to detect ophthalmo-
compression of the optic chiasm and produce VEP logic disorders. However, these disorders may coex-
abnormalities. Common disorders that may cause ist with neurologic disorders for which VEPs are
such changes include pituitary tumors and cranio- performed and can affect the VEP. Ophthalmologic
pharyngiomas. Compression of the optic chiasm disorders affect amplitude of the P100 waveform
causes distortion of the topography and a loss of more than the latency.
amplitude of the VEP. The latency may or may not Refractive errors affect visual acuity and can
be prolonged. When a “crossed asymmetry” of the result in loss of amplitude and prolongation of
parasagittal P100 waveform with monocular full-field latency of the P100 waveform. Various types of
stimulation is noted, a chiasmatic lesion is suspected. amblyopia also cause similar VEP abnormalities.
PRVEP with monocular hemifield stimulation are Disorders of the anterior chamber of the eye can
needed for confirmation when this pattern is seen. result in impaired vision and degradation of the
VEP. Glaucoma affects the optic nerve and causes
Postchiasmatic Disorders changes in the VEP. Similarly, various types of ret-
Many neurological disorders can cause postchias- inopathy can also affect the VEP. Since it is mostly
matic visual pathway dysfunction. There are too the central retina that contributes to the VEP, it is
many disorders to list individually, but they can only when this part of the retina is involved in a
be grouped depending on whether the lesions they retinopathy that the VEP is affected.
produce are symmetric or asymmetric.
Disorders with asymmetric lesions Examples
Any disease process affecting visual pathways in the In this section examples of VEP are presented
temporal, parietal, and occipital lobes can cause an (Figures 2.14–2.31, Examples 1-18). Technical issues
abnormality of the VEP. Common conditions caus- and various types of abnormalities are highlighted.
ing this include strokes, tumors, and demyelinating Since VEP abnormalities are not specific to a disease
lesions. Similar to chiasmatic lesions, these lesions state, the abnormalities will be highlighted and inter-
predominantly cause a change in topography and pretation discussed. Whenever possible, the abnor-
reduction of the P100 waveform amplitude. The mality will be correlated with neuroimaging findings.
Age: 52 years
Sex: Female
Filters: 1–100 Hz
Preauricular–preauricular distance: 34 cm (+↓)
(A)
Eye: Left P100 Amplitude Ratio
Visual angle: 30′ Location Ratio
Visual acuity: 20/50 OS–OD 1.09
LO–RO 1.60

MO–AU 73.0 98.5 127.0 MO–MF 14.6 19.7 17.2
Oz–AU 73.0 98.5 127.0 MO–AU 9.1 14.2 11.6
MF–AU 101.0 Oz–AU 11.1 18.6 14.9
LO–MF 76.0 99.5 129.0 LO–MF 10.6 12.7 11.7
RO–MF 77.5 101.0 128.0 RO–MF 6.1 8.4 7.3
Figure 2.14 (A and B) Example 1. (continued )

(+↓)
Absolute Latency
MO–MF 73.0 99.0 129.0
MO–AU 73.5 97.0 126.0
Oz–AU 73.5 97.0 126.0
MF–AU 102.0
LO–MF 74.5 100.0 130.0
RO–MF 74.0 99.0 127.0
Amplitude
N75–P100 P100–N145
MO–MF 14.3 23.3 18.8
MO–AU 7.0 15.1 11.1
Oz–AU 10.2 20.0 15.1
LO–MF 11.0 15.4 13.2
RO–MF 6.9 11.7 9.3
(B)
Eye: Right
History: This is a 52-year-old woman with a history of vascular headaches and a remote history of questionable optic neuritis. MRI
of the brain was normal. This study is being done to rule out multiple sclerosis. No medications are listed.
Discussion: This is a normal VEP study. It shows high amplitude, reproducible VEP responses after monocular stimulation of the left
(A) and right (B) eyes. The responses are correctly tagged by the technologist. Based on the patient’s preauricular to preauricular
distance, the upper limit of normal for the P100 waveform latency is 111.0 ms. This patient’s P100 waveforms after left and right
eye stimulation are within normal limits at 99.0 ms after stimulation of either eye.
The P100 waveform latency and amplitude are usually measured from the MO–MF derivation, as this derivation typically has the
greatest amplitude. This is because the MO–MF derivation takes advantage of the N105 waveform which is seen in the frontal
regions. The parasagittal waveforms are symmetric as denoted by their amplitude ratios, and the amplitude ratio of the P100 wave-
form between left and right eye stimulation is also normal at 1.09. This study does not confirm a prior history of optic neuritis, and
by extension, does not support a diagnosis of multiple sclerosis.

Age: 38 years
Sex: Female
(A)
Eye: Left
Visual angle: 30′
Absolute Latency
MO–MF 79.0 107.0 154.0
MO–AU 80.0 107.0 150.0
Oz–AU 80.5 107.0 150.0
MF–AU 112.0
LO–MF 79.5 108.0 157.0
RO–MF 80.5 111.0 158.0
Amplitude
N75–P100 P100–N145 Mean
Derivation (μV) (μV) (μV)
MO–MF 17.6 26.5 22.1
MO–AU 15.5 20.9 18.2
Oz–AU 17.4 26.3 21.9
LO–MF 7.0 12.6 9.8
RO–MF 8.2 12.3 10.3

Location Ratio
OS–OD 1.02
LO–RO 1.05
Figure 2.15 (A–D) Example 2. (continued )

(B)
Eye: Right
Absolute Latency
MO–MF 78.5 106.0 146.0
MO–AU 79.0 106.0 144.0
Oz–AU 79.0 107.0 142.0
MF–AU 106.0
LO–MF 79.0 107.0 156.0
RO–MF 79.5 109.0 156.5
Amplitude
Derivation N75–P100 P100–N145 Mean (μV)
(μV) (μV)
MO–MF 19.40 24.0 21.70
MO–AU 16.80 19.0 17.90
Oz–AU 19.10 23.6 21.40
LO–MF 8.40 10.4 9.41
RO–MF 8.76 12.0 10.40

Location Ratio
OS–OD 1.02
LO–RO 1.18
History: This is a 38-year-old patient who is undergoing a visual
evoked potential study to evaluate for visual pathway dysfunction.
She has a history of subtle white matter changes on MRI and is sus-
pected of having a demyelinating disease.
Discussion: This is a normal VEP study. It shows high amplitude,

reproducible VEP responses after monocular stimulation of the
left (A) and right (B) eyes. The responses are correctly tagged by
the technologist. Based on the patient’s preauricular to preau-
ricular distance, the upper limit of normal for the P100 waveform
latency is 111.0 ms. This patient’s P100 waveforms after left and
right eye stimulation are within normal limits at 107.0 and 106.0
ms, respectively. These measurements are obtained in the MO–MF
derivation. The P100 waveform in this derivation is not of substan-
tially higher amplitude than in MO–AU or Oz–AU derivations. This
is because the N105 waveform is of small amplitude in the MF–AU
derivation, and the contribution of the N105 waveform negativity
to the P100 waveform in the MO–MF derivation is negligible. Thus, (C) (D)
the P100 waveform amplitude is similar in all three derivations. The
parasagittal waveforms are symmetric with normal amplitude ratios. The MRI abnormalities are suggestive but not diagnostic of demye-
linating disease (multiple sclerosis). If the PRVEP had revealed a defi-
The brain MRI shows areas of signal abnormality in the posterior part nite lesion, it may have provided support for this diagnosis. However,
of the corpus callosum (C; T2 image) and periventricular white matter because this study is normal, it does not add to the diagnostic certainty
(D; FLAIR image). of multiple sclerosis.
FLAIR, fluid-attenuated inversion recovery; PRVEP, pattern reversal visual evoked potential; VEP, visual evoked potential.

Age: 48 years
Sex: Female
Preauricular–preauricular distance: 36
(A)
Eye: Left
Visual angle: 30′
Absolute Latency
MO–MF 84.0 109.0 135.0
MO–AU 84.0 107.0 135.0
Oz–AU 84.0 108.0 135.0
MF–AU 108.0
LO–MF 84.0 109.0 134.0
RO–MF 84.5 110.0 133.0
Amplitude
N75–P100 P100–N145
MO–MF 4.1 5.6 4.8
MO–AU 3.2 4.7
Oz–AU 3.3 4.6
LO–MF 3.0 3.0 3.0
RO–MF 2.7 2.0 2.3

Location Ratio
OS–OD 1.2
LO–RO 1.3

(B)
Eye: Right

Location Ratio
OS–OD 1.2
LO–RO 1.3

MO–AU 82.5 107.0 144.0 MO–MF 4.4 6.9 5.7
Oz–AU 82.5 106.0 144.0 MO–AU 3.3 6.5
MF–AU 121.0 Oz–AU 3.2 6.6
LO–MF 80.5 114.0 142.0 LO–MF 3.1 4.9 4.0
RO–MF 82.5 112.0 140.0 RO–MF 2.6 3.7 3.1
History: This is a 48-year-old patient who is complaining of decreased vision in her right eye. She denies any pain in her eye. Her
MRI scan is normal.
Discussion: This is a normal PRVEP study. It shows reproducible but noisy VEP responses after monocular stimulation of the left
(A) and right (B) eyes. The responses are correctly tagged by the technologist. Based on the patient’s preauricular to preauricular
distance, the upper limit of normal for the P100 waveform latency is 113.5 ms. This patient’s P100 waveforms after left and right
eye stimulation are within normal limits at 109.0 ms. The responses have considerable artifact, but they are clearly identifiable. The
parasagittal waveforms are also symmetric. Despite the artifact this study is normal and does not explain the patient’s symptoms.

Age: 71 years
Sex: Male
Filters: 1–100 Hz (+↓))
(A) P100 Amplitude Ratio

Eye: Left
LO–RO 1.21

MO–AU 91.5 115.0 141.0 MO–MF 5.6 7.4 6.5
Oz–AU 91.5 115.0 141.0 MO–AU 4.5 4.8 4.7
MF–AU 123.0 Oz–AU 4.9 5.9 5.4
LO–MF 91.0 123.0 152.0 LO–MF 3.4 4.8 4.1
RO–MF 98.0 126.0 154.0 RO–MF 3.1 3.7 3.4

(B)
Eye: Right
Absolute Latency
MO–MF 88.5 114.0 144.0
MO–AU 89.0 116.0 140.0
Oz–AU 89.0 116.0 140.0
MF–AU 126.0
LO–MF 89.0 117.0 153.0
RO–MF 89.5 118.0 154.0
Amplitude
Derivation N75–P100 (μV) P100–N145 (μV) Mean (μV)
MO–MF 6.0 6.6 6.3
MO–AU 5.2 4.8 5.0
Oz–AU 5.8 6.0 5.9

LO–MF 4.0 5.6 4.8
RO–MF 2.7 2.2 2.5

Location Ratio
OS–OD 1.03
LO–RO 1.92
Even though the P100 waveforms in this PRVEP appear

prolonged at first glance, when analyzed with respect to
head size, this study is normal. Whereas the MRI findings
History: This is a 71-year-old male with gait difficulty who is may help explain the patient’s symptoms, the PRVEP does
complaining of deteriorating vision in his left right eye. His MRI not extend the lesions to the visual pathways.
shows multiple white matter lesions.
Discussion: This is a normal PRVEP study. The P100

waveform after left (A) and right (B) eye stimulation
are 115.0 ms and 114.0 ms, respectively. At first glance
these latencies suggest bilateral prolongation of the
P100 waveform, but the preauricular to preauricular
distance is 38 cm. The normal P100 waveform latency
for this head size is 116 ms. It is important to use
normative data that accounts for head size; alternatively
different norms for males and females can also be used.
MRI of the brain reveals multiple areas of signal

abnormality in the left temporal area (C; T2 image) and
subcortical white matter bilaterally (D; T2 image).
(C) (D)

Age: 25 years
Sex: Female
Filters: 1–100 Hz
(A)
Eye: Left
Visual angle: 30′
Absolute Latency
MO–MF 70.8 109.8 136.2
MO–AU
In–AU 76.5 105.3 140.1
MF–AU 108.6
LO–MF 71.1 108.6 136.2
RO–MF 71.7 112.2 131.7
Amplitude
MO–MF 6.1 8.5 7.3
MO–AU
In–AU 10.5 21.2 15.9
LO–MF 2.3 4.7 3.5
RO–MF 4.6 5.7 5.2

Location Ratio
OS–OD 1.2
LO–RO 1.5

(B)
Eye: Right
Visual angle: 30′
Absolute Latency
MO–MF 85.5 114.3 151.5
MO–AU 123.0
In–AU 82.2 106.2 150.3
MF–AU 110.4
LO–MF
RO–MF 83.4 115.2 151.5
Amplitude
N75–P100 P100–N145
History: This is a 25-year-old female with recurrent left sided numb-
ness and slurred speech. MRI of brain shows a contrast enhancing
right sided splenium of the corpus callosum lesion and other areas of MO–MF 3.5 9.1 6.3
abnormality. This study is being done to evaluate for multiple sclerosis.
MO–AU
Discussion: This is a normal PRVEP study. The P100 waveforms
In–AU 9.5 23.2 16.4
after left and right eye stimulation are 105.3 and 106.2 ms,
respectively. Notice that the MO–MF and MO–AU derivations do LO–MF
not show as well-defined P100 waveforms as the In–AU deriva-
tion. The latter has a much higher amplitude and better mor-
RO–MF 4.1 6.3 5.2
phology. As discussed previously, sometimes the P100 waveform
vector projects such that the highest amplitude is recorded more
posteriorly than the MO electrode. Instead of the Oz position, the P100 Waveform Amplitude Ratio
In position is used in this case as that is more posterior and inferior Location Ratio
than Oz. If the In–AU derivation was not recorded, the PRVEP after
right sided stimulation may have been erroneously called abnor- OS–OD 1.2
mal as the latency of the P100 waveform in the MO–AU deriva- LO–RO N/A
tion is 123.0 ms. Notice that some of the values in the tables are
missing; this is because latencies and amplitudes of some wave-
forms that are not clearly reproducible are not listed. The parasag-
ittal waveforms are also poorly formed. Recall that the LO and
RO electrodes are placed in the same horizontal plane as the MO
electrode. Had they been moved to the In electrode plane, better
morphology parasagittal waveforms could have been recorded.
The brain MRI shows a contrast enhancing lesion on the right side
of the splenium of the corpus callosum (C; T1 with contrast image).
There are multiple other white matter lesions as well (D; FLAIR image).
The brain MRI in this patient is very suggestive of multiple sclero-

sis but a definite diagnosis could not be reached by the clinician,
which is why the PRVEP was performed. If the In–AU derivation
was not recorded in this study, it could have been misinterpreted
and thought to help confirm the diagnosis suspected. However,
with analysis of the In–AU derivation, the PRVEP is normal and
does not help confirm the diagnosis of multiple sclerosis.
FLAIR, fluid-attenuated inversion recovery; PRVEP, pattern

reversal visual evoked potential. (C) (D)

Age: 54 years
Sex: Female
Filters: 1–100 Hz
(+↓)
(A)
Eye: Left P100 Amplitude Ratio
Visual angle: 30′

MO–AU 98.0 125.0 151.0 MO–MF 3.0 3.5 3.3
Oz–AU 99.5 126.0 152.0 MO–AU 2.3 3.6 3.0
MF–AU Oz–AU 3.1 4.0 3.6
LO–MF 118.0 137.0 156.0 LO–MF 5.1 2.4 3.7
RO–MF 104.0 125.0 149.0 RO–MF 2.9 1.7 2.3
Figure 2.19 (A–F) Example 6. (continued )

(+↓)
(B)
Eye: Right P100 Amplitude Ratio
Visual angle: 30′

MO–AU 83.5 115.0 141.0 MO–MF 5.1 6.5 5.8
Oz–AU 84.5 118.0 142.0 MO–AU 2.7 6.2 4.5
MF–AU 120.0 Oz–AU 3.5 6.0 4.8
LO–MF 106.0 131.0 147.0 LO–MF 4.3 1.4 2.9
RO–MF 97.5 121.0 137.0 RO–MF 5.9 3.3 4.6
Figure 2.19 (A–F) (continued )

(+↓)
(C)
Eye: Left
Visual angle: 60′

MO–AU 82.5 115.0 154.0 MO–MF 7.7 7.1 7.4
Oz–AU 85.5 120.0 154.0 MO–AU 7.5 7.2 7.4
MF–AU Oz–AU 7.7 7.3 7.5
LO–MF 88.5 115.0 134.0 LO–MF 5.6 2.5 4.0
RO–MF 91.0 123.0 153.0 RO–MF 6.1 2.3 4.2

Location Ratio
OS–OD 1.1
LO–RO 1.0

(D)
Eye: Right
(+↓) Number of repetitions: 200
Visual angle: 60′
Absolute Latency
MO–MF 79.0 110.0 148.0
MO–AU 79.0 108.0 147.0
Oz–AU 80.0 114.0 149.0
MF–AU
LO–MF 88.0 115.0 138.0
RO–MF 84.0 123.0 150.0
Amplitude
N75–P100 P100–N145
MO–MF 9.1 7.3 8.2
History: This is a 54-year-old female with a history of multiple MO–AU 8.5 7.8 8.2
sclerosis complaining of visual deterioration. She recalls that her Oz–AU 10.0 5.8 7.9
visual problems predated her diagnosis of multiple sclerosis and
is worried that some of her visual problems might be due to her LO–MF 6.2 3.9 5.0
underlying disease. RO–MF 6.7 3.0 4.9
Discussion: This is a normal PRVEP study. The P100 waveforms
after left (A) and right (B) eye stimulation with 30′ check size
are prolonged at 124.0 and 117.0 ms, respectively. However, the
visual acuity, despite the use of corrective lenses is 20/400 and P100 Waveform Amplitude Ratio
20/100 on the left and right sides, respectively. When a larger
Location Ratio
check size is used (60′) to compensate for the poor visual acuity,
the P100 waveform latencies normalize to 115.0 and 110.0 ms OS–OD 1.1
(upper limit of normal for this head circumference is 116.0) (C, D).
LO–RO 1.0
The MRI of her brain shows abnormal signal on the left side of
the splenium of the corpus callosum (E; FLAIR image) as well as
other areas of abnormal signal in subcortical white matter. MRI
of the cervical and thoracic spine shows an area of increased T2
signal in the thoracic spine (F; T2 image).
This patient has been diagnosed as having multiple sclerosis but

does not have definite involvement of the visual pathways. If this
PRVEP had not been done with the larger check size, it might
have been interpreted as being abnormal, consistent with bilat-
eral optic nerve lesions. However, the larger check size compen-
sated for the poor visual acuity, and the P100 waveforms were
recorded at a normal latency. This suggests that the visual dete-
rioration is not due to optic nerve demyelination as may occur in
multiple sclerosis but likely due to other causes.
FLAIR, fluid-attenuated inversion recovery; PRVEP, pattern (E) (F)
reversal visual evoked potential.

Age: 66 years
Sex: Female
Filters: 1–100 Hz (+)
(A)
Eye: Left P100 Waveform Amplitude Ratio
Visual angle: 30′

MO–AU 75.5 121.0 180.0 MO–MF 5.0 6.6 5.8
Oz–AU 82.0 121.0 180.0 MO–AU 2.5 6.0 4.3
MF–AU 126.0 Oz–AU 2.1 5.0 3.6
LO–MF 75.5 121.0 172.0 LO–MF 4.3 3.0 3.6
RO–MF 75.5 121.0 172.0 RO–MF 5.6 5.1 5.3
Figure 2.20 (A–E) Example 7. (continued )

↓(+)
(B)
Eye: Right P100 Waveform Amplitude Ratio
Visual angle: 30′
OS–OD 1.5

MO–AU 85.5 125.0 169.0 MO–MF 4.9 6.6 5.8
Oz–AU 86.5 125.0 169.0 MO–AU 5.6 5.0 5.3
MF–AU 130.0 Oz–AU 6.4 5.3 5.8
LO–MF 81.0 122.0 166.0 LO–MF 3.5 2.0 2.7
RO–MF 79.5 122.0 164.0 RO–MF 5.3 2.7 4.0

OS
↓(+)
OD
(C)
Visual evoked potential (flash)
Eye: Both
Number of repetitions:
Visual acuity: N/A
Corrective lenses: N/A
Filters: 1–100 Hz
P1 Waveform
OS Present
OD Present


(D) (E)
History: This is a 66-year-old female undergoing evaluation for rapidly progressive dementia. She is sleepy and is unable to maintain
focus on the PRVEP stimulus screen.
Discussion: The PRVEP study is uninterpretable. The P100 waveforms after left (A) and right (B) eye stimulation appear to be
prolonged at 123.0 ms and 128.0 ms, respectively. However, the patient was sleepy during this study and was unable to focus on
the stimulus. Note that the PRVEP after right eye stimulation only has one repetition; due to the patient’s inability to focus, the
technologist aborted the study without replicating data. These factors make it very likely that the PRVEP is unreliable. Additionally,
her visual acuity was 20/400 bilaterally, which could also affect the P100 waveform latency. A FVEP was also performed (C), which
does not require patient cooperation. It was normal, with responses present after bilateral stimulation.
MRI of brain shows mild atrophy (D; T1 image) thought to be consistent with age and no other lesions were seen (E; T2 image).
The technologist must be vigilant and make certain that the patient is attending to the stimulus. In this case, had the technologist
not noted that the patient was drowsy and unable to concentrate on the stimulus, the test may have been stopped after the PRVEP
were obtained and no comment made by the technologist about patient cooperation. The study may then have been interpreted
as being abnormal, indicating visual pathway lesions. However, since the technologist detected problems with patient participation
with the PRVEP, FVEP were performed. An abnormal FVEP would indicate a severe visual pathway lesion; a normal test does not
rule out a mild lesion. Since the latter test was normal, it is impossible to interpret this study. Certainly the PRVEP may be abnormal,
however, it should not be interpreted as such since these changes may be seen when the patient does not attend to the stimulus.
A normal MRI also does not provide evidence of a lesion in the visual pathways. The PRVEP should be repeated when the patient is
able to participate more fully with testing.
FVEP, flash visual evoked potential; PRVEP, pattern reversal visual evoked potential.

Age: 46 years
Sex: Female
Filters: 1–100 Hz
for (A) and (B)
Scale: Amplitude = 2 μV/div; Latency = 25 ms/div for (C) and (D)
(A)
Eye: Left
Visual angle: 60′

MO–AU 86.5 104.0 128.0 MO–MF 4.1 10.1 7.12
Oz–AU 87.0 105.0 128.0 MO–AU 4.3 9.2 6.80
MF–AU 100.0 Oz–AU 4.2 9.1 6.70
LO–MF 84.0 99.5 122.0 LO–MF 3.2 9.9 6.60
RO–MF 115.0 148.0 173.0 RO–MF 7.5 3.4 5.50

Location Ratio
OS–OD 1.5
LO–RO 1.2

(+↓)
(B)
Eye: Right
Visual angle: 60′

MO–AU 84.5 102.0 129.0 MO–MF 7.1 13.9 10.5
Oz–AU 84.5 103.0 128.0 MO–AU 5.7 12.9 9.3
MF–AU 98.5 Oz–AU 5.6 12.8 9.2
LO–MF 84.0 101.0 126.0 LO–MF 6.0 10.4 8.2
RO–MF RO–MF

Location Ratio
OS–OD 1.5
LO–RO

(+↓)
(C)
Eye: Left
Visual angle: 30′

MO–AU 82.0 103.0 124.0 MO–MF 5.2 6.1 5.7
Oz–AU 82.0 103.0 124.0 MO–AU 4.3 4.7 4.5
MF–AU Oz–AU 4.4 4.8 4.6
LO–MF LO–MF
RO–MF 86.0 129.0 162.0 RO–MF 3.9 3.2 3.3

Location Ratio
OS–OD 1.0
LO–RO

(D)
Eye: Right
Visual angle: 30′
Absolute Latency
MO–MF 77.0 102.0 126.0
MO–AU 79.0 102.0 126.0
Oz–AU 79.0 102.0 126.0
MF–AU 98.5
LO–MF
RO–MF 78.0 125.0 170.0
Amplitude
N75–P100 P100–N145
MO–MF 5.7 6.0 5.9
MO–AU 4.2 4.9 4.6
Oz–AU 4.3 5.0 4.7
History: This is a 46-year-old female with chronic fatigue and memory
difficulties. She has history of breast cancer and brain MRI shows ques- LO–MF
tionable white matter lesions. RO–MF 2.5 3.6 3.1
Discussion: This is a normal PRVEP study. PRVEP with 60′ check size was
performed first as her visual acuity was impaired and she did not have her
corrective lenses. The morphology of the P100 waveforms after left (A) and
right (B) eye stimulation is unusual, demonstrating a “W-shaped” pattern. P100 Waveform Amplitude Ratio
There are two positive peaks, with the first one occurring at about 75 ms.
Only the second peak is tagged. With a smaller check size (30′), the morphol-
Location Ratio
ogy of the P100 waveform improves, and a much clearer P100 waveform is OS–OD 1.0
evident at 105 ms and 102 ms after left (C) and right (D) eye stimulation,
respectively. There are many reasons that the P100 waveform can appear LO–RO
to have a “W-shape” (see discussion in text). One reason is that peripheral
retinal stimulation with larger check sizes activates slower nerve fibers. This
may have happened in this patient with 60′ check size stimulation. Using a
30′ check size stimulates the foveal fibers more selectively, activating only
the faster conducting fibers and resulting in a better defined P100 wave-
form. However, the latency did not shorten with foveal stimulation in this
patient, which would be expected with selective foveal stimulation. Recall,
though, that the patient’s visual acuity was also impaired, which may have
contributed to lack of latency shortening with the smaller check size.
The brain MRI shows a right (E; FLAIR image) and left (F; FLAIR image)
sided area of signal abnormality that did not enhance with contrast.
The MRI raises the suspicion for a demyelinating condition, even though
the findings are not classic. The PRVEP was done to help confirm this diag-
nosis. Because the PRVEP was normal, it does not support the presence of
a demyelinating process in the visual pathways.
FLAIR, fluid-attenuated inversion recovery; PRVEP, pattern reversal
visual evoked potential. (E) (F)

Age: 30 years
Sex: Female
Filters: 1–100 Hz
(A)
Eye: Left
Visual angle: 30′

Derivation N75 (ms) P100 (ms) N105 (ms) N145 (ms) N75–P100 P100–N145 Mean
MO–MF 80.5 109.0 151.0 Derivation (μV) (μV) (μV)
MO–AU 82.0 105.0 149.0 MO–MF 4.6 4.7 4.6
Oz–AU 81.0 106.0 149.0 MO–AU 3.4 5.3 4.4
MF–AU 112.0 Oz–AU 4.4 5.6 5.0
LO–MF 80.0 113.0 151.0 LO–MF 3.6 3.3 3.4
RO–MF 80.5 109.0 139.0 RO–MF 3.5 3.0 3.2

Location Ratio
OS–OD 1.1
LO–RO 1.1

(B)
Eye: Right
Visual angle: 30′
Absolute Latency
N145
Derivation N75 (ms) P100 (ms) N105 (ms) (ms)
MO–MF 94.0 121.0 164.0
MO–AU 98.0 121.0 155.0
Oz–AU 98.0 121.0 155.0
MF–AU 122.0
LO–MF 88.0 121.0 165.0
RO–MF 91.0 121.0 164.0
Amplitude
N75–P100 P100–N145
MO–MF 2.9 5.7 4.3
History: This is a 30-year-old female with impaired vision in
MO–AU 1.6 4.2 2.9
her right eye for the last month. She has a history of right sided
optic neuritis several years ago that improved with steroids. Oz–AU 2.1 5.0 3.6
Discussion: This is an abnormal PRVEP due to prolongation LO–MF 2.3 4.7 3.5
of the P100 waveforms after right eye stimulation. The P100 RO–MF 2.4 4.1 3.2
waveform is normal at 109 ms after left eye stimulation (A)
but is prolonged at 121 ms after right eye stimulation (B). The
parasagittal waveforms are symmetric, suggesting that the
pathology is prechiasmatic. The morphology of the P100 wave-
form after right eye stimulation is normal, which is typical for Location Ratio
a demyelinating lesion. Other types of optic nerve pathology, OS–OD 1.1
such as ischemia, are more likely to distort waveform morphol-
ogy in addition to causing latency prolongation. The latency LO–RO 1.1
prolongation is usually greater with demyelinating lesions com-
pared to ischemic or compressive lesions.
The MRI of the orbits (C; T1 with contrast image) shows
enhancement of the right optic nerve with administration of
contrast. Brain MRI shows multiple hyperintense lesions in the
periventricular area (D; FLAIR image).
The PRVEP findings are consistent with a lesion of the right sided
prechiasmatic visual pathway, as is seen in optic neuritis. The pre-
served morphology of the waveforms with the prolonged latency
is consistent with this diagnosis. The MRI showing enhancement
of the optic nerve further confirms this impression. The brain MRI
suggests that the diagnosis is likely multiple sclerosis.
FLAIR, fluid-attenuated inversion recovery; PRVEP, pattern reversal
visual evoked potential.
(C) (D)

Age: 60 years
Sex: Male
Filters: 1–100 Hz
(A)
Eye: Left
Visual angle: 30′

MO–AU 81.0 107.0 134.0 MO–MF 6.9 9.7 8.3
Oz–AU 81.0 107.0 134.0 MO–AU 6.3 8.7 7.5
MF–AU 110.0 Oz–AU 4.9 7.4 6.2
LO–MF 81.0 107.0 134.0 LO–MF 4.3 3.7 4.0
RO–MF 84.5 109.0 126.0 RO–MF 5.0 3.9 4.5

Location Ratio
OS–OD 1.4
LO–RO 1.1

(B)
Eye: Right
Visual angle: 30′
Absolute Latency
MO–MF 106.0 130.0 171.0
MO–AU 108.0 132.0 167.0
Oz–AU 108.0 132.0 167.0
MF–AU
LO–MF 108.0 131.0 168.0
RO–MF 102.0 126.0 142.0
Amplitude
N75–P100 P100–N145
MO–MF 4.5 7.2 5.8
MO–AU 5.3 7.8 6.6
Oz–AU 4.1 6.7 5.4
History: This is a 60-year-old male with gait difficulty and spastic LO–MF 4.4 2.7 3.6
lower extremities. He has a history of abnormal brain and spine MRI
and is being evaluated for demyelinating disease. He denies any acute RO–MF 2.9 2.8 2.9
vision problems.
Discussion: This is an abnormal PRVEP due to prolongation of the

P100 waveform after right eye stimulation. The P100 waveform is P100 Waveform Amplitude Ratio
normal at 107 ms after left eye stimulation (A) but is prolonged at
Location Ratio
130 ms after right eye stimulation (B). The parasagittal waveforms
have poor morphology but their symmetry is within normal limits. OS–OD 1.4
These findings suggest that the lesion is in the right prechiasmatic
pathway. The patient’s visual acuity is impaired, and consideration
LO–RO 1.2
should be given to whether that may be causing the latency prolon-
gation. However, the visual acuity of the left eye is worse than the
right eye, and the abnormal P100 waveform latency is seen after right
eye stimulation. Moreover, the study is repeated with 60′ check size
as well, and the results are very similar (not shown). Thus visual acuity
issues are unlikely be responsible for this abnormality.
The brain MRI shows multiple areas of hyperintensity in the white

matter (C; T2 image). The cervical spine MRI shows a hyperintensity
at the C3 level (D; fast spin echo image).
The PRVEP findings are consistent with a lesion of the right sided pre-
chiasmatic visual pathway. Given the extent of the latency prolongation
and the retained morphology of the P100 waveform, the lesion is likely
demyelinating. The MRI shows multiple lesions in the white matter of
the brain and spinal cord, making a demyelinating disease likely.
PRVEP, pattern reversal visual evoked potential. (C) (D)

Age: 47 years
Sex: Female
Stimulation rate: 3.9/s (+)↓
Filters: 1–100 Hz
(A)
Eye: Left
Visual angle: 30′
Absolute Latency
MO–MF 86.5 116.0 157.0
MO–AU 86.0 116.0 157.0
Oz–AU 86.0 116.0 157.0
MF–AU 136.0
LO–MF 86.0 116.0 157.0
RO–MF 86.0 116.0 157.0
Amplitude
MO–MF 10.4 12.0 11.2
MO–AU 9.5 11.5 10.5
Oz–AU 8.2 10.4 9.3
LO–MF 3.7 3.5 3.6
RO–MF 6.1 7.2 6.6

Location Ratio
OS–OD 1.0
LO–RO 1.8
Figure 2.24 (A–C) Example 11. (continued )

(B)
Eye: Right
(+)↓ Number of repetitions: 200
Visual angle: 30′
Absolute Latency
MO–MF 96.5 126.0 166.0
MO–AU 96.5 126.0 166.0
Oz–AU 96.0 126.0 167.0
MF–AU 139.0
LO–MF 96.5 129.0 173.0
RO–MF 96.5 129.0 175.0
Amplitude
MO–MF 8.4 13.5 10.9
MO–AU 8.2 12.9 10.6
Oz–AU 7.5 12.0 9.8
History: This is a 47-year-old female with dyses- LO–MF 5.0 6.5 5.7
thesia of the right upper extremity. She has not had
previous neurologic symptoms. A recent brain MRI RO–MF 4.3 9.0 6.7
shows multiple white matter lesions. She is being
evaluated for multiple sclerosis. P100 Waveform Amplitude Ratio
Discussion: This is an abnormal PRVEP due to pro- Location Ratio
longation of the P100 waveforms after both left (A)
and right (B) eye stimulation. The upper limit of nor- OS–OD 1.0
mal of the P100 waveform for this head size (32 cm)
LO–RO 1.2
is 109 ms. The P100 waveform is more prolonged
after right eye stimulation at 126 ms, but it is also
prolonged after left eye stimulation at 116 ms. Whenever the latency prolongation is less than 10 ms above
the upper limit of normal, consideration should be given to whether visual acuity problems may be causing the
latency prolongation. Because of this, the PRVEP is repeated with 60′ check size, but no significant change in
latencies is noted (not shown). Even with the larger check size, the P100 waveform latency does not decrease to
within the normal range. Thus the bilateral latency prolongation cannot be explained by the reduced visual acuity.
Bilateral P100 waveform latency prolongation can occur when the lesion(s) are bilateral prechiasmatic, chiasmatic, or
postchiasmatic. As noted previously, to distinguish between these possibilities, the parasagittal derivations should be
analyzed. If the P100 waveforms in the LO–MF and RO–MF derivations are symmetric (i.e., amplitude ratio <2.5 and
latency asymmetry <8 ms), bilateral prechiasmatic lesions are most likely. However, if the parasagittal derivations show
asymmetric P100 waveforms, a chiasmatic or postchiasmatic lesion becomes more likely. Differentiation between
these two can also be made based on whether one side (either LO–MF or RO–MF) is consistently lower in amplitude or
more prolonged in latency with independent stimulation of either eye. This is referred to as an “uncrossed asymmetry”
and implies a postchiasmatic lesion contralateral to the side (LO or MO) from which the lower amplitude or longer
latency was recorded. On the other hand, if the asymmetry shifts from one side to the other (i.e., after left eye stimu-
lation the LO–MF shows the lower amplitude or longer latency of the P100 waveform and after right eye stimulation
(C)
the MO–MF shows the lower amplitude or longer latency), this is referred to as an “crossed asymmetry.” This implies a
chiasmatic lesion. Whenever parasagittal asymmetry is noted, PRVEP with hemifield stimulation should be performed to confirm the site of the lesion. In
this patient, the parasagittal derivations showed symmetric P100 waveforms.
The brain MRI shows multiple areas of hyperintensity in the white matter. Some lesions enhance with contrast administration (C; T1 with contrast
image). No clear optic nerve lesions are noted.
The PRVEP findings are consistent with bilateral prechiasmatic visual pathway lesions with right side being more severely affected. The PRVEP sug-
gests additional lesions beyond those seen on MRI, helping make the diagnosis of MS.
Figure 2.24 (A–C) (continued )

Age: 18 years
Sex: Female
Filters: 1–100 Hz (+)↓
(A)
Eye: Left
Visual angle: 30′

MO–MF Derivation (μV) (μV) Mean (μV)
MO–AU MO–MF
Oz–AU MO–AU
MF–AU Oz–AU
LO–MF LO–MF
RO–MF RO–MF

Location Ratio
OS–OD
LO–RO

(B)
Eye: Right
(+)↓ Visual angle: 30′
164ms
105ms
132ms

MO–AU 106.0 132.0 164.0 MO–MF 3.4 4.5 3.9
Oz–AU 106.0 133.0 163.0 MO–AU 2.4 3.8 3.1
MF–AU Oz–AU 2.4 4.7 3.6
LO–MF LO–MF
RO–MF 120.0 143.0 172.0 RO–MF 3.0 3.2 3.1

Location Ratio
OS–OD
LO–RO

(C) (D) (E)

History: This is an 18-year-old female with bilateral eye pain and vision loss, left greater than right. She has a history of optic neu-
ritis, antiphospholipid antibody syndrome, membranous nephritis, and thrombocytopenia.
Discussion: This is an abnormal PRVEP due to prolongation of the P100 waveforms after both left (A) and right (B) eye stimulation.
After stimulation of the left eye, no reproducible response is noted. Absence of a response may be due to technical reasons such
a machine malfunction, inappropriate recording parameters, and so on. Additionally, an absent response may occur if the patient
does not attend to the pattern reversal stimulus or looks away. The technologist should troubleshoot equipment and confirm that
technical issues are not responsible for an absent waveform. The patient should be carefully observed to note if he or she is coop-
erating with the examination and looking at the stimulus. In this case the technologist carefully ruled out technical problems and
noted the patient to be cooperative and attending to the stimulus. As noted previously, reduced visual acuity can cause latency
prolongation of the P100 waveform, and the visual acuity for this eye is 20/200. It is unlikely, though, that an absent P100 waveform
can be explained on this basis. In this patient, the PRVEP is repeated with a 60′ check size and still a P100 waveform is not seen
(not shown). Whereas an absent P100 waveform can occur due to prechiasmatic, chiasmatic, or postchiasmatic lesions, it is mostly
severe prechiasmatic lesions that result in this degree of abnormality. Since the waveforms are absent, parasagittal derivation cannot
be evaluated to help confirm the localization of the lesion.
Stimulation of the right eye results in a P100 waveform that has a prolonged latency at 138.0 ms. The morphology of the response
is preserved. This degree of latency prolongation is unlikely to be explained simply by visual acuity problems. Stimulation with a
larger check size does not change the latency significantly (not shown). The parasagittal waveforms are not well formed and of
low amplitude. The RO–MF response is better formed than the LO–MF response, but all responses are of low amplitude, precluding
hemifield testing. The localization of the abnormality after right eye stimulation is more difficult and it may be prechiasmatic, chi-
asmatic, or postchiasmatic.
The MRI of the orbits shows enhancement of the optic chiasm and enlargement of the distal optic nerves (C; T1 with contrast
image). Brain MRI shows the optic nerve enhancement as well (D; T1 with contrast image) and rare, scattered punctate white matter
lesions (E; T2 image.)
The PRVEP abnormalities are not clearly localizing, but the absence of the response after left eye stimulation suggests a prechias-
matic visual pathway lesion on that side. The MRI findings suggest a chiasmatic or bilateral prechiasmatic lesion.

Age: 61 years
Sex: Female
Filters: 1–100 Hz
(+↓)
(A)
Eye: Left P100 Waveform Amplitude Ratio
Visual angle: 30′

MO–AU 103.0 124.0 145.0 MO–MF 4.5 5.9 5.2
Oz–AU 103.0 124.0 146.0 MO–AU 2.8 5.7
MF–AU 142.0 Oz–AU 3.0 7.0 5.0
LO–MF 103.0 137.0 165.0 LO–MF 4.4 3.3 3.8
RO–MF 105.0 136.0 157.0 RO–MF 2.2 1.4 1.8

(+↓)
(B)
Visual angle: 30′

MO–AU 102.0 127.0 154.0 MO–MF 4.4 4.0 4.2
Oz–AU 101.0 128.0 155.0 MO–AU 5.3 3.6 4.5
MF–AU ?148.0 Oz–AU 5.0 4.6 4.8
LO–MF 106.0 138.0 160.0 LO–MF 3.6 2.8 3.2
RO–MF 111.0 138.0 174.0 RO–MF 1.9 2.1 2.0

(C) (D)
History: This is a 61-year-old female who has had left sided weakness and dizziness for one month. MRI brain revealed an
infiltrating right hemispheric lesion. She is suspected to have an infiltrating neoplasm, multiple sclerosis, or vasculitis. This study is
done to help narrow the differential diagnosis and determine if other sites of demyelination are present.
Discussion: This is an abnormal PRVEP due to prolongation of the P100 waveform after stimulation of the left (A) and right (B) eyes.
The waveforms are reasonably reproducible. After stimulation of the left eye, the P100 waveform latency is prolonged to 125 ms.
This degree of prolongation is more than can be explained by visual acuity impairment. The parasagittal waveforms are present but
poorly defined. The P100 waveform in the RO–MF derivation is of very low amplitude and may be incorrectly tagged.
After right eye stimulation the P100 waveform is prolonged to 128.0 ms. The parasagittal derivations do not show clearly repro-
ducible waveforms, especially in the RO–MF channel. The N105 waveform appears to be present at about 148 ms, even though the
waveform is not tagged by the technologist.
Prolongation of the P100 waveforms after stimulation of the left and right eyes suggests that the localization of the lesion may be
prechiasmatic, chiasmatic, or postchiasmatic. Review of the parasagittal waveform amplitude ratios listed in the table suggests that
the lesions are likely to be bilateral prechiasmatic as the ratios are normal. However, closer review of the parasagittal waveforms
suggests that the RO–MF derivation shows a lower amplitude response after left eye stimulation and no response at all after right
eye stimulation (even though a waveform has been tagged). These findings suggest that a postchiasmatic lesion on the left side
may be present, but this should be confirmed with hemifield stimulation. However, the amplitude of the waveforms is low, with the
parasagittal waveforms being less than 5 mA, precluding hemifield testing.
MRI of the brain shows an infiltrating lesion on right side with a smaller area of high signal on the left side as well (C; FLAIR image).
The right sided lesion also has patchy enhancement with contrast (D; T1 with contrast image).
The PRVEP abnormalities appear to be prechiasmatic on review of the tables; however, review of the data suggests that the
abnormalities may be postchiasmatic, likely on the left side (cannot be confirmed as hemifield testing is not possible). The MRI
abnormalities are more severe on the right but are present on the left as well. This PRVEP does not establish definite prechias-
matic lesions, and makes a diagnosis of multiple sclerosis somewhat less likely as separate lesions from the one seen on MRI are
not confirmed.
FLAIR, fluid-attenuated inversion recovery; PRVEP, pattern reversal visual evoked potential.

Age: 69 years
Sex: Female
(A) P100 Waveform Amplitude Ratio

Eye: Left
Corrective lenses: No LO–RO

MO–AU 86.5 106.0 142.0 MO–MF 7.2 13.1 10.2
Oz–AU 88.0 108.0 142.0 MO–AU 7.9 13.5 10.7
MF–AU ?126 Oz–AU 7.4 13.3 10.5
LO–MF LO–MF
RO–MF 87.5 124.0 143.0 RO–MF 5.6 5.6 5.6
Figure 2.27 (A–H) Example 14. (continued )

(+↓)
(B)
Visual angle: 30′
Corrective lenses: No LO–RO 1.9

MO–AU 82.0 104.0 135.0 MO–MF 13.5 25.3 19.4
Oz–AU 82.0 104.0 133.0 MO–AU 13.6 23.5 18.6
MF–AU 119.0 Oz–AU 12.3 23.3 17.8
LO–MF 83.5 109.0 144.0 LO–MF 3.4 5.0 4.2
RO–MF 83.5 109.0 140.0 RO–MF 6.2 10.0 8.1
Figure 2.27 (A–H) (continued )

(+↓)
(+↓)
(C) (D)
(+↓)
(+↓)
(E) (F)

(G) (H)
History: This is a 69-year-old female with a history of transient vision loss lasting a few seconds that occurred several years ago. Several months later
she had an episode of speech difficulty. She underwent a brain MRI scan and was told she had multiple sclerosis. She has been free of symptoms
since then, but would like to have this evaluated. This study is being done to evaluate for multiple sclerosis.
Discussion: This is a normal PRVEP study with full-field and monocular hemifield stimulation. With full-field stimulation of the left (A) and right
(B) eyes, the P100 waveform latencies are normal at 107 and 105 ms, respectively. However, after stimulation of the left eye, parasagittal P100
waveforms are asymmetric, with the response being absent in the LO–MF derivation. After stimulation of the right eye, the parasagittal P100 wave-
forms are within normal limits of symmetry (LO–RO ratio is 1.9, upper limit of normal is 2.5). However, the LO–MF response is somewhat lower in
amplitude. Thus, even though the P100 waveforms to full-field stimulation are within normal limits, the asymmetric parasagittal derivations raise the
possibility of chiasmatic or postchiasmatic lesion. Since the LO–MF derivation is the abnormal one after left eye stimulation and of lower amplitude
after right eye stimulation, this is an “uncrossed asymmetry.” An uncrossed asymmetry implies a postchiasmatic lesion, which in this case is more
likely to be in the right hemisphere. This possible right hemispheric lesion must be confirmed by hemifield testing. Since the P100 waveform ampli-
tudes are large (>5 mV in most locations), hemifield testing can be performed.
Binocular hemifield stimulation can be performed when there is an uncrossed asymmetry noted on full-field stimulation. Monocular hemifields must
be performed when a crossed asymmetry is noted. Monocular hemifields are more sensitive, but take longer to perform and require more patient
cooperation. In this case, monocular hemifields were performed despite there being what appeared to be an uncrossed asymmetry on full-field
testing. This was done since there wasn’t a clear parasagittal abnormality after right eye stimulation.
Stimulation of the left eye, left hemifield (C) is normal with a P100 waveform noted in the LO–MF derivation. Left eye, left hemifield stimulation acti-
vates the nasal retina of the left eye, and impulses are transmitted through the optic chiasm to the right optic tract and ultimately to the RO cortex.
This results in a P100 waveform that is better visualized in the LO–MF derivation (contralateral projection of the P100 waveform vector; see Figure
2.9). In the author’s laboratory, latencies and amplitudes of waveforms obtained with hemifield stimulation are not measured, only their presence
or absence is noted. Stimulation of the left eye, right hemifield (D) is normal with a P100 waveform noted in the RO–MF derivation. Left eye, right
hemifield stimulation activates the temporal retina of the left eye, and impulses are transmitted through the optic chiasm to the left optic tract and
ultimately to the LO cortex. This results in a P100 waveform that is better visualized in the RO–MF derivation.
Stimulation of the right eye, left hemifield (E) is normal with a P100 waveform noted in the LO–MF derivation. Right eye, left hemifield stimulation
activates the temporal retina of the right eye, and impulses are transmitted through the optic chiasm to the right optic tract and ultimately to the RO
cortex. This results in a P100 waveform that is better seen in the LO–MF derivation. Stimulation of the right eye, right hemifield (F) is normal with a P100
waveform noted in the RO–MF derivation. Right eye, right hemifield stimulation activates the nasal retina of the right eye, and impulses are transmitted
through the optic chiasm to the left optic tract and ultimately to the LO cortex. This results in a P100 waveform that is better seen in the RO–MF derivation.
MRI of the brain shows multiple white matter lesions bilaterally (G; T2 image). There is no enhancement with contrast suggesting that these lesions
are not acute (H; T1 with contrast image).
This PRVEP is normal despite the parasagittal derivations on full-field testing suggesting a right hemispheric lesion. Monocular hemifield testing does
not confirm this finding. If hemifield testing was not performed, the parasagittal abnormality may have been thought to be consistent with the MRI
findings, but with the additional testing, this does not appear to be the case.

Age: 16 years
Sex: Male
Filters: 1–100 Hz ↓(+)
N145
(A)
Eye: Left
Number of repetitions: 200 P100 Waveform Amplitude Ratio
Visual angle: 30′ Location Ratio
OS–OD 1.7
LO–RO 1.1

MO–AU 83.5 108.0 136.0 MO–MF 4.3 8.2 6.2
Oz–AU MO–AU 1.5 5.0 3.3
MF–AU 109.0 Oz–AU
LO–MF 83.0 105.0 137.0 LO–MF 5.7 11.3 8.5
RO–MF 104.0 142.0 174.0 RO–MF 7.3 10.6 8.9
Figure 2.28 (A–H) Example 15. (continued )

(B)
Visual angle: 30′
OS–OD 1.7

MO–AU 81.5 108.0 136.0 MO–MF 7.9 12.8 10.3
Oz–AU MO–AU 3.8 8.5 6.1

LO–MF 79.5 103.0 131.0 LO–MF 5.7 10.8 8.3
RO–MF 102.0 133.0 179.0 RO–MF 4.3 12.6 8.4

↓(+) ↓(+)
(C) (D)
↓(+)
↓(+)
(E) (F)

(G) (H)
History: This is a 16-year-old male who has been having numbness and tingling of his right upper and lower extremities, academic decline, and
recent seizures. This study is being done to evaluate for demyelinating lesions in the visual pathway.
Discussion: This is an abnormal PRVEP study with full-field and monocular hemifield stimulation. With full-field stimulation of the left (A) and right
(B) eyes, the P100 waveform latencies are normal at 108 ms bilaterally. However, the P100 waveform in the RO–MF derivation after stimulation of
both sides is considerably prolonged compared to the P100 waveform in the LO–MF derivation. The upper limit of normal asymmetry is 8 ms, and
in this patient the asymmetries are 37 ms and 30 ms after stimulation of the left and right eyes, respectively. P100 waveform latency asymmetry has
the same significance as amplitude asymmetry. This represents an uncrossed asymmetry, implying a lesion in the left hemisphere. Recall that this is
because the P100 waveform vector projects to the contralateral hemisphere and is better recorded over the contralateral occipital electrode. Thus
prolonged latencies of the P100 waveform recorded over the RO electrode imply a lesion in the left hemisphere. PRVEP with hemifield stimulation
must be done to confirm this suspected postchiasmatic abnormality. Since the parasagittal abnormality is uncrossed, PRVEP with binocular hemifield
stimulation could have been done; however, the technologist performed monocular hemifield stimulation. The latter is more sensitive but takes
longer.
Stimulation of the left eye, left hemifield (C) is normal with a P100 waveform noted in the LO–MF (and LT–MF) derivation. Left eye, left hemifield
stimulation activates the nasal retina of the left eye, and impulses are transmitted through the optic chiasm to the right optic tract and ultimately
to the RO cortex. This results in a P100 waveform that is better visualized in the LO–MF derivation. Stimulation of the left eye, right hemifield (D) is
abnormal without a clear P100 waveform noted in the RO–MF derivation. Left eye, right hemifield stimulation activates the temporal retina of the
left eye, and impulses are transmitted through the optic chiasm to the left optic tract and ultimately to the LO cortex. This should result in a P100
waveform that is better visualized in the RO–MF derivation (absent in this case).
Stimulation of the right eye, left hemifield (E) is normal with a P100 waveform noted in the LO–MF (and LT–MF) derivation. Right eye, left hemifield
stimulation activates the temporal retina of the right eye, and impulses are transmitted through the optic chiasm to the right optic tract and ulti-
mately to the RO cortex. This results in a P100 waveform that is better seen in the LO–MF derivation. Stimulation of the right eye, right hemifield (F)
is abnormal without a clear P100 waveform noted in the RO–MF derivation. Right eye, right hemifield stimulation activates the nasal retina of the
right eye, and impulses are transmitted through the optic chiasm to the left optic tract and ultimately to the LO cortex. This should result in a P100
waveform that is better seen in the RO–MF derivation (absent in this case).
MRI of the brain shows a large region of signal abnormality in the left temporal, parietal, and occipital white matter (G and H; FLAIR images).
In this PRVEP the abnormal P100 waveforms noted in parasagittal derivations on full-field testing is confirmed with monocular hemifield testing.
The hemifield testing confirms the presence of a left postchiasmatic lesion. The MRI of brain also shows a large area of demyelination in the left
temporal, parietal, and occipital areas. The PREVEP localized the same abnormality as seen by the MRI and not another lesion in the anterior visual
pathways. This patient was later diagnosed to have adrenoleukodystrophy.
FLAIR, fluid-attenuated inversion recovery; PRVEP, pattern reversal visual evoked potential.

Age: 74 years
Sex: Female
Stimulation Rate: 3.9/s
Filters: 1–100 Hz
(A)
Eye: Left
Visual angle: 30′

MO–AU 129.6 MO–MF 7.1 12.8 10.0
In–AU 125.7 MO–AU
LO–MF 94.8 130.8 168.6 LO–MF 7.9 17.4 12.7
RO–MF 94.5 132.6 168.6 RO–MF 2.9 7.0 4.9

Location Ratio
OS–OD 1.1
LO–RO 2.6

(B)
Eye: Right
Visual angle: 30′

MO–AU 140.1 MO–MF 9.7 9.0 9.4
In–AU 133.8 MO–AU
LO–MF 102.0 136.8 184.8 LO–MF 10.6 13.8 12.2
RO–MF 100.8 136.8 195.0 RO–MF 4.4 5.1 4.8

Location Ratio
OS–OD 1.1
LO–RO 2.6

(E)
History: This is a 74-year-old female complaining of right sided numbness and pain.
(C) She does not have any weakness. There is also a history of chronic fatigue syndrome. A
CT scan of the brain showed no abnormalities. This study is being done to evaluate for
lesions in the optic pathways.
Discussion: This is an abnormal PRVEP study with full field and binocular hemifield stim-
ulation. With full-field stimulation of the left (A) and right (B) eyes, the P100 waveform
latencies are prolonged at 128.4 and 139.5 ms, respectively. The visual acuity cannot
explain these findings (20/25 and 20/40). Review of the parasagittal derivations shows
that the P100 waveform in the RO–MF derivations are significantly lower in amplitude
than those seen in the LO–MF derivations after stimulation of both eyes (amplitude ratio
is 2.6; upper limit of normal is 2.5). This is an uncrossed asymmetry and implies a lesion
in the left hemisphere. This finding, however, must be confirmed by performing PRVEP
with hemifield stimulation. Since the abnormality is uncrossed, binocular hemifields can
be performed.
Stimulation of the left hemifield of both eyes (C) is normal with a P100 waveform noted
in the LO–MF (and LT–MF) derivation. Binocular left hemifield stimulation activates the
nasal retina of the left eye and temporal retina of the right eye, and impulses are trans
mitted through the optic chiasm to the right optic tract and ultimately to the RO cortex.
This results in a P100 waveform that is better visualized in the LO–MF derivation.
Stimulation of the right hemifield of both eyes (D) is abnormal without a clear P100 wave-
form noted in the RO–MF derivation. Binocular right hemifield stimulation activates the
temporal retina of the left eye and nasal retina of the right eye, and impulses are trans-
mitted through the optic chiasm to the left optic tract and ultimately to the LO cortex.
This should result in a P100 waveform that is better visualized in the RO–MF derivation
(absent in this case).
A CT scan was performed prior to the PRVEP study and was normal (E).
In this PRVEP the P100 waveform latencies are prolonged after stimulation of both eyes.
(D) Because of the uncrossed asymmetry of the P100 waveform amplitudes noted in the
parasagittal derivations (with RO–MF derivation having the lower amplitude P100 waveform), the localization of the abnormality is most likely post-
chiasmatic in the left hemisphere. These findings are confirmed with the PRVEP with binocular hemifield stimulation. Prior CT of the head does not
show the lesion; however, given the PRVEP findings, additional more detailed neuroimaging should be considered.

Age: 8 months
Sex: Female
Visual acuity: N/A
Filters: 1–100 Hz
Eyes: Left and right, independent
(+↓)
(B)
History: This is an 8-month-old female recently diagnosed with
mucopolysaccharidosis IVA who is being considered for stem
cell transplantation. This is baseline visual pathway testing prior
to transplant. No definite visual symptoms have been noted.
Discussion: This is a normal FVEP. The left and right eyes are
stimulated with a flash stimulus using goggles (A). FVEPs were
preformed because of the patient’s age and inability to cooper-
ate with PRVEP. A P1 (also sometimes called the P100) waveform
is noted at approximately 125 ms after stimulation of either
eye. FVEPs are interpreted simply as being present or absent.
Because of significant variability, the latency of the responses is
not typically considered in determining abnormality.
(A) An MRI of the brain was normal for age (B; FLAIR image).
P1 Waveform FVEP is a very rudimentary method of analyzing the visual path-

way, and so a present response simply implies that there are
some fibers connecting the retinae to the visual cortices.
OS Present Presence of vision cannot be inferred from the presence of a
P1 waveform. In this patient, this study confirms presence of
OD Present some visual pathway fibers connecting the retinae to the visual
cortices. These findings are consistent with a normal MRI.
FLAIR, fluid-attenuated inversion recovery; FVEP, flash visual evoked potential; PRVEP, pattern reversal visual evoked potential.
Figure 2.30 (A and B) Example 17.

Age: 5 years
Sex: Male
Visual acuity: N/A
Filters: 1–100 Hz
Eyes: Left and right, independent
(+↓)
(B)
(A)
P1 Waveform
OS Absent
OD Absent
FVEP, flash visual evoked potential; PRVEP, pattern reversal visual

evoked potential.
(C)
Figure 2.31 (A–C) Example 18 (continued ).

History: This is a 5-year-old male with Tay Sachs disease who is having gradually progressive decline in cognitive and motor abilities.
He is displaying less visual and auditory responsiveness to his environment than before. This study is being done to assess visual
pathways. Previous FVEP 2 years ago was normal.
Discussion: This is an abnormal FVEP study. There is no reproducible P1 response seen after left or right eye stimulation with gog-
gles (A). FVEPs were performed as the patient’s cognitive status precluded performing PRVEP.
A brain MRI shows confluent foci of T2 hyperintensities in the periventricular region (B; T2 image) and diffuse volume loss (C; T1
image).
The absence of P1 waveforms in this FVEP suggests that there are no connections between the retinae and visual cortices. Whereas
the presence of FVEP does not mean that vision is present, absent FVEP makes presence of vision very unlikely. At this point this
patient likely does not have visual perception. This represents a worsening for this patient as a prior FVEP was reported to be normal.
The white matter changes and atrophy noted on MRI also show worsening compared to previous studies.
Figure 2.31 (A–C) Example 18 (continued ).
Conclusions
VEPs are an effective way to functionally evaluate a lesion. Assessment in patients who cannot coop-
the visual pathways. Careful analysis of the latency erate with PRVEPs can be done with FVEPs. VEPs
and amplitude of the P100 waveform is important, continue to be used in many neurological diseases
but further evaluation of the topography and mor- in which there is preferential involvement of the
phology are necessary as well. If necessary, hemi- visual pathways.
field stimulation can further clarify the location of
References
1. Husain AM. Evoked potentials overview. In: Ebersole retinotopic map of human primary visual cortex.
JS, Husain AM, Nordli DR, eds. Current Practice of Clin Neurophysiol. 1999;110:1793-1800.
Clinical Electroencephalography. Philadelphia, PA: 7. Misulis KE. Spehlmann’s Evoked Potential Primer. 2nd
Wolters Kluwer Health; 2014:442-487. ed. Boston, MA: Butterworth-Heinemann; 1994.
2. Husain AM, Hayes S, Young M, Shah D. Visual 8. Odom JV, Bach M, Barber C, et al. Visual evoked
evoked potentials with CRT and LCD monitors: potentials standard (2004). Doc Ophthalmol.
when newer is not better. Neurology. 2009;72:162-164. 2004;108:115-123.
3. ACNS. Guideline 9B: Guidelines on visual evoked 9. Brigell MG, Celesia GG. Visual evoked poten-
potentials. J Clin Neurophysiol. 2006;23:138-156. tials: advances in clinical and basic sciences.
4. Di Russo F, Martínez A, Sereno MI, et al. Cortical Electroencephalogr Clin Neurophysiol Suppl.
sources of the early components of the visual evoked 1999;49:95-102.
potential. Hum Brain Mapp. 2002;15:95-111. 10. Celesia GG, Brigell MG. Recommended stan-
5. Kurita-Tashima S, Tobimatsu S, Nakayama- dards for pattern electroretinograms and visual
Hiromatsu M, Kato M. The neurophysiologic sig- evoked potentials. The International Federation of
nificance of frontal negativity in pattern-reversal Clinical Neurophysiology. Electroencephalogr Clin
visual-evoked potentials. Invest Ophthalmol Vis Sci. Neurophysiol Suppl. 1999;52:53-67.
1992;33:2423-2428. 11. Taylor MJ, McCulloch DL. Visual evoked poten-
6. Slotnick SD, Klein SA, Carney T, et al. Using tials in infants and children. J Clin Neurophysiol.
multi-stimulus VEP source localization to obtain a 1992;9:357-372.
12. Porciatti V. Temporal and spatial properties of the 18. Aldosari M, Altuwaijri M, Husain AM. Brain-
pattern-reversal VEPs in infants below 2 months of stem auditory and visual evoked potentials in
age. Hum Neurobiol. 1984;3:97-102. children with Krabbe disease. Clin Neurophysiol.
13. Levy SR. Visual evoked potentials in pediatrics. 2004;115:1653-1656.
In: Chiappa KH, ed. Evoked Potentials in Clinical 19. Husain AM, Altuwaijri M, Aldosari M. Krabbe
Medicine. 3rd ed. Philadelphia, PA: Lippincott- disease: neurophysiologic studies and MRI correla-
Raven Publishers; 1997:147-156. tions. Neurology. 2004;63:617-620.
14. Wright KW, Eriksen KJ, Shors TJ, Ary JP. 20. De Meirleir LJ, Taylor MJ, Logan WJ. Multimodal
Recording pattern visual evoked potentials evoked potential studies in leukodystrophies of chil-
under chloral hydrate sedation. Arch Ophthalmol. dren. Can J Neurol Sci. 1988;15:26-31.
1986;104:718-721. 21. Markand ON, Garg BP, DeMyer WE, et al. Brain
15. Riemslag FC, Spekreijse H, Van Wessem TN. stem auditory, visual and somatosensory evoked
Responses to paired onset stimuli: implications potentials in leukodystrophies. Electroencephalogr
for the delayed evoked potentials in multiple Clin Neurophysiol. 1982;54:39-48.
sclerosis. Electroencephalogr Clin Neurophysiol. 22. Halliday AM, McDonald WI. Visual evoked poten-
1985;62:155-166. tials. In: Stalberg E, Young RR, eds. Neurology I:
16. Streletz LJ, Bae SH, Roeshman RM, et al. Visual Clinical Neurophysiology. London: Butterworths; 1981:
evoked potentials in occipital lobe lesions. Arch 228-258.
Neurol. 1981;38:80-85. 23. Chiappa KH, Hill RA. Pattern-shift visual evoked
17. Thompson PD, Mastaglia FL, Carroll WM. Anterior potentials: interpretation. In: Chiappa KH, ed.
ischaemic optic neuropathy. A correlative clinical Evoked Potentials in Clinical Medicine. 3rd ed.
and visual evoked potential study of 18 patients. J Philadelphia, PA: Lippincott-Raven Publishers;
Neurol Neurosurg Psychiatry. 1986;49:128-135. 1997:95-130.
3
Brainstem Auditory Evoked Potentials
Brainstem auditory evoked potentials (BAEPs) are AEPs can also be classified based on the site of
obtained by activation of the auditory pathway. recording. BAEP and middle- and long-latency
BAEP waveforms are far-field, short-latency AEPs are recorded using a vertex electrode. An
potentials that are usually recorded within 10 ms electrode placed near the cochlea can record the
of stimulation. They are of small amplitude, and ECochG. In this type of recording, high-amplitude
recording them can be challenging if the patient is waveforms arising from the very distal auditory
not relaxed. This chapter discusses BAEP pathway are clearly identified, though occasionally
classification, anatomy, methodology, interpre-
some of them may be seen in the BAEP.
tation, special considerations in pediatrics, and
clinical applications. BAEP examples with neuro-
imaging correlations will follow.
Anatomy
The auditory pathway extends from the external
auditory canal to the auditory cortex in the
Classification temporal lobe. Sound enters the external auditory
BAEPs are a type of auditory evoked potentials canal and produces vibrations of the tympanic
(AEPs). AEPs are classified primarily based on membrane (TM). These vibrations are transmitted
their latency. Short-latency AEPs are a series of through the middle ear ossicles to the oval window
peaks with amplitudes of about 0.2 µV that occur and the cochlea. The cochlea contains a b asilar
within 10 ms of stimulus onset and are thought to membrane that extends from the base to the

arise from various structures within the cochlea, apex, and on the basilar membrane are hair cells.
auditory nerve, and brainstem. BAEPs are a type of Movement of the hair cells and basilar membrane
short-latency AEPs. Another type of short-latency produces activation of cochlear nerve (also known
AEP is the electrocochleogram (ECochG). The as the auditory nerve) fibers. Parts of the basilar
ECochG includes responses from the cochlea and membrane closest to the base of the cochlea respond
the distal auditory nerve. Middle-latency AEPs to high-frequency sounds, while the part closest to
occur between 10 and 50 ms after stimulation, have the apex responds to low-frequency sounds.
amplitudes of about 1 µV, and are thought to arise Sound impulses travel in the cochlear nerve
from primary auditory cortical areas. Long-latency to the brainstem. The cochlear nerve joins with
AEPs occur after 50 ms of stimulation, are much the vestibular nerve carrying impulses from the
larger in amplitude, and thought to arise from audi- semicircular canals. Together these nerves form

tory association areas of the cortex. BAEPs are the the vestibulocochlear nerve (cranial nerve VIII).
only AEP routinely used for clinical purposes, and Fibers from the vestibulocochlear nerve travel to
this chapter is limited to the discussion of this type the rostral medulla and synapse on the dorsal and
of AEPs. ventral cochlear nuclei. About 50% of second order
113
neurons arising from the cochlear nuclei d ecussate

and ascend on the contralateral side to the s uperior
Methodology
olivary complex in the pons. The decussation The methods used for eliciting BAEPs can affect the
is known as the trapezoid body. The remaining morphology, amplitude, and latency of the wave-
50% of the fibers ascend to the ipsilateral superior forms. The nature of the auditory stimulus, as well
olivary complex. Thus, auditory fibers are less
as the equipment used to provide the stimulus, can
lateralized than many other types of sensory fiber influence the BAEP. This section focuses on the
tracts. From the superior olivary complex, auditory methods used for obtaining BAEPs for neurologic
fibers ascend in the lateral lemniscus to the inferior indications.
colliculus in the midbrain. Auditory fibers then
travel to the medial geniculate body in the thalamus Patient Variables
and then in the thalamic auditory radiations on to The patient undergoing a BAEP study is seated
the primary auditory cortex in the temporal lobe. in a reclining chair or lying in bed. It is important
Because of only partial decussation, activation of that the patient be comfortable and relaxed during
the auditory system on one side produces a ctivation this test. The waveforms to be recorded are very
of both primary auditory cortices (1). This pathway small in amplitude, and electromyographic (EMG)
is illustrated in Figure 3.1. Since BAEPs evaluate activity from the neck muscles may make these
only brainstem auditory transmission, only a frac- responses hard to record. A soft pillow under
tion of this pathway (until the inferior colliculus) is the neck often helps relax these muscles, making
evaluated in this test. it easier to record the BAEP. Unlike other types
of evoked potentials, cortical responses are not
recorded in BAEP. As such, it is not necessary for
the patient to remain awake and alert during test-
Auditory Auditory
cortex cortex
ing. In fact, if the patient falls asleep, neck mus-
cles may become more relaxed, leading to a better
recording (Figure 3.2A and B). At times sedation
Medial Medial is administered to help patients fall asleep during
geniculate geniculate
body body
the study. However, prior to doing so, conscious
sedation guidelines must be followed.
The auditory stimulus used to elicit the BAEP is
Inferior Inferior delivered to the ears with headphones. When very
colliculus colliculus young children are being tested, appropriately
sized headphones should be used. Large head-
phones may collapse the external auditory canal
Superior Superior
olivary olivary
in young children, causing spurious changes in the
complex complex BAEP. Alternatively, the sound can be delivered to
the ears through a tube connected to a transducer.
The latter technique is preferred due to infection
Ventral and control issues and because of the large and cumber-
dorsal cochlear
nuclei
some nature of the headphones. If a transducer with
ear insert tubes is used, the latency of the BAEP is
longer, as it takes the sound longer to reach the
Cochlea TM. The testing environment should be very quiet,
and the patient should not be able to hear anything
other than the test stimulus and masking noise.
Ipsilateral Contralateral
Stimulation Parameters
Figure 3.1 Diagrammatic representation of the auditory BAEP waveform can be significantly affected by
pathway arising from the cochlea. changes in the stimulation parameters. Not only
Chapter 3. Brainstem Auditory Evoked Potentials ■ 115
do stimulation rate and intensity affect the BAEP Intensity

waveform, stimulus polarity and delivery method Sound intensity is measured in decibels (dB)
are some of the other parameters that can have a and is expressed as the actual intensity com-
profound effect. pared to a reference value. The different ways in
which sound intensity can be expressed are the
Type following:
Several types of stimuli can be used to elicit BAEPs.
When BAEPs are performed for neurologic indica-
• Hearing level (HL): Average hearing thresh-
tions, broadband clicks are most appropriate. These
old in young adults when tested with pure
contain a wide range of frequencies and have a
tones of 500 µs.
duration is 100 µs with a rapid rise and fall of stimu-
• Normal hearing level (nHL): Average hearing
lus intensity (2). Because they are “broadband” and
threshold in young adults when tested with
contain many frequencies, these clicks stimulate a
broadband clicks of 100 µs. For clinical BAEP
large section of the cochlea. Pure tones, tone pips,
studies, nHL is used most often.
and filtered clicks use sounds in specific frequency
• Sensory level (SL): The hearing threshold for a
ranges, which allow specific parts of the cochlea
given patient.
to be stimulated. Tone pips and filtered clicks are
• Sound pressure level (SPL): The physical
used mostly for audiologic testing. Yet another
pressure of the sound; it is measured in rela-
type of stimulus, tone bursts, have a slower rise and
tion to a zero reference. This is used most often
fall of the stimulus and are used for middle- and
to measure the sound output of devices like
long-latency AEPs.
speakers.
Polarity
The polarity if the broadband click changes BAEP intensity is often measured in dBnHL. This
depending on whether the transducer diaphragm is the dB above normal hearing. For example, if a
moves toward or away from the TM. Movement 70-dB intensity click is to be delivered, and the nHL
away from the TM produces rarefaction clicks, and is 5 dB (averaging hearing threshold of broadband
movement toward the TM results in condensation clicks for young individuals), the stimulus inten-
clicks. Clicks of different polarities can affect the sity is set to 75 dBnHL. If the actual hearing thresh-
latencies of the BAEP. At times, especially when old (SL) of a patient is measured and found to be
high-intensity stimulation is used, clicks of alter- 10 dB, then a 70-dB stimulus would be delivered as
nating polarity (alternating between rarefaction 80 dBSL (3).
and condensation) are used. Alternating clicks With increasing stimulus intensity, the BAEP
result in elimination of the stimulus artifact as it waveforms increase in amplitude and become
cancels due to opposing polarities, but the laten- shorter in latency (Figure 3.6A and B). With very
cies of the various waveforms are slightly dif- high intensity, extra waveforms may also be noted,
ferent. The latency difference makes averaging making it hard to identify the waveforms of interest
responses less efficient and results in lower ampli- (Figure 3.7). As the stimulus intensity is reduced,
tude responses (Figure 3.3A–C). waveforms gradually reduce in amplitude and
Rarefaction clicks often produce shorter latency, eventually disappear. The last waveform to disap-
better morphology waveforms (Figure 3.4). How- pear is the wave V.
ever, in some patients condensation clicks produce
better waveforms (Figure 3.5). There is no audiolog- Rate
ically detectable difference between the polarities. The rate of stimulation can affect the latency and
Laboratories should have the capability of per- amplitude of the BAEP waveforms. Typically,
forming BAEPs of both polarities. If abnormalities clinical studies are performed with “slow” stimu-
are present with one polarity, it is helpful to try the lation rates of 8 to 12/s. Occasionally fast rates of
other polarity to see if the responses improve. If 50/s can also be used. The advantage of fast rates is
high-stimulus intensities are required due to hear- that the study can be completed much more quickly;
ing loss, alternating polarity may be needed. this is especially important when young patients
Age: 29 years
Sex: Female
Scale: Amplitude = 0.2 µV/div; Latency = 1.5 ms/div
Montage: Cz–Ai, Cz–Ac, Ac–Ai
(A)
Side: Right
Absolute Latencies
Waveform Latency (ms) Amplitude (µV)
III 4.44 Waveforms Latency (ms)
V 5.97 0.65 (V–Vn) I–Vc 3.90
Vc 6.27 I–III 2.07
2.07 (V/I ratio) III–Vc 1.83
Figure 3.2 (A and B) Effects of muscle relaxation on the recording. (continued )

(B)
Side: Right
Absolute Latencies
V 6.27 0.43 (V–Vn) I–Vc 4.09
Vc 6.37 I–III 2.22
Discussion: In this figure, note that the montage on the waveform is inverted, and the correct montage is noted above. Without
averaging, BAEP waveforms are of very low amplitude. To visualize reliable and reproducible waveforms, noise must be m inimized.
A common source of noise is EMG activity, especially from neck muscles. When no attempt is made to relax the patient, muscle activ-
ity can make it difficult to visualize the BAEP. In this example, initially no attempt is made to relax the patient, and there is a lot of
EMG contamination in the recording (A). To be able to average despite the EMG activity, the amplifier sensitivity had to be increased.
Thereafter, an attempt is made to relax the patient by putting a pillow under her neck (B). This considerably increases the signal-to-
noise ratio and improves the BAEP morphology. In BAEP studies, it is important to try to relax the patient as much as possible.
BAEP, brainstem auditory evoked potential; dBnHL, decibel above normal hearing level; EMG, electromyographic.

Age: 29 years
Sex: Female
(A)
Side: Right
Absolute Latencies
V 6.27 0.43 (V–Vn) I–Vc 4.09
Vc 6.37 I–III 2.22
Figure 3.3 (A–C) Effects of rarefaction, condensation, and alternating clicks. (continued )
(B)
Side: Right
Absolute Latencies
V 6.30 0.56 (V–Vn) I–Vc 4.01
Vc 6.36 I–III 2.05

(C)
Side: Right
Absolute Latencies
V 6.30 0.61 (V–Vn) I–Vc 4.00
Vc 6.40 I–III 2.10
Discussion: In this example, note that the montage on the waveform is inverted, and the correct montage is noted above. Broadband
clicks of rarefaction or condensation polarity can be delivered to the ear to produce BAEP. Though the same waveforms are seen with
both types of clicks, the morphology and the latency of the various waves can vary. Often the waveforms are of shorter latency with
rarefaction clicks (A). Notice the upward deflection of the stimulus artifact (arrow). Condensation clicks may have a slightly longer
absolute latency, but IPLs are comparable (B). Notice the downward deflection of the stimulus artifact (arrow). Alternating clicks (C) are
obtained by alternating between rarefaction and condensation clicks. BAEPs with rarefaction and condensation clicks can be electron-
ically averaged to produce alternating clicks. Since the latencies of the various waveforms are different in the two polarities of clicks,
averaging them together may result in slightly lower amplitudes. In the example shown, this does not happen much as the latencies of
the various waveforms in the rarefaction and condensations clicks are not much different. However, the stimulus artifact is eliminated
(arrow). Alternating clicks are used when using high-stimulation intensity as that can result in significant stimulus artifact.
BAEP, brainstem auditory evoked potential; dBnHL, decibel above normal hearing level.

Age: 7 years
Sex: Male
Stimulation rate: 11.1/s (+)↑
Scale: Amplitude = 0.1 µV/div; Latency = 1 ms/div
Side: Left
Click polarity: Rarefaction and condensation
Rarefaction clicks
Absolute Latencies
V 6.42 0.67 (V–Vn) I–Vc 4.20
Vc 6.50 I–III 2.40
Condensation clicks
Absolute Latencies
V 6.68 0.41 (V–Vn) I–Vc 4.56
Vc 6.86 I–III 2.32
Discussion: Rarefaction clicks often result in shorter latency and better morphology than condensation clicks. In the sample shown here,
the BAEP obtained with rarefaction clicks showed more clearly defined waves I, III, V, and Vc than those obtained with condensation clicks.
This makes determining absolute latencies and IPLs easier. The latencies are not shorter for rarefaction clicks in this example. It is important
to obtain BAEPs in rarefaction and condensation polarities to ensure that the best morphology, shortest latency waveforms are evaluated.
BAEP, brainstem auditory evoked potential; IPL, interpeak latency.
Figure 3.4 Better waveforms with rarefaction clicks.

Age: 4 years
Sex: Male
Stimulation rate: 11.1/s ↑(+)
Side: Left
Click polarity: Rarefaction, condensation, and alternating
Rarefaction clicks
Absolute Latencies
I 2.22 0.17 (I–In)
III 6.99
V 11.20 0.10 (V–Vn)
Vc N/A
0.59 (V/I ratio)
Interpeak Latencies
I–Vc N/A
I–III 4.77
III–Vc N/A
Condensation clicks
Absolute Latencies
V 10.80 0.10 (V–Vn) I–Vc 7.74
Vc 10.70 I–III 4.05
Alternating clicks
Absolute Latencies
V 11.20 0.07 (V–Vn) I–Vc N/A
Vc N/A I–III 3.75
0.70 (V/I ratio) III–Vc N/A
Discussion: The waveforms in this BAEP study are best seen with condensation clicks. Though the latency is slightly longer with
this polarity, the waveforms, particularly wave V, are better defined. With rarefaction and alternating clicks, the waves V and Vn are
difficult to identify. With alternating clicks, the wave I is also of much lower amplitude. All the IPLs in this example are prolonged,
but it demonstrates how it is often unpredictable which polarity will show the best BAEP waveforms.
BAEP, brainstem auditory evoked potential; dBnHL, decibel above normal hearing level; IPL, interpeak latency.
Figure 3.5 Better waveforms with condensation clicks.

are being studied. However, faster stimulation rates With these electrodes, an adequate montage can
result in longer latencies and lower amplitudes of be constructed to define the BAEP waveforms.
the waveforms (Figure 3.8A and B). When faster
rates are used, normative data for that rate should Recording montage
be consulted. In the author’s laboratory, when the The BAEP montage should have a minimum of two
BAEP is abnormal with fast stimulation rates, it is channels. Unlike other types of evoked p otentials,
repeated with slow rates to confirm the abnormality. BAEPs are often displayed with positivity as
an upward deflection. All BAEP studies should
Masking noise be clearly marked showing this convention to
Broadband clicks delivered to the external ear eliminate any confusion.
canal are transmitted by air conduction to stimu- The montage recommended by the American
late the ipsilateral cochlear hair cells and basement Clinical Neurophysiology Society (ACNS) for
membrane. Additionally, sound is transmitted recording BAEPs is shown below (2). Additional
through bone conduction and reaches not only the channels can be added if needed to better define the
ipsilateral but also the contralateral cochlea. There waveforms.
is approximately 30 to 40 dB attenuation when
sound is transmitted via bone conduction; thus Channel 1: Cz–Ai (ipsilateral)
for a 70-dBnHL stimulus delivered to the left ear, Channel 2: Cz–Ac (contralateral)
approximately 30 to 40 dBnHL reach the right ear
through bone conduction. This amount of sound In the author’s laboratory, an additional channel
is sufficient to activate the contralateral cochlea is added as follows:
and produce a BAEP. This can produce confusion
in interpretation as wave I is seen in the contralat- Channel 3: Ac–Ai
eral derivation (see below). To prevent activation
through bone conduction, white masking noise is BAEP waves II–V are seen as surface positive,
delivered to the contralateral ear at an intensity far-field potentials from the Cz electrode. Because
that is 30 to 40 dB less than used for stimulation. positive waveforms are displayed as upward deflec-
The white masking noise is used routinely for tions, all four of these waves appear as upward
clinical BAEP studies. deflections. Wave I is a surface negative, near-field
potential recorded from the ipsilateral ear or mas-
Recording Parameters toid process electrode. Because the ipsilateral ear
BAEPs are a recording of the auditory signal as electrode is in Input 2 (G2) of the amplifier, a neg-
it ascends in the brainstem. Cortical waveforms ative potential in this input is also displayed as an
are not recorded. Most BAEP waveforms are far- upward deflection. Thus, all five BAEP waveforms
field potentials, and the auditory volley can be are displayed as upward deflections.
recorded with very few electrodes as it ascends. In the Cz–Ai derivation, all five BAEP waveforms
However, in order to correctly identify the wave- are seen. However, in the Cz–Ac derivation, the
forms, it is important to record the BAEP in different wave I is not recorded as it is a near-field potential
derivations. arising near the Ai electrode. Only waves II–V are
recorded in this derivation. The Ac–Ai derivation
Electrode placement is used to identify wave I; other waveforms are less
Surface electrodes are used to record BAEP. All clear in this derivation as there is no scalp electrode.
electrodes should have an impedance of less than In this derivation, the wave I is again an upward
5 kΩ (2). The following electrodes should be placed deflection, as the Ai electrode is connected to Input
to record the BAEP: 2 (G2) of the amplifier (Table 3.1).
• Vertex—located at the Cz location Amplifier settings

• A1/2 or M1/2—Bilateral ear or mastoid The ACNS guidelines recommend filter setting
processes of 10 to 3,000 Hz for clinical BAEP studies (2).
• Ground—anywhere on the scalp; forehead is Increasing the low-frequency filter to 150 Hz can
often a reasonable location help reduce EMG artifact, but it can also cause
Age: 26 years
Sex: Female
Filters: 150–3,000 Hz (+↑)
(A)
Side: Left
Rarefaction clicks
Absolute Latencies
V 6.80 0.41 (V–Vn) I–Vc 4.50
Vc 6.84 I–III 2.82
Condensation clicks
Absolute Latencies
V 7.04 0.36 (V–Vn) I–Vc 4.42
Vc 7.12 I–III 2.50
Figure 3.6 (A and B) Effects of increasing stimulation intensity. (continued )

(B)
Side: Left
(+↑) Masking intensity: 55 dBnHL
Absolute Latencies
V 6.30 0.65 (V–Vn) I–Vc 4.22
Vc 6.44 I–III 2.32
Discussion: There are several ways to improve the morphology of the BAEP waveforms. Doing so helps in proper identification
of the waveforms. In this example, with stimulation intensity of 70 dBnHL, the BAEP waveforms were of low amplitude with
both rarefaction and condensation clicks (A). By increasing the stimulus intensity to 85 dBnHL, the morphology and amplitude
of the waveforms increases considerably, making waveform identification clearer (B). Alternating clicks were used because with
high-stimulation intensity, the stimulus artifact becomes larger. With alternating clicks, the stimulus artifact is cancelled.

Age: 52 years
Sex: Male
↑(+)
Side: Right
Absolute Latencies
V 6.48 0.33 (V–Vn) I–Vc 4.26
Vc 6.69 I–III 2.34
Discussion: At times, waves in addition to the typical I–V waveforms can be recorded in a BAEP. This is especially true if high-intensity
stimulation is used. In this example, waves VI (thin arrow) and VII (thick arrow) are seen after the wave V. Wave V can be clearly
identified since it is fused with wave IV in the Cz–Ai derivation and in the Cz–Ac derivation they are clearly separated. Wave VI is
thought to arise from the medial geniculate body and wave VII from the auditory radiations. If these extra waveforms cause difficulty
in identification of wave V, the stimulus intensity should be decreased. In this case, the stimulus intensity was 85 dBnHL, but it was
not reduced as waves I–V could be clearly identified.
Figure 3.7 Extra waveforms with excessively high stimulation intensity.

Age: 40 years
Sex: Female
(+)↑
(A)
Side: Right
Absolute Latencies
I 2.38 0.13 (I–In)
III 4.30
V 6.36 0.32 (V–Vn)
Vc 6.40
2.5 (V/I ratio)
Interpeak Latencies
I–Vc 4.02
I–III 1.92
III–Vc 2.10
(B)
Side: Right
Absolute Latencies
V 6.14 0.40 (V–Vn) I–Vc 1.84
Vc 6.20 I–III 1.88
Discussion: The morphology and latency of the BAEP waveforms can change with stimulation rate. When a fast stimulation rate of 51/s is used,
the study can be completed quickly. However, with a slow stimulation rate of 11/s, the waveforms are of better morphology, shorter latency, and
higher amplitude. In the author’s laboratory, often patients are initially studied with fast rates. If the BAEP is abnormal at a fast rate, the study
is repeated with slow rates. In the example shown, initially BAEPs are obtained with rarefaction clicks at a stimulation rate of 51/s (A), and then
the study is repeated with a stimulation rate of 11/s (B). With the slower rate, notice that the BAEP morphology is clearer, IPLs are shorter, and
amplitudes are larger than with faster rates. The BAEP is normal at both stimulation rates in this example; the difference is fast and slow rates are
even more evident when the study is abnormal. When BAEPs are used in the operating room, a rate of 31/s is used as a compromise between
quickly replicating the waveforms and having good morphology.
Figure 3.8 (A and B) Effects of changing stimulation rate.

Table 3.1 The Channels of a BAEP and the Waveforms Seen Table 3.2 BAEP Waveforms and Their Neural Generators
in Those Channels
Wave I Distal vestibulocochlear nerve
Montage Waveform(s) Wave II Proximal vestibulocochlear nerve and
Cz–Ai I, II, III, IV, V cochlear nuclei
Cz–Ac II, III, IV, V Wave III Superior olivary complex
Ac–Ai I Wave IV Lateral lemniscus
Wave V Inferior colliculus
a shift in latency (Figure 3.9A–C). In the author’s BAEP, brainstem auditory evoked potential.
laboratory, all clinical BAEP studies are performed In the Cz–Ac derivation, the waveforms are labeled
with filter settings of 150 to 3,000 Hz. In the oper- IIc, IIIc, IVc, and Vc; the after-going negativity is
ating room (OR), the high-frequency filter is some- called IIcn, IIIcn, IVcn, and Vcn.
times changed to 1,500 Hz to reduce high-frequency The morphology of the BAEP waveforms appears
noise. Clinical studies should use the same filter slightly different in the ipsilateral and contralateral
settings as were used to acquire the normative data. derivations. These differences can be used to dis-
Because BAEPs are of low amplitude, typically tinguish the waveforms. The general principles in
1,000 to 4,000 repetitions are required to obtain reli- the following list can be helpful in identifying the
able, reproducible responses. If the signal-to-noise waveforms (4,5) (Figure 3.10).
ratio is high, fewer repetitions may be enough. Channel: Cz–Ai
Each BAEP must be replicated at least once to con-
firm presence of waveforms. If the responses are • Starting from the right and looking to the left,
noisy, more than one replication may be appro- wave V is peak just before (to the left of) the
priate. Analysis time (time window) should be at most significant negative waveform.
least 10 ms; at times with very delayed responses, a • Waves IV and V often fused, and wave V is a
15 ms analysis time may be more appropriate. deflection on wave IV
• Waves II and III separated
Interpretation • Wave I present, typically not earlier than
1.5 ms.
BAEP waves I–V are evaluated in routine clinical
studies to determine if there are problems with the Channel Cz–Ac
auditory pathways in the brainstem. Depending on • Wave Vc slightly later in latency than wave V
the individual, BAEP waveforms may be better seen • Waves IVc and Vc separated
with rarefaction or condensation clicks. The polar- • Waves IIc and IIIc often fused
ity with which better data with shorter latencies are • Wave I not present
obtained is used for interpretation. It is important
Channel Ac–Ai
to compare test data to normative data that were
acquired with exactly the same parameters. • Wave I seen as a prominent upward deflection
• Other waves not as clearly identifiable
Expected Waveforms
There are five BAEP waveforms that arise from Parameters Measured
brainstem neural generators (Table 3.2). Even Analysis of BAEP waveforms should start with con-
though a single brainstem structure is often referred firming that all waveforms are present. The various
to as the generator of a particular waveform, in fact montage derivations should be evaluated to verify
many structures contribute to a single waveform, that waveforms are correctly tagged. The absolute
and one neural structure may contribute to more latencies of waves I, III, V, and Vc are determined.
than one waveform. However, interpeak latencies (IPLs) are better for
The BAEP waveforms in the Cz–Ai derivation are assessment of brainstem conduction as absolute
identified in the order in which they appear: waves latencies also include the peripheral part of the
I, II, III, IV, and V. The negativity or valley after the vestibulocochlear nerve. Amplitude of the BAEP
positive peak is labelled In, IIn, IIIn, IVn, and Vn. waveforms can be used to determine normalcy, but
this is not as definite a criterion as latency (discussed obligate ones. Using alternating clicks can eliminate
below). Side-to-side IPL asymmetry should also be stimulation artifact, making it easier to see wave I.
evaluated, and if it exceeds normal limits, the side This latter technique may, however, result in loss
with the longer latency is abnormal even if it falls of amplitude if the latencies of the various wave-
within “normal limits.” forms are different for the two stimulation polarities.
Obligate waveforms Absolute latency
In neurologically and audiologically normal individ- Absolute latencies of waves I, III, and V should be
uals, waves I, III, and V should always be present evaluated. Approximate latencies of these waves
(Figure 3.10). Absence of any one of these waves is are 1.5, 3.5, and 5.5 ms; however, these latencies
considered an abnormality. Waves II and IV may not should not be used to identify these waveforms.
always be evident and may be fused with other wave- If a transducer with ear inserts is used to deliver
forms. Before any waveform is deemed to be absent, the broadband click to the ear and normative data
all derivations (Cz–Ai, Cz–Ac, and Ac–Ai) should were acquired with headphones, an additional
be evaluated and alternate stimulation parameters 0.9 to 1.0 ms is added to the absolute latency to com-
should be used (i.e., different polarity, slower rate). pensate for the extra tubing length that sound must
Correctly identifying the various BAEP peaks travel. Assessment of the absolute latency includes
is important. The technique for picking the correct conduction of sound through the external, middle,
peaks as waves I–V are discussed earlier and sum- and inner ear; the vestibulocochlear nerve; and the
marized in the next paragraph. The absolute latencies brainstem. In particular, the wave I absolute latency
should not be used to determine which waveform is is a measure of conduction along the ear and ves-
wave I or wave V, that is, just because a waveform tibulocochlear nerve. Pathology affecting the ear
occurs at approximately 1.5 and 5.5 ms does not make or the nerve, such as various types of deafness, can
it a wave I or wave V, respectively. Additionally, the result in absolute latency prolongation of wave I.
first waveform should not be picked as wave I, and Other BAEP waveforms will also be proportionally
subsequent ones identified as waves II–V. prolonged in this situation.
When starting to tag BAEP waveforms, it is best In the author’s laboratory, the absolute latencies
to start from the right side of the tracing and identify of waves I and V are evaluated and normative data
the greatest negative deflection. The positive peak are present in Table 3.3. Published normative data
before the negative deflection is the wave V. This can are also available from other sources (6).
be done in the Cz–Ai or Cz–Ac derivation; if the lat-
ter is used, the wave Vc will be identified. Thereafter, Interpeak latency
counting waveforms backward helps identify waves BAEP IPLs are a better measure of conduction along
IV, III, II, and I. As noted earlier, waves IV and V the auditory pathways in the central nervous sys-
are often fused in the Cz–Ai derivation, but they tem. The IPLs that are usually evaluated are waves
are more separated in the Cz–Ac derivation. On the I–III, III–V, and I–V. In some laboratories, wave Vc
other hand, waves II and III are separated in the is used to calculate the IPLs instead of wave V. This
Cz–Ai derivation, but fused in the Cz–Ac deriva- is because the definition of wave Vc is often better
tion. The Cz–Ai and Ac–Ai derivations demonstrate than wave V (as waves IV and V are more sepa-
wave I, but the Cz–Ac derivation does not. rated in the Cz–Ac derivation). This is the case in
Waveforms can be enhanced by slowing the stim- the author’s laboratories, and IPLs are expressed
ulation rate and increasing stimulus intensity. This as waves I–III, III–Vc, and I–Vc (Table 3.3). Many
results in an increase in the amplitude and decrease laboratories will have additional IPL normative
in the latency of the waveforms, particularly wave I. data based on age and gender. Patients older than
However, as noted before, a very high stimulation 60 years often have a longer waves I–V IPL and
intensity may result in waveforms after wave V, women then to have a shorter wave I–V IPL.
making it more difficult to identify the waveforms of The waves I–III IPL is a measure of auditory
interest. In this situation, reducing the stimulus inten- impulse conduction between the distal vestibuloco-
sity helps. At times, changing between rarefaction chlear nerve and the superior olivary complex. This
and condensation clicks improves the morphology IPL is considered a measure of conduction in the
of the waveforms and makes it easier to identify the ipsilateral lower brainstem. The waves III–V IPL
Age: 29 years
Sex: Female
(A)
Side: Right
Absolute Latencies
V 6.40 0.40 (V–Vn) I–Vc 4.07
Vc 6.50 I–III 2.16
Figure 3.9 (A–C) Effects of changing filter settings. (continued )

(B)
Side: Right
Stimulation Intensity: 70 dBnHL
Masking Intensity: 40 dBnHL
Absolute Latencies
V 6.27 0.43 (V–Vn) I–Vc 4.09
Vc 6.37 I–III 2.22

(C)
Side: Right
Absolute Latencies
V 6.39 0.48 (V–Vn) I–Vc 4.08
Vc 6.56 I–III 2.12
Discussion: In this example, note that the montage on the waveform is inverted, and the correct montage is noted above. The
filter settings recommended for BAEPs by the ACNS are 10 to 3,000 Hz. A BAEP obtained at these filter settings is presented in this
example (A). Notice that with 2,000 repetitions, there is still baseline sway. When the low-frequency filter is increased to 150 Hz,
the sway artifact can be significantly reduced (B). Many clinical laboratories prefer to use this filter setting to enhance the ability
to obtain reliable BAEPs. Normative data should, however, have been obtained at whichever filter setting is being used to collect
patient data. In the OR, the high-frequency filter is often reduced to 1,500 Hz (150–1,500 Hz) to eliminate the high-frequency noise
from that environment (C). This results in only a minor change, which is not clinically important in the OR as the patient serves as
his or her own control.
BAEP, brainstem auditory evoked potential; dBnHL, decibel above normal hearing level; OR, operating room.

Age: 34 years
Sex: Female
(+↑)
Side: Left
Absolute Latencies
I 2.42 0.44 (I–In)
III 4.56
V 6.46 0.38 (V–Vn)
Vc 6.50
0.86 (V/I ratio)
Interpeak Latencies
I–Vc 4.08
I–III 2.14
III–Vc 1.94
Discussion: The BAEP waveforms look slightly different in the Cz–Ai, Cz–Ac, and Ac–Ai derivations. In the Cz–Ai derivation, it is
best to look for the wave V first. This is done by starting at the right and looking toward the left for the most prominent negative
(downward) waveform. Immediately before (to the left of) this waveform is the wave V. In this derivation, the waves IV and V are
close together and even fused (in this example they are not fused). Further to the left of wave IV is wave III. Wave III and wave II
are often well separated from each other. Further to the left of wave II is wave I. The absolute latency of wave I is seldom earlier
than 1.4 ms, though this should not be relied upon to identify the wave I. It is important to not start counting waves from left to
right. If that was done, the first wave would have been tagged as the small deflection that is earlier than the correctly tagged wave I.
The Cz–Ac derivation should be used to confirm that waveforms have been correctly identified. Once again, wave V is sought be
scanning from right to left. The wave immediately before (to the left of) the most significant negative deflection is wave Vc. Notice
that waves Vc and IVc are separated more than waves V and IV. Further to the left, notice that waves IIIc and IIc are closer together
than their counterparts, waves III and II (which are nicely separated). Finally, wave I is not present in this derivation as it is a near-field
potential recorded from the Ai electrode.
The Ac–Ai derivation is used mainly to confirm the location of wave I. Wave I should be seen in this derivation in about the same
location as in the Cz–Ai derivation, while it is absent in the Cz–Ac derivation. Other waveforms may be seen in this derivation, but
they are not as well defined as in the other derivations.
It is important to look at the BAEP in the different derivations so that waveforms can be correctly tagged and latencies appropriately
calculated.
Figure 3.10 Ipsi and contralateral derivations.

Table 3.3 Normative Data for BAEPs From the Author’s Other waveforms
Laboratory Occasionally, parts of the ECochG can be seen in the
Wave Absolute Latency or IPL (ms)*+ BAEP as well, but if they are seen, they are of very
I 2.0 low amplitude. The ECochG has three components,
V 6.3 the cochlear microphonic, summating potential,
I–III 2.6 and cochlear nerve action potential. These are seen
III–Vc 2.3 only in the Cz–Ai (and occasionally Ac–Ai) deriva-
I–Vc 4.6^
tion as they are near-field potentials recorded from
the Ai electrode. The cochlear microphonic arises
* Latency values 3 SD beyond mean from the movement of cochlear hair cells against
+ Note age/gender effects below the membrane above it (tectorial membrane). This is
^ Maximum wave I–Vc side-to-side latency asymmetry = often a polyphasic waveform that changes polarity
.4 ms with change in stimulation polarity. Because of this
Stimulation rate ≈ 11/s
property, the cochlear microphonic must be differ-
For a rate of ≈ 31/s add 0.1 ms, and for ≈ 51/s add 0.2 ms to
latencies entiated from the stimulus artifact. The summating
Wave V/I amplitude ratio ≥ 0.7 potential is generated by the hair cells and is even
smaller in amplitude than the cochlear microphonic.
Gender and age effects on wave I–Vc IPL The cochlear nerve action potential arises from the
Age (yrs.) Male Female most distal fibers of the cochlear nerve, and can be
<60 4.6 ms 4.5 ms regarded as the same as wave I of the BAEP. Neither
>60 4.7 ms 4.6 ms the summating potential nor cochlear nerve action
potential reverses polarity with changing stimulus
BAEP, brainstem auditory evoked potential; IPL, interpeak latency; SD,
polarity, thus differentiating them from the cochlear
standard deviation.
microphonic. The presence or absence of these wave-
is a measure of conduction between the superior forms does not affect the interpretation of the BAEP.
olivary complex and the inferior colliculus. This is
considered a measure of conduction in the upper Abnormalities
brainstem; however, laterality is not possible to The types of abnormalities that are seen with BAEP
determine because of the partial decussation of the are: loss of an obligate waveform, prolongation of
auditory pathways. The waves I–V IPL are a mea- IPLs, prolongation of absolute latency, and reduced
sure of conduction in the entire brainstem (from the amplitude. Complete loss of a waveform is the most
vestibulocochlear nerve to the inferior colliculus), significant abnormality and implies a lesion of the
with lateralization not being possible (7). generator of that waveform or in the auditory path-
In addition to determining if the IPLs are within way distal to it. IPL prolongations imply a lesion in
normal limits, they should be compared from side the auditory pathway between the generators of the
to side. Symmetry data from the author’s laboratory relevant peaks. Prolongations of the absolute latency
are presented in Table 3.3. Even if both waves I–V are less specific, and they imply a lesion in the audi-
IPL are within normal limits, but the side to side tory pathway distal to the peak in question. The lesion
difference is beyond the upper limit of normal, the could affect the most distal parts of the auditory
longer IPL should be interpreted as being abnormal. system, including the ear and the vestibulocochlear
nerve. IPLs and absolute latencies are considered
Amplitude
abnormal if they exceed the upper limit of the nor-
Amplitude abnormalities are of less certain clini- mative data. Even if latencies do not exceed this limit,
cal significance than latency abnormalities. When but the side to side asymmetry is greater than nor-
they occur along with latency prolongation, they mal, that is also abnormal. A low amplitude is not as
are more meaningful. The amplitudes of waves I convincing an abnormality as latency prolongation.
and V are measured, and the ratio of waves V/I is
reported. An amplitude ratio of waves V/I of less Loss of obligate waveforms
than 0.7 is considered abnormal (in some laborato- The obligate waveforms of a BAEP are waves I, III,
ries, 0.5 is considered abnormal). This implies that and V. These waveforms are present in all normal
under normal circumstances, wave V amplitude is individuals and their absence is abnormal. Before
greater than wave I. any of these waveforms is considered absent,
techniques for enhancing and correctly identifying wave I absolute latency is suggestive of a lesion in
the waveform should be undertaken, as discussed the ear or distal vestibulocochlear nerve. This abnor-
earlier. Absence of a waveform implies that there is mality is often seen with hearing loss. In this situ-
a lesion at its generator or distal to it. Loss of wave I ation, the IPLs are normal as there is no brainstem
or III is usually due to an ipsilateral lesion. Because pathology. Before the wave I latency is considered
of the incomplete decussation of the auditory fibers, prolonged, compensation must be allowed for use of
loss of wave V may represent an ipsilateral or con- ear inserts and tubing if a transducer was used and
tralateral lesion. The following can be used as an normative data were acquired with headphones.
interpretation guide to loss of obligate waveforms:
Reduced amplitude ratio
Loss of all obligate waves: Technical issues must A low amplitude is the least reliable marker of
be ruled out; lesion in the ear or distal vestibulo- abnormality in a BAEP as waveform amplitudes
cochlear nerve, ipsilateral to side of stimulation. are highly variable. To help minimize the effect of
Loss of waves III and V: Lesion in the auditory variable amplitudes, an amplitude ratio of waves
pathway in the brainstem at or below the level of V/I is calculated. If waves IV and V are fused,
the superior olivary complex, usually ipsilateral the waves IV-V/I amplitude ratio is determined.
to side of stimulation. A wave V/I amplitude ratio of less than 0.7 is
Loss of wave V: Lesion in the auditory pathway considered abnormal in the author’s laboratory;

above the level of the superior olivary complex others consider 0.5 abnormal. When the amplitude
up to the level of the inferior colliculus, ipsilat- abnormality is seen in isolation, its significance is
eral or contralateral to side of stimulation. uncertain. However, when it is associated with a
Loss of wave I; waves III and V prolonged: loss of waveform or prolongation of an IPL, it is
Lesion in the auditory pathway below the level more meaningful. Though the exact localization
of the inferior colliculus (or below superior oli- of this amplitude abnormality is uncertain, it most
vary complex, depending on waves III–V IPL), likely represents upper brainstem dysfunction, ipsi-
ipsilateral or contralateral to side of stimulation. lateral or contralateral to side of stimulation.
Prolongation of Interpeak Latencies
The three IPLs evaluated in BAEP are waves I-III, Pediatric Considerations
III–V, and I–V. In some laboratories, instead of
wave V, Vc is used as it is often easier to identify. In children, BAEPs are used to assess not only brain-
An abnormality of IPLs is more indicative of a neu- stem function but also auditory function, as there
rologic dysfunction than prolongation of absolute are few tests to effectively test hearing in this age
latencies. The following can be used as an interpre- group. There are predictable changes in the BAEP
tation guide to prolonged IPLs: as the infant and child’s nervous system matures.
The clinical applications of BAEP in pediatrics will
Prolongation of waves I–III IPL: Lesion between be discussed later in this chapter.
the distal vestibulocochlear nerve and supe-
rior olivary complex (lower brainstem lesion),
ipsilateral to side of stimulation.
Maturational Issues
Prolongation of waves III–V IPL: Lesion between BAEP can be recorded in preterm neonates as early
the superior olivary complex and inferior collicu- as 30 weeks conceptional age. At this age, only wave
lus (upper brainstem), ipsilateral or contralateral I is present. By term, waves I, III, and V are present
to side of stimulation. but are of low amplitude. All five waveforms appear
Prolongation of waves I–V IPL: Lesion between at 3 to 4 months of age, but several of the waves may
the distal vestibulocochlear nerve and inferior col- be fused. The BAEP starts to resemble that of an
liculus, ipsilateral or contralateral to side of stim- adult by 1 year of age, and by age 2 years, the ampli-
ulation. Further localization can be possible if the tudes are also comparable to those of an adult (8–11).
waves I–III or III–V IPL is also prolonged. At term the absolute latencies of all waveforms
are prolonged. Wave I absolute latency reaches
Prolongation of absolute latencies adult values by 2 months, and the wave V reaches
The absolute latencies of waves I, III, and V can be adult values by age 2 to 3 years. Thus, the wave I–V
evaluated. When wave I is prolonged, other absolute IPL is longer than in adults and shorter as central
latencies will also be prolonged. Prolongation of the nervous system myelination occurs.
Methodology Table 3.4 Normative Data for Neonatal and Baby BAEPs
Used in the Author’s Laboratory
There are a few methodological considerations that
should be remembered when obtaining BAEPs in Age (conceptional) I–V IPL (ms)*
children. BAEPs no not include any cortical poten- 30 wk or less 11.4
tial, so maintaining alertness in the patient is not 31–32 wk 8.8
important. For children, in fact, sleep is preferred 33–34 wk 6.7
so that movement and EMG artifact is minimized.
35–36 wk 6.8
Thus, BAEPs should be done after feeding in young
children to facilitate sleep, or if conscious sedation 37–38 wk 6.4
is used, appropriate institutional guidelines should Full term 5.8
be followed. The ground electrode is often placed 3 mo 5.4
on the forehead; in young children the anterior fon- 6 mo 5.1
tanelle may not be closed, and in this situation, the
9 mo 4.9
ground electrode should be placed anterior to this
landmark. Headphones should not be used as they 12 mo 4.8
can collapse the external auditory canal in young 15 mo 4.7
children. A transducer with ear inserts is the pre- 18 mo 4.6
ferred technique for providing the auditory stim-
ulus. The stimulus intensity should be started at *Latency values mean + 3SD or 99% TL
70 dBnHL in young children; if a normal response
BAEP, brainstem auditory evoked potential; IPL, interpeak latency.
is not obtained, the intensity can be increased, but
not more than 95 dBnHL. Because the latencies of can help determine if the hearing loss is conductive
the waveforms are longer, the acquisition time win- or sensorineural. A behavioral audiogram should not
dow should be increased to 15 to 20 ms, rather than be regarded as definitive proof of hearing loss; rather
the 10 ms used for adults. if hearing loss is suspected, it should be confirmed
with a complete audiological evaluation.
In a behavioral audiogram, the patient’s hearing
Interpretation is tested with pure tones at various frequencies and
In children over the age of 2 years, the BAEP is intensities. The sound is delivered through head-
interpreted in the same way as in adults. In younger phones to one ear at a time. The testing is started
children, wave III is often not clearly defined, and with pure tones at a frequency of 500 Hz. The start-
the absolute latency of wave I is affected by hearing stimulation intensity is 20 dBHL. If the patient
ing loss. Thus, typically only the waves I–V IPL can hear the tones at this intensity, the intensity is
are evaluated in young children. This is compared reduced by 10 dB and testing repeated. The lowest
to normative data that are ideally obtained in the intensity that the patient can hear is noted. If the
same laboratory for multiple age groups. However, patient cannot hear the pure tones at 20 dBHL, the
if that is not possible, published normative data can stimulation intensity is increased in increments of
be used, but the methodology has to be kept consis- 10 dB until the hearing threshold is found. Since
tent with that used in the reference laboratory. The this testing is not typically done in a soundproof
waves I–V IPL normative data used in the author’s room, most laboratories used a correction factor to
laboratory, adapted from published sources, are compensate for ambient noise. In the author’s lab-
presented in Table 3.4 (12–14). oratory, this correction factor is dependent on the
frequency of the pure tone tested. This procedure
is repeated at frequencies of 1,000, 2,000, 4,000, and
Behavioral Audiogram 8,000 Hz. Frequencies between these can also be
A behavioral audiogram is a quick bedside assess- tested if there is a need.
ment of hearing that is often performed before a BAEP Normal hearing threshold is 20 dBHL for all
study is conducted. It helps determine if hearing loss frequencies. Young individuals often have lower
is present, and if so, the stimulation intensities can be thresholds, even −10 or −15 dBHL. This implies
appropriately adjusted to compensate. Sometimes it that they can hear better than the average p erson.
Patients with conductive hearing loss have an changes are discussed in the following sections. In
elevated hearing threshold at all frequencies. With infants, BAEP can be used to screen for hearing loss,
sensorineural hearing loss, the hearing loss is and this is also briefly discussed.
greatest with higher frequencies. Sample behav-

ioral audiograms are presented in Figure 3.11A–C. Multiple Sclerosis
BAEPs have been used to detect multiple sclero-
Latency/Intensity Series sis (MS) lesions in the brainstem. If the MS lesions
A BAEP latency/intensity (L/I) series is performed affect the auditory pathways, BAEP abnormalities
to evaluate for possible hearing loss. There are many can be seen. These abnormalities can be loss of
ways to test hearing in adults, and many are more waves III and V or prolongation of IPLs. The wave
sensitive than a L/I series. Consequently, a L/I series III–V IPL is prolonged more often than the wave
is seldom performed. However, when the wave I I–III IPL. The wave V/I amplitude ratio may also
absolute latency is prolonged and c onductive or sen- be affected (15,16). Depending on which waveform
sorineural hearing loss is suspected, this can be eval- is absent or which IPL is prolonged, the lesion can
uated in the evoked potentials laboratory quickly. be further localized.
The principle of the L/I series is that the absolute
latency of wave V increases in a predictable manner Dysmyelinating Conditions
as the stimulation intensity is decreased. This latency
Childhood dysmyelinating diseases can affect the
change is measured in microsecond of latency pro-
BAEP. In Krabbe disease, the upper brainstem is often
longation per dB of intensity reduction (µs/dB).
affected (17). This results in loss of waves III and V
The L/I series is obtained by reducing the
or prolongation of waves III to V IPL early in the
stimulation intensity by increments of 20 dB until
disease course. As the disease advances, the BAEP
30 dBnHL. Thus if the initial BAEP was obtained
abnormalities worsen (18). Adrenoleukodystrophy
with 70 dBnHL stimulation, it is repeated with 50 and
and metachromatic leukodystrophy can also affect
30 dBnHL stimulation. The wave V absolute latency
the BAEP but usually later in the disease course.
is determined at each intensity and the change in
latency per dB is calculated (Figure 3.12A–C). If
the latency change is greater than 60 µs/dB at any Cerebellopontine Angle Tumors
point, a sensorineural hearing loss is suspected Tumors such as vestibular schwannomas and
(Figure 3.13A–C). If the wave V is not recordable at meningiomas in the cerebellopontine (CP) angle
an intensity of 50 or 30 dBnHL, a conductive hear- can affect the BAEP. These tumors can cause hear-
ing loss is suspected (Figure 3.14A–C). If hearing ing loss, and this is seen earlier with vestibular
loss of either type is suspected based on the L/I schwannomas because they arise from the vestib-
series, it should be confirmed with other types of ulocochlear nerve. When the tumor is small, the
audiologic testing. initial change is a prolongation of the waves I–III
IPL. Soon thereafter, the waves I–V IPL may also
become prolonged. If the tumor continues to grow,
Clinical Correlates waves III–V IPL is also affected, but this is only after
BAEP changes can occur with any disease state the wave I–III has also become prolonged. Often,
that affects the auditory pathways. This may occur the BAEP from contralateral ear stimulation also
due to an otologic or neurologic reason. Most often has prolonged waves III–V IPL if the tumor is large.
BAEPs are performed for assessment of the audi- Finally, there may be loss of all waveforms if the
tory pathways in the brainstem. In fact, the BAEP is tumor is very large and distal. If the tumor is more
often used as a surrogate physiologic marker for any proximal, there may be loss of waves II or III and
lesion affecting the brainstem. Though many neu- others after those. After successful surgical resec-
rologic conditions can affect the BAEP, the changes tion, the BAEP may recover to near baseline (19,20).
they produce are not specific for a particular condi-
tion. Rather the BAEP abnormality can help localize Other Brainstem Tumors and Lesions
the lesion, but it cannot identify its cause. The com- Tumors intrinsic to or those causing compression
mon neurologic conditions that can result in BAEP of the brainstem can produce BAEP changes if the
(text continues on page 146)
(A) (B)
Age: 29 years Age: 34 years
Sex: Female Sex: Female
Stimulation rate: 51.1/s Stimulation rate: 51.1/s
Side: Left and right Side: Left and right
Click polarity: Condensation Click polarity: Condensation
Masking intensity: 40 dBnHL Masking intensity: 40 dBnHL
Figure 3.11 (A–C) Behavioral audiogram. (continued )

(C)
Age: 71 years
Sex: Male
Side: Left and right
Discussion: A behavioral audiogram is obtained in all patients undergoing a BAEP study, unless the patient’s ability to cooperate is
limited. Pure tones at frequencies of 500, 1,000, 2,000, 4,000, and 8,000 Hz are delivered at gradually increasing intensities. The
hearing threshold for each frequency is determined. A normal behavioral audiogram (A) has hearing thresholds at or below 20 dBHL
(HL is used instead of nHL as pure tones are used instead of broadband clicks). Patients with conductive hearing loss have elevated
hearing thresholds across all frequencies (B). In sensorineural hearing loss, the hearing thresholds for the higher frequencies are
elevated, but normal for lower frequencies (C).
BAEP, brainstem auditory evoked potential; dBHL, decibels hearing level; dBnHL, decibel above normal hearing level; HL, hearing level; nHL,
normal hearing level.

Age: 7 years
Sex: Male
(+)↑
(+)↑
70 dBnHL
∆ = 26.0 µsec/dB
70 dBnHL
∆ = 19 µsec/dB
50 dBnHL
∆ = 42.0 µsec/dB
50 dBnHL
∆ = 41 µsec/dB
30 dBnHL
30 dBnHL
(A) (B)
Latency/Intensity series Latency/Intensity series
Side: Left Side: Right
Montage: Cz–Ai Montage: Cz–Ai
Stimulation intensity: 70, 50, and 30 dBnHL Stimulation intensity: 70, 50, and 30 dBnHL
Masking intensity: 40, 20, and 0 dBnHL Masking Intensity: 40, 20, and 0 dBnHL
Number of repetitions: 3,000 Number of repetitions: 3,000
Wave V Latency Latency shift Wave V Latency Latency shift
Intensity (dBnHL) Intensity (dBnHL)
(ms) (µs/dB) (ms) (µs/dB)
70 7.34 — 70 7.44 —
50 7.72 19.0 50 7.96 26.0
30 8.54 41.0 30 8.80 42.0
Figure 3.12 (A–C) Normal L/I series. (continued )

(C)
Behavioral audiogram
Discussion: The L/I series is used to evaluate for hearing loss. In this case, the prolongation of the wave V as the stimulation intensity
is reduced is within normal limits (A and B). When the stimulation intensity is reduced from 70 to 50 dBnHL, the latency change is
19 and 26 µs/dB after left and right ear stimulation. With a further decrease from 50 to 30 dBnHL, the latency change is 41 and
42 µs/dB. A latency prolongation of less than 60 µs/dB is considered normal. The behavioral audiogram (C) also suggests lack of
hearing loss.
dBnHL, decibel above normal hearing level; L/I, latency/intensity.

Age: 34 years
Sex: Male
Ear inserts (add 0.9 ms): No
70 dBnHL
70 dBnHL
Δ = 48 μsec/dB
Δ = 70 μsec/dB
50 dBnHL
50 dBnHL
Δ = 87.5 μsec/dB
Δ = N/A
30 dBnHL
30 dBnHL
(A) (B)

Montage: Cz–Ai Montage: Cz–Ai
Masking intensity: 40, 20, and 0 dBnHL Masking intensity: 40, 20, and 0 dBnHL
Intensity (dBnHL) (ms) (µs/dB) Intensity (dBnHL) (ms) (µs/dB)
70 6.89 — 70 6.17 —
50 8.29 70.0 50 7.13 48.0
30 N/A N/A 30 8.88 87.5
Figure 3.13 (A–C) L/I series with suspected sensorineural hearing loss. (continued )
(C)
3K Behavioral audiogram
3K 6K
−5 −5 −10
Discussion: The L/I series in this example is suggestive of sensorineural hearing loss (A and B). After left-sided stimulation, the
latency shift between 70 and 50 dBnHL is abnormally high at 70 µs/dB. With a further reduction of stimulation intensity to
30 dBnHL, a wave V is no longer seen. With right-sided stimulation, the latency shift from 70 to 50 dBnHL is normal at 48 µs/dB,
but with further reduction to 30 dBnHL, the shift is abnormally large at 87 µs/dB. The greater than 60 µs/dB latency shift is sugges-
tive of sensorineural hearing loss. The behavioral audiogram (C) also suggests sensorineural hearing loss bilaterally as the hearing
thresholds for the higher frequencies are elevated above 20 dBHL, while those for lower frequencies are normal. Because of the
high-frequency hearing loss in the behavioral audiogram, notice that additional frequencies (3,000 and 6,000 Hz) were also tested.
dBHL, decibels hearing level; dBnHL, decibel above normal hearing level; L/I, latency/intensity.

Age: 31 years
Sex: Male
Stimulation rate: 51.1/s Filters: 150–3,000 Hz
85 dBnHL
85 dBnHL
#3 #3
#4 #4
∆ = 2.67 µsec/dB ∆ = 34.7 µsec/dB
70 dBnHL
70 dBnHL
#1 #1
#2
#2
∆ = N/A
∆ = NA
50 dBnHL
50 dBnHL
(A) (B)
Montage: Cz–Ai, Cz–Ac Montage: Cz–Ai, Cz–Ac
Click polarity: Rarefaction Click polarity: Rarefaction
Masking intensity: 55, 40, and 20 dBnHL Masking intensity: 55, 40, and 20 dBnHL
Intensity (dBnHL) (ms) (µs/dB) Intensity (dBnHL) (ms) (µs/dB)
85 6.98 — 85 7.22 —
70 7.02 2.67 70 7.74 34.7
50 N/A N/A 50 N/A N/A
Figure 3.14 (A–C) L/I series with suspected conductive hearing loss. (continued )
(C)
Discussion: The L/I series in this example is suggestive of conductive hearing loss (A and B). Stimulation is started at a higher
intensity than usual (85 dBnHL), suggesting that to obtain reliable waveforms a high-stimulation intensity was needed. When the
intensity is reduced to 70 dBnHL, the latency shift is normal at 2.67 and 34.7 µs/dB after left and right ear stimulation. However,
with a further reduction of intensity to 50 dBnHL, a wave V is not seen after stimulation of either side. Since wave V was not seen at
50 dBnHL stimulation, stimulation at 30 dBnHL was not attempted. This suggests conductive hearing loss. The behavioral audiogram
(C) also suggests conductive hearing loss bilaterally as the hearing thresholds for various frequencies are elevated above 20 dBHL.
dBHL, decibels hearing level; dBnHL, decibel above normal hearing level; L/I, latency/intensity.

auditory pathways are affected. Tumors such as childbirth and could affect hearing a ssessment. The
brainstem gliomas, ependymomas, pinealomas, cli- hearing threshold for newborns is 25 to 30 dBnHL.
vus chordomas, thalamic gliomas, and others can BAEP can be rapidly obtained in newborns.
result in BAEP abnormalities. Depending on their Stimulation is performed at 25 to 30 dBnHL, the usual
location, they can cause a loss of waveforms or pro- hearing threshold. Generally, only a presence of waves
longation of one or more of the IPLs. I and V are sought, and if present the baby “passes”
Vascular lesions, such as arteriovenous malfor- the hearing screen. If the waveforms appear very
mations, in the brainstem can affect BAEPs as well. delayed, the BAEP can be repeated in a few months
The location of the lesion dictates the possible BAEP to confirm the presence of waveforms and ensure
abnormality. If the lesion is in the lower medulla, the an improvement in latency. If BAEP waveforms are
BAEP may be normal but the patient may be very absent, the baby is considered to have “failed” the
impaired clinically. Similarly, a lesion in the cortex screening, and the study is repeated in 3 months. If
may not be detected with BAEP as it only evaluates the BAEP remains absent, the child is referred for
the brainstem. Vascular tumors in the brainstem can more detailed audiologic assessment (25,26).
cause a loss of waveforms or IPL prolongations. A baby who fails the hearing screening does
not necessarily have hearing loss. About 20% of
Coma babies undergoing screening fail the screen at
When using BAEPs to evaluate coma, it is import- birth; however, only about 1% to 5% continue to
ant to remember that only brainstem pathways (not have abnormalities at 3 months (27). Thus, while
thalamic and cortical structures) are evaluated. In false-positive results can occur, false-negative
brain death, the brainstem is nonfunctional and results are very uncommon. When BAEPs are pres-
BAEP waveforms are either completely absent or ent, hearing loss is very unlikely. Because of this,
absent after wave I (21). Wave I may occasionally BAEPs are a good screening tool for hearing loss,
be present as it is generated by the distal vestibulo- which when absent can lead to further testing, but
cochlear nerve, which may not be affected in brain when present are reassuring.
death. However, when all waveforms are absent, Automated, handheld devices for recording
technical adequacy of the test must be established AEPs are also available. The stimulus is provided at
before declaring it abnormal. Moreover, there are 35 dBnHL. They record AEPs and transient evoked
many other reasons why all BAEP waveforms might otoacoustic emissions, which are cochlear sounds
be absent, so absence of all waveforms should not made in response to broadband clicks. These
be assumed to be due to brain death. devices can also automatically analyze the results
Comatose states in which the brainstem is spared and appear to have a high sensitivity for detecting
do not cause BAEP abnormalities. Persistent vege- hearing loss (28).
tative state (PVS) often occurs due to thalamocorti-
cal injury. BAEPs are often normal in this situation. Neurophysiologic Intraoperative Monitoring
If cortical injury is severe enough to cause edema BAEPs are commonly used in the OR for patients
and brainstem compression, the BAEP may become undergoing neurosurgical procedures in which
abnormal (22,23). In coma due to metabolic causes, auditory pathways and other structures in the
the brainstem auditory pathways are not structur- brainstem are at risk. An overview of the use of
ally affected, and so the BAEP is normal. BAEPs in the OR is presented in Chapter 5.
Newborn Hearing Screening

While there are many ways to test hearing in adults
Examples
and older children who can participate with a behav- In this section, various BAEP examples are presented
ioral audiological examination, BAEPs remain a (Figures 3.15–3.28, Examples 1–14). Technical issues
useful and quick way to screen for hearing loss in and various types of pathologies are h ighlighted.
newborns. All newborns are screened for hearing BAEP abnormalities are nonspecific, so the abnor-
loss as close to the day of discharge from the nursery mality and how it should be interpreted are dis-
as possible (24). This allows absorption of fluid from cussed. MRI correlations are presented whenever
the middle ear that may have accumulated during possible.
Age: 33 years
Sex: Female
Scale: Amplitude = 0.1 µV/div; Latency = 1.0 ms/div (+↑)
(A)
Side: Left
Rarefaction clicks
Absolute Latencies
V 6.74 0.24 (V–Vn) I–Vc 4.48
Vc 6.80 I–III 2.22
Condensation clicks
Absolute Latencies
V 6.64 0.40 (V–Vn) I–Vc 4.30
Vc 6.80 I–III 2.20

(+↑)
(B)
Side: Right
Rarefaction clicks
Absolute Latencies
V 6.68 0.39 (V–Vn) I–Vc 4.40
Vc 6.76 I–III 2.28
Condensation clicks
Absolute Latencies
V 6.82 0.38 (V–Vn) I–Vc 4.18
Vc 6.86 I–III 2.06

(+↑)
(D)
(C)
Side: Left and right (E)
History: This is a 33-year-old female with complaints of headaches, abnormal facial sensations, visual symptoms, weakness, sense
of disorientation, sexual difficulty, and hearing difficulty. Her MRI had been reported to be concerning for MS. She is referred for
neurophysiologic assessment of auditory pathways for MS evaluation.
Discussion: The BAEPs after (A) left- and (B) right-sided stimulation were normal. The behavioral audiogram was normal (C).
A brain MRI showed scattered hyperintense T2 foci scattered throughout bifrontal subcortical and periventricular white matter.
Lesions were not noted in the brainstem. T2 axial MRI images of the subcortical periventricular white matter (D) and pons (E) are
shown.
The BAEP was performed with fast stimulation rates (51/s) and good morphology and normal latency waveforms were seen. The
wave I absolute latencies were normal, not suggestive of hearing loss. This was corroborated by the behavioral audiogram. All the
IPLs were normal as well, suggesting normal conduction in brainstem auditory pathways. Patients with vague symptoms are often
referred for evoked potential studies to identify silent lesions that may occur with MS. Despite having lesions in the subcortical white
matter, this BAEP study did not help establish lesions in the brainstem auditory pathways.
BAEP, brainstem auditory evoked potential; dBnHL, decibel above normal hearing level; IPL, interpeak latency; MS, multiple sclerosis.

Age: 8 years
Sex: Male
70 dBnHL
(A)
Side: Left
Click polarity: Rarefaction and alternating
Number of repetitions: 4,000 85 dBnHL
Rarefection clicks
Absolute Latencies
I N/A N/A (I–In) Interpeak Latencies
Vc N/A I–III N/A
Alternating clicks
Absolute Latencies
Vc N/A I–III N/A

(B)
Side: Right
Click polarity: Rarefaction and alternating
(+↑) Stimulation intensity: 70 dBnHL
70 dBnHL
85 dBnHL
Rarefection clicks
Absolute Latencies
Vc N/A I–III N/A
Alternating clicks
Absolute Latencies
Vc N/A I–III N/A

(C)
History: This is an 8-year-old male with Tay Sachs disease who underwent a bone marrow transplant 7 years ago. Currently his
neurological exam is significant for severe developmental regression, quadriplegia, and spasticity. This study is being done for
neurophysiologic assessment.
Discussion: The BAEPs are abnormal after left (A) and right (B) ear stimulation because of absence of all waveforms. A behavioral
audiogram could not be performed.
BAEPs were initially attempted with rarefaction clicks at 70 dBnHL. No response was seen with stimulation of either ear. A wave I
is tagged after stimulation of both sides; however, this is not a reproducible wave and thus should not be considered a wave I. As
no clear wave I is seen, a hearing deficit cannot be ruled out, so the stimulation intensity is increased to 85 dBnHL with alternating
clicks. Despite this increase in intensity, no waveforms are seen. A wave I is again tagged, but it also is not reliable.
The T2 weighted brain MRI shows severe atrophy and extensive white matter changes (C). No clear lesion is seen in the brainstem.
The absence of all BAEP waveforms after stimulation of the left and right sides could be due to a severe bilateral hearing loss or
a technical problem preventing adequate stimulation of the cochlea. A behavioral audiogram could not be performed to confirm
hearing loss. Technical problems should be excluded by double checking the setup. The presence of a stimulus artifact in the rarefac-
tion clicks makes technical issues less likely. Another possibility is severe, bilateral brainstem dysfunction. However, in that case, at
least a wave I should have been present as that arises from the distal vestibulocochlear nerve. The most likely interpretation for these
findings is that absence of all waveforms is consistent with bilateral lesions distal to the distal part of the vestibulocochlear nerve,
that is, ear- or hearing-related problems. Absence of clear MRI abnormalities in the brainstem is consistent with this interpretation.

Age: 8 years
Sex: Male
(A)
Side: Left
Rarefaction clicks
Absolute Latencies
Vc N/A I–III 3.42
Condensation clicks
Absolute Latencies
Vc N/A I–III N/A

(B)
Side: Right
(+↓)
Rarefaction clicks
Absolute Latencies
Vc N/A I–III N/A
Condensation clicks
Absolute Latencies
Vc N/A I–III N/A

(C)
History: This is a 8-year-old male with metachromatic leukodystrophy who underwent a bone marrow transplant 6 years ago.
She is able to follow simple commands, but is nonverbal. She has diffuse quadriplegia and hyperreflexia. This study is done to eval-
uate for neurophysiologic assessment.
Discussion: The BAEP is abnormal after left ear stimulation because of absence of wave V with rarefaction clicks and absence
of waves III and V with condensation clicks (A). Wave I is also prolonged with rarefaction clicks, but it is normal with condensa-
tion clicks. After right ear stimulation, waves III and V are absent after both rarefaction and condensation clicks (B). A behavioral
audiogram could not be performed.
The T2 weighted sagittal brain MRI shows remarkable brain atrophy, including in the brainstem (C). Diffuse white matter disease
was also noted.
The absence of all BAEP waveforms after wave I suggests a lesion proximal to the vestibulocochlear nerve but before the superior
olivary complex (lower brainstem). Presence of a wave III after left ear stimulation with rarefaction clicks is of uncertain significance,
but may imply that the lesion(s) is proximal to (or more severe proximal to) the superior olivary complex. Because the findings are
bilateral, the brainstem lesion are also likely bilateral/diffuse in the lower brainstem. The brainstem atrophy seen on MRI is consistent
with this interpretation.

Age: 2 years
Sex: Male
Stimulation rate: 11.1/s ↑(+)
(A)
Side: Left
Rarefaction clicks
Absolute Latencies
V 9.29 0.19 (V–Vn) I–Vc N/A
Vc 9.33 I–III N/A
N/A (V/I ratio) III–Vc 4.30
Condensation clicks
Absolute Latencies
V 9.23 0.21 (V–Vn) I–Vc N/A
Vc 9.10 I–III N/A

(B)
↑(+) Side: Right
Rarefaction clicks
Absolute Latencies
V 10.00 0.19 (V–Vn) I–Vc N/A
Vc 9.90 I–III N/A
Condensation clicks
Absolute Latencies
V 9.39 0.24 (V–Vn) I–Vc N/A
Vc 9.57 I–III N/A

(C)
History: This is a 2-year-old male born prematurely, had intraventricular hemorrhages at birth, was later diagnosed with
alpha-mannosidosis, who is now presenting with delayed development and dysmorphic facial features. He is being considered for
bone marrow transplantation. This study is performed to assess for brainstem function.
Discussion: The BAEPs are abnormal after left (A) and right (B) ear stimulation because of absence of wave I and prolonged latency
of waves III and V. The waves III–Vc IPL is also prolonged bilaterally. A behavioral audiogram could not be performed.
A T1 weighted sagittal brain MRI image shows a Chiari malformation with 12 mm herniation of the cerebellar tonsils below the
foramen magnum (as indicated by the markings on the MRI scan) (C). The MRI was otherwise normal for age.
The absence of wave I after stimulation of the left and right ears with preserved waves III and V suggests a bilateral brainstem lesion
that is more severe between the vestibulocochlear nerve and the superior olivary complex (lower brainstem). Prolongation of the
waves III–Vc IPL also suggests involvement of the brainstem between the superior olivary complex and inferior colliculus (upper
brainstem). Since both BAEPs from both ears are affected, the lesion is likely to be bilateral. How the MRI abnormalities relate to the
BAEPs is not entirely clear; however, the Chiari malformation could be causing a bilateral lower brainstem dysfunction. The upper
brainstem dysfunction is harder to explain based on the MRI.

Age: 48 years
Sex: Male
(+↑)
(A)
Side: Left
Absolute Latencies
V 6.60 0.33 (V–Vn) I–Vc 4.20
Vc 6.63 I–III 2.25

(B)
Side: Right
(+↑) Masking intensity: 55 dBnHL
Absolute Latencies
Vc N/A I–III N/A

(C)
History: This is a 48-year-old male with hearing loss in his right ear. His neurological exam is otherwise normal. Neuroimaging
revealed a petroclival meningioma. This study is being done to evaluate presurgery for suitability for neurophysiologic intraoperative
monitoring (NIOM).
Discussion: The BAEP after left ear stimulation is normal (A). After right ear stimulation, the BAEP is abnormal due to absence of all
waveforms (B). A behavioral audiogram was not performed as this was a presurgical evaluation. A rate of 31/s was used as this was
a presurgical evaluation, and during surgery, the stimulation rate would be 31/s.
The FLAIR coronal brain MRI shows a large right sided petroclival meningioma (C).
The absence of all waveforms after right ear stimulation suggests a severe dysfunction of the distal right vestibulocochlear nerve or
the ear more distal to it. Since the abnormality is seen only on one side, this is a lateralized abnormality on the right side. The MRI
lesion supports this interpretation as the tumor is likely affecting the distal vestibulocochlear nerve and causing both the hearing
loss and absent BAEP response. The value of NIOM using BAEP monitoring on the operative ear will be limited due to absence of
that response preoperatively.
BAEP, brainstem auditory evoked potential; dBnHL, decibel above normal hearing level; FLAIR, fluid-attenuated inversion recovery.

Age: 18 years
Sex: Male
(+↑)
(A)
Side: Left
Absolute Latencies
I Approx. 1.9 — (I—In) Interpeak Latencies
Vc N/A I–III N/A

(B)
Side: Right
(+↑)
Absolute Latencies
V 8.39 0.16 (V–Vn) I–Vc 4.98
Vc 7.02 I–III 2.67

(C)
History: This is an 18-year-old male who presented with a 2-month history of vision problems. He was noted to have papilledema.
The remainder of his neurological examination was normal. This study was done as presurgical evaluation for suitability for NIOM.
Discussion: The BAEP after left ear stimulation shows a possible wave I, though it is not tagged (A). It is present in the Cz–Ai
and Ac–Ai derivation and not in the Cz–Ac derivation, making it most likely to be a wave I. No other waveform is seen after
left-side stimulation, however. After right-sided stimulation, all waveforms are present; however, the waves I–Vc IPL is prolonged
(B). The waves I–III and III–Vc IPLs are not prolonged. Note that stimulation rate was 31/s as this was a presurgical evaluation and
the same rate would be used in the OR. Because of the faster rate of stimulation, the waves I–III and III–Vc IPLs are not abnormal. A
behavioral audiogram was not performed as this was a presurgical evaluation.
The contrast-enhanced, coronal brain MRI shows a large left-sided posterior fossa tumor (C). It appears to be distorting the middle
and upper brainstem.
After left ear stimulation, the absence of waves III and V is suggestive of a lesion on that side affecting the proximal v estibulocochlear
nerve or lower brainstem below the level of the superior olivary complex. The much subtler abnormality after right-sided stimu-
lation may suggest either an independent lesion affecting the auditory pathways arising from the right ear between the distal
vestibulocochlear nerve and the inferior colliculus or that a large left-sided lesion is affecting the contralateral response as well.
Given the degree of asymmetry of the abnormality and that single lesions are more common, the latter possibility is more likely.
Thus, the interpretation of this study would be a lesion affecting the proximal left vestibulocochlear nerve that may also be causing
brainstem dysfunction. Certainly the MRI shows that a large tumor is affecting the left cerebellopontine angle and most likely is
affecting the proximal left vestibulocochlear nerve. Moreover, the degree of brainstem distortion could be accounting for the slightly
prolonged waves I–Vc IPL after right-sided stimulation. These BAEPs may still be of value to monitor during surgery as the nonoper-
ative side BAEP is present, and the lesion is large enough that the nonoperative BAEP could be affected during the surgery as well.
BAEP, brainstem auditory evoked potential; dBnHL, decibel above normal hearing level; IPL, interpeak latency; NIOM, neurophysiologic
intraoperative monitoring.

Age: 43 years
Sex: Female
Filters: 150–3,000 Hz +↑
(A)
Side: Left
Rarefaction clicks
Absolute Latencies
V 6.42 0.23 (V–Vn) I–Vc 4.12
Vc 6.54 I–III 2.20
Condensation clicks
Absolute Latencies
V 6.50 0.60 (V–Vn) I–Vc 4.02
Vc 6.56 I–III 2.04

(B)
Side: Right
Rarefaction clicks
Absolute Latencies
Vc N/A I–III N/A
Condensation clicks
Absolute Latencies
Vc N/A I–III N/A

(C) (D)
History: This is a 43-year-old female with complaints of weakness in left lower extremity for several months. There is numbness associated with the
weakness. She is having problems with balance recently. She is being evaluated for MS and this is a work up to identify other silent lesions.
Discussion: The BAEP after left-sided stimulation has normal absolute latencies and IPLs (A). However, the waves V/I amplitude ratio is slightly
reduced. The BAEP after right ear stimulation showed a robust wave I but no other waveforms. The behavioral audiogram is suggestive of a very
mild right-sided hearing loss, likely conductive (C).
The brain MRI scan is normal. The brainstem did not show signs of demyelination. A T2 weighted brain MRI scan is shown (D).
The loss of waves III and V after right ear stimulation is suggestive of a lesion at or distal to the ipsilateral superior olivary complex (likely lower
brainstem). The abnormal wave V/I amplitude ratio after left ear stimulation is more difficult to localize. As the only abnormality, its significance is
uncertain. If it is accompanied by other abnormalities, it may suggest upper brainstem dysfunction. However, it is possible that the lesion detected by
right ear stimulation is causing the amplitude abnormality. The behavioral audiogram shows a very mild right-sided hearing loss, which appears to be
conductive. This hearing loss is not seen on the BAEP, as the absolute latency of wave I is within normal limits after right ear stimulation. Thus, the most
appropriate interpretation of this study would be a lesion in the auditory pathways arising from the right ear, distal to the superior olivary complex.
The normal brain MRI scan does not demonstrate the lesion. In a patient being evaluated for MS, such a BAEP abnormality may indicate a silent lesion.

Age: 4 years
Sex: Female
Scale: Amplitude = 0.1 µV/div; Latency = 1.5 ms/div (+)↑
(A)
Side: Left
Condensation clicks
Absolute Latencies
V 10.30 0.23 (V–Vn) I–Vc 6.63
Vc 10.40 I–III 4.53
Rarefaction clicks
Absolute Latencies
V 10.40 0.22 (V–Vn) I–Vc 6.75
Vc 10.50 I–III 4.56

(B)
Side: Right
(+)↑ Masking intensity: 40 dBnHL
Condensation clicks
Absolute Latencies
V 10.40 0.26 (V–Vn) I–Vc 7.23
Vc 10.50 I–III 4.62
Rarefaction clicks
Absolute Latencies
V 10.50 0.16 (V–Vn) I–Vc 6.90
Vc 10.60 I–III 4.59

(C) (D)
History: This is a 4-year-old female with history of metachromatic leukodystrophy who underwent a bone marrow transplant at
1 year of age. He has significant cognitive impairment and quadriparesis. He is nonverbal. This study is to assess brainstem function.
Discussion: The BAEPs after (A) left- and (B) right-sided stimulation is similar. It shows prolonged waves I–Vc and I–III IPLs bilaterally.
Waves III–Vc IPL is normal after left ear stimulation and slightly prolonged after right-sided stimulation. Wave I has an abnormally
prolonged absolute latency bilaterally. The findings for both condensation and rarefaction clicks are similar, though wave I is seen
better with rarefaction clicks. A behavioral audiogram was not performed.
A brain MRI scan shows diffuse supratentorial white matter, signaling abnormality and atrophy. Clear lesions in the brainstem are
not noted. An axial FLAIR MRI scan through the brainstem (C) shows subcortical white matter changes and a relatively normal
appearing brainstem, while an axial T2 weighted image shows the periventricular white matter changes (D).
The BAEP shows remarkable prolongation of the waves I–Vc IPL, with more localized prolongation of waves I–III. This suggests a
lesion of the brainstem between the distal vestibulocochlear nerve and superior olivary complex (lower brainstem). The brainstem
between the superior olivary complex and the inferior colliculus (upper brainstem) appears to be less involved. Since the prolon-
gations are bilateral, the lesion is diffuse. Such lesions are common in dysmyelinating diseases, even if they are not clearly seen on
MRI. The MRI shows the typical white matter lesions of metachromatic leukodystrophy. Additionally, bilateral wave I prolongations
suggest possible hearing loss; however, that could not be confirmed with a behavioral audiogram as the patient was unable to
cooperate with it.
BAEP, brainstem auditory evoked potential; dBnHL, decibel above normal hearing level; FLAIR, fluid-attenuated inversion recovery; IPL,
interpeak latency.

Age: 8 years
Sex: Male
Filters: 150–3,000 Hz (+↑)
Scale: Amplitude = 0.1 lV/div; Latency = 1.5 ms/div
(A)
Side: Left
Rarefaction clicks
Absolute Latencies
V 8.13 0.23 (V–Vn) I–Vc 5.61
Vc 8.16 I–III 2.73
Condensation clicks
Absolute Latencies
V 8.13 0.30 (V–Vn) I–Vc 5.67
Vc 8.16 I–III 2.73

(B)
Side: Right
(+↑) Stimulation intensity: 70 dBnHL
Rarefaction clicks
Absolute Latencies
V 8.46 0.18 (V–Vn) I–Vc 5.85
Vc 8.49 I–III 2.67
Condensation clicks
Absolute Latencies
V 8.34 0.27 (V–Vn) I–Vc 5.64
Vc 8.43 I–III 2.49

(C) (D)
History: This is an 8-year-old male with history of metachromatic leukodystrophy who underwent a bone marrow transplant at
7 years of age. He has mild cognitive impairment, and is able to communicate and follow simple commands. This study is to assess
brainstem function.
Discussion: The BAEPs after (A) left- and (B) right-sided stimulation show prolongation of almost all IPLs. However, the most signif-
icant prolongation is of the waves III–Vc IPL (along with waves I–Vc IPL). Though the waves I–III IPLs are also prolonged (except after
right-sided stimulation with condensation clicks), they are not as severely affected as the other IPLs. Abnormalities after right-sided
stimulation are more severe. These findings are noted in both the rarefaction and condensation clicks. Additionally, with rarefaction
clicks the waves V/I amplitude is also reduced. A behavioral audiogram was not performed.
The brain MRI scan shows diffuse white matter signal abnormality in the subcortical white matter. Subtle hyperintensities were
noted in the brainstem as well. An axial FLAIR MRI scan shows subcortical white matter changes (C). Another axial FLAIR image
through the upper pons shows subtle hyperintensities (arrows) (D).
These BAEPs show a widespread, bilateral abnormality in the brainstem by virtue of all IPLs being prolonged. However, the selectively
greater effect on waves III–Vc IPL suggests that the most severe abnormality is between the superior olivary complex and the infe-
rior colliculus (upper brainstem) bilaterally. This is supported by the low waves V/I amplitude ratio seen with the rarefaction clicks.
Brainstem involvement is often seen with neurophysiologic testing in children with metachromatic leukodystrophy, even when it
cannot be seen on neuroimaging. In this child, the metachromatic leukodystrophy is affecting not only the subcortical white matter,
but also the white matter tracts in the brainstem.
BAEP, brainstem auditory evoked potential; dBnHL, decibel above normal hearing level; FLAIR, fluid-attenuated inversion recovery; IPL,
interpeak latency.

Age: 43 years
Sex: Male
Scale: Amplitude = 0.1 µV/div; Latency = 1.5 ms/div ↑(+)
(A)
Side: Left
Absolute Latencies
V 7.89 0.32 (V–Vn) I–Vc 5.46
Vc 7.96 I–III 3.57

(B)
Side: Right
↑(+) Stimulation intensity: 70 dBnHL
Absolute Latencies
V 6.60 0.32 (V–Vn) I–Vc 4.29
Vc 6.78 I–III 2.31

(C)
History: This is a 43-year-old male who presented with episodic dizziness, loss of balance, and headaches for the last 6 months. He
denied any hearing loss. The remainder of his neurological examination was normal. This study was done as presurgical evaluation
for suitability for NIOM.
Discussion: The BAEP after left ear stimulation shows prolonged waves I–Vc and I–III IPLs (A). The absolute latency of wave I is nor-
mal. After right sided stimulation, the IPLs and absolute latency of wave I are normal. A behavioral audiogram was not performed
as this was a presurgical evaluation.
The brain MRI scan shows a 1.7 × 1.4 cm contrast enhancing mass in the left cerebellopontine angle that extends into the left
internal auditory canal, most consistent with vestibular schwannoma. A contrast enhanced axial image is shown demonstrating the
mass (C).
These BAEPs show a lateralized lesion after left-sided stimulation. The prolongation of the waves I–Vc and I–III IPLs suggests a lesion
affecting the auditory pathways between the distal vestibulocochlear nerve and the superior olivary complex (lower brainstem).
This type of abnormality can be seen with vestibular schwannomas. As they get larger, the IPLs may increase and eventually there
is loss of waveforms (particularly waves III and V). With very large tumors, contralateral BAEP may be affected. In this example,
this tumor produces a localized abnormality. The normal absolute latency of wave I after left ear stimulation is consistent with the
patient’s lack of hearing loss. The BAEPs are adequate for monitoring during surgery.

Age: 43 years
Sex: Male
(+↑)
(A)
Side: Left
Absolute Latencies
V 7.06 0.14 (V–Vn) I–Vc 4.47
Vc 7.05 I–III 2.22

(B)
Side: Right
(+↑)
Absolute Latencies
I 2.49 0.14 (I–In)
III N/A
V 8.49 0.13 (V–Vn)
Vc 8.52
0.93 (V/I ratio)
Interpeak Latencies
I–Vc 6.03
I–III N/A
III–Vc N/A
History: This is a 43-year-old male who presented with a 2-month

history of headaches, nausea, and vomiting. He denied hearing loss.
The remainder of his neurological examination was normal. This study
was done as presurgical evaluation for suitability for NIOM.
Discussion: The BAEP after left ear stimulation was normal except
for a slightly reduced waves V/I amplitude ratio (A). In isolation, the
significance of this finding is uncertain. After right ear stimulation, the
BAEP shows prolonged waves I–Vc IPL and wave III is absent (B). The (C)
absolute latency of wave I is normal. A behavioral audiogram was not
performed as this was a presurgical evaluation. colliculus, but the absence of wave III implies a particularly severe
involvement at the level of the superior olivary complex. The lateralized
The brain MRI scan shows a large enhancing mass within the fourth abnormality implies the lesion is mostly on one side. The side which is
ventricle extending through the right foramen of Luschka. Associated abnormal is a little harder to determine as the waves I–III and III–Vc IPLs
tonsillar herniation through the foramen magnum and ventricular cannot be determined. However, the absence of wave III suggests that
dilation is noted as well. A T2 weighted axial image of the brainstem the predominant abnormality is at the level of the superior olivary com-
shows the large enhancing mass (C). plex or distal to it, making it more likely that the lesion is ipsilateral to
the side of stimulation (right-sided lesion). The normal absolute latency
The BAEP abnormalities after right ear stimulation suggest a diffuse of wave I after left ear stimulation is consistent with the patient’s lack
lesion between the distal vestibulocochlear nerve and the inferior of hearing loss. The BAEPs are adequate for monitoring during surgery.

Age: 5 years
Sex: Male
(A)
Side: Left
Rarefaction clicks
Absolute Latencies
V 6.50 0.06 (V–Vn) I–Vc 4.37
Vc 6.56 I–III 2.13
Condensation clicks
Absolute Latencies
V 6.41 0.08 (V–Vn) I–Vc 4.43
Vc 6.59 I–III 2.19
Figure 3.26 (A–B) Example 12. (continued )

Interpeak Latencies
I–Vc 4.11
(+↑)
I–III 1.83
III–Vc 2.28
Condensation clicks
Absolute Latencies
I 2.25 0.19 (I–In)
III 4.20
V 6.33 0.07 (V–Vn)
Vc 6.33
0.37 (V/I ratio)
Interpeak Latencies
I–Vc 4.08
I–III 1.95
III–Vc 2.13
History: This is a 5-year-old male with Gaucher’s disease who

presents for evaluation of brainstem function. He has abdomi-
nal discomfort but denies any neurological symptoms. He does
not appear to have hearing loss according to parents.
(B)
Side: Right Discussion: The BAEPs after (A) left- and (B) right-sided stim-
Click polarity: Rarefaction and condensation ulation with rarefaction and condensation clicks show normal
Stimulation intensity: 70 dBnHL IPLs and absolute latencies to wave I. The waves V/I amplitude
Masking intensity: 40 dBnHL ratio is reduced bilaterally. A behavioral audiogram was not
Number of repetitions: 4,000 performed due to patient’s inability to cooperate.
Rarefaction clicks
Neuroimaging was not performed in this patient. Individuals
Absolute Latencies with Gaucher’s disease usually have abdominal, skeletal, and
hearing complaints.
I 2.10 0.46 (I–In) The only abnormality in this BAEP study is the reduced waves
V/I amplitude ratio. Because all the IPLs and absolute latencies
III 3.93 are normal, the significance of the amplitude abnormality is
V 6.09 0.11 (V–Vn) uncertain. This study could be interpreted as normal with the
amplitude abnormality described or as mildly abnormal with
Vc 6.21
the clinical significance of the abnormality as uncertain. The
0.24 (V/I ratio) most important aspect of interpretation would be to tem-
per the amplitude abnormality because no other abnormality
BAEP, brainstem auditory evoked potential; dBnHL, decibel above is seen.
normal hearing level; IPL, interpeak latency.
Figure 3.26 (A–B) (continued )

Age: 7 years
Sex: Male
(A)
Side: Left
Rarefaction clicks
Absolute Latencies
V 8.01 0.31 (V–Vn) I–Vc 4.23
Vc 8.07 I–III 2.49
Condensation clicks
Absolute Latencies
V 8.07 0.26 (V–Vn) I–Vc 4.23
Vc 8.10 I–III 1.92

(B)
Side: Right
(+↑)
Rarefaction clicks
Absolute Latencies
V 8.94 0.30 (V–Vn) I–Vc 4.02
Vc 9.03 I–III 2.40
Condensation clicks
Absolute Latencies
V 9.15 0.29 (V–Vn) I–Vc 3.96
Vc 9.18 I–III 2.49

(C)
History: This is a 7-year-old male with Niemann–Pick disease who underwent a bone marrow transplant at the age of 3 years.
She is currently significantly cognitively impaired, quadriplegic, and nonverbal. She is presented for evaluation of brainstem function.
Discussion: The BAEPs after left-sided stimulation show a prolonged wave I absolute latency with rarefaction and condensation
clicks (A). IPLs are normal with rarefaction clicks, and with condensation clicks wave I–III and I–Vc IPLs are normal, while the wave
III–Vc IPL is slightly prolonged. After stimulation of the right ear, the wave I absolute latency is prolonged with rarefaction and
condensation clicks (B). All the IPLs are normal with rarefaction and condensation clicks. The waves V/I amplitude ratios are normal
bilaterally. A behavioral audiogram was not performed due to patient’s inability to cooperate.
An MRI scan of the brain shows diffuse atrophy but no focal lesions. A coronal T1 weighted image is shown to demonstrate the
atrophy (C).
The main abnormality in this BAEP is the prolongation of the wave I absolute latency after left and right ear stimulation. This is most
consistent with bilateral lesions distal to the distal vestibulocochlear nerve (hearing loss). The only IPL that is prolonged is the wave
III–Vc IPL after left ear stimulation with condensation clicks. Further review of that waveform indicates that the wave III morphology
is not well defined, and it could have been tagged a little to the left or right. This would have made the IPLs normal. Because of
this, the borderline IPL “abnormality” should be discounted. Despite the neurological illness, this patient’s auditory pathways in the
brainstem do not demonstrate a lesion, but there does appear to be possible bilateral hearing loss. This should be confirmed with
audiological testing.

Age: 47 years
Sex: Male
↑(+)
(A)
Side: Left
Rarefaction clicks
Absolute Latencies
V 6.56 0.27 (V–Vn) I–Vc 4.22
Vc 6.64 I–III 2.24
Condensation clicks
Absolute Latencies
V 6.74 0.27 (V–Vn) I–Vc 4.18
Vc 6.82 I–III 2.18

(B)
Side: Right
↑(+) Masking intensity: 40 dBnHL
Rarefaction clicks
Absolute Latencies
V 8.38 0.21 (V–Vn) I–Vc 4.36
Vc 8.48 I–III 2.58
Condensation clicks
Absolute Latencies
V 8.50 0.21 (V–Vn) I–Vc 4.30
Vc 8.66 I–III 2.36

(C)
Side: Right
↑(+)
Alternating clicks
Absolute Latencies
V 8.26 0.23 (V–Vn) I–Vc 4.34
Vc 8.30 I–III 2.68

6k
(E)
1500
(D) (F)
History: This is a 47-year-old male with a history of left lower extremity weakness and numbness who developed hearing loss in his right ear about 8 months
ago. His neurological examination shows mild left upper and lower extremity weakness and reduced vibration and position in his left lower extremity. Bedside
hearing assessment demonstrates right-sided hearing loss. His brain MRI shows white matter lesions. He is referred for evaluation of brainstem function.
Discussion: The BAEPs after left ear stimulation at 70 dBnHL show normal IPLs and a normal wave I absolute latency with rarefaction and conden-
sation clicks (A). After right ear stimulation at 70 dBnHL, all the IPLs are normal, but the wave I absolute latency is prolonged with rarefaction and
condensation clicks (B). Right ear stimulation is repeated at 85 dBnHL, and the wave I absolute latency improves slightly, but it remains prolonged (C).
The IPLs remain normal. A behavioral audiogram shows a mild decrease in hearing for higher frequencies in the left ear. On the right side, there is a
significant decrease in hearing at all frequency levels (D).
MRI of the brain shows numerous T2 signal abnormalities in juxtacortical locations bilaterally, suggestive of a demyelinating disease. No clear
brainstem lesion is noted. An axial T2 weighted MRI is shown demonstrating linear subcortical periventricular white matter lesions (E). An axial image
through the pons does not show any brainstem hyperintensities (F).
The only abnormality in this BAEP study is the prolongation of the absolute latency of wave I after stimulation of the right ear, which persisted despite
a higher intensity of stimulation. After left ear stimulation, the wave I absolute latency is normal. The IPLs after stimulation of both ears are normal,
as is the waves V/I amplitude ratio. This finding is suggestive of a lesion distal to the distal right vestibulocochlear nerve, that is, hearing problem. This
is corroborated by the behavioral audiogram, which shows a significant decrease in hearing in all frequency ranges on the right side. The left-sided
behavioral audiogram suggests only mild hearing loss, and that may be why the absolute latency of wave I after left ear stimulation is not prolonged.
This study was performed to evaluate for brainstem dysfunction in a patient suspected of having MS. It is important to not interpret this study as
suggestive of a brainstem lesion, as might occur in an MS patient. Rather, this study is suggestive of hearing loss on the right side.
BAEP, brainstem auditory evoked potential; dBnHL, decibel above normal hearing level; IPL, interpeak latency; MS, multiple sclerosis.

Conclusions
BAEPs are very small potentials that can be difficult is important to correctly identify the waveforms
to record. With proper attention to detail and appro- and interpret the data. Most uses for BAEP are to
priate patient preparation they can be recorded evaluate neurologic disorders; however, they can
successfully in most patients. Changes in stimula- be used in certain situations to assess hearing, such
tion parameters can change the BAEP remarkably. as newborn hearing screening. BAEPs are also used
Analysis of the waveforms in various derivations extensively in the OR.
References
1. Legatt AD, Arezzo JC, Vaughan HG Jr. The anatomic Niedermeyer’s Electroencephalography. 6th ed.
and physiologic bases of brain stem auditory evoked Philadelphia, PA: Wolters Kluwer Lippincott
potentials. Neurol Clin. 1988;6:681-704. Williams & Wilkins; 2011:1057-1082.
2. ACNS. Guideline 9C: guidelines on short-latency 13. Levy SR. Brain stem auditory evoked potentials in
auditory evoked potentials. J Clin Neurophysiol. pediatrics. In: Chiappa KH, ed. Evoked Potentials
2006;23:157-167. in Clinical Medicine. 3rd ed. Philadelphia, PA:
3. Erwin CW, Husain AM. Brainstem auditory evoked Lippincott-Raven Publishers; 1997:269-282.
potentials. In: Ebersole JS, Pedley TA, eds. Current 14. Thivierge J, Cote R. Brain-stem auditory
Practice of Clinical Electroencephalography. 3rd ed. evoked response: normative values in children.
Philadelphia, PA: Lippincott Williams & Wilkins; Electroencephalogr Clin Neurophysiol. 1990;77:309-313.
2003:864-891. 15. Chiappa KH, Harrison JL, Brooks EB, Young RR.
4. ACNS. Guideline 9B: guidelines on visual evoked Brainstem auditory evoked responses in 200
potentials. J Clin Neurophysiol. 2006;23:138-156. patients with multiple sclerosis. Ann Neurol.
5. Misulis KE. Spehlmann’s Evoked Potential Primer. 1980;7:135-143.
2nd ed. Boston, MA: Butterworth-Heinemann; 1994. 16. Prasher DK, Gibson WP. Brain stem auditory
6. Moncho D, Poca MA, Minoves T, et al. Interside evoked potentials: a comparative study of monaural
latency differences in brainstem auditory and versus binaural stimulation in the detection of mul-
somatosensory evoked potentials. Defining upper tiple sclerosis. Electroencephalogr Clin Neurophysiol.
limits to determine asymmetry. J Clin Neurophysiol. 1980;50:247-253.
2015;32:424-427. 17. Aldosari M, Altuwaijri M, Husain AM. Brain-
7. Chiappa KH, Hill RA. Brain stem auditory evoked stem auditory and visual evoked potentials in
potentials: interpretation. In: Chiappa KH, ed. Evoked children with Krabbe disease. Clin Neurophysiol.
Potentials in Clinical Medicine. 3rd ed. Philadelphia, 2004;115:1653-1656.
PA: Lippincott-Raven Publishers; 1997:199-249. 18. Husain AM, Altuwaijri M, Aldosari M. Krabbe
8. Eggermont JJ. Development of auditory evoked disease: neurophysiologic studies and MRI correla-
potentials. Acta Otolaryngol. 1992;112:197-200. tions. Neurology. 2004;63:617-620.
9. Goldie WD. The brainstem auditory evoked poten- 19. Moffat DA, Baguley DM, Hardy DG, Tsui YN.
tial in infants and children. J Clin Neurophysiol. Contralateral auditory brainstem response abnor-
1992;9:394-407. malities in acoustic neuroma. J Laryngol Otol.
1989;103:835-838.
10. Picton T, Taylor MJ, Durieux-Smith A. Brain
stem auditory evoked potentials in pediatrics. In: 20. Moller MB, Moller AR. Brainstem auditory evoked
Aminoff MJ, ed. Electrodiagnosis in Clinical Medicine. potentials in patients with cerebellopontine angle
New York, NY: Churchill Livingstone; 1992:537-569. tumors. Ann Otol Rhinol Laryngol. 1983;92:645-650.
11. Starr A, Amlie RN, Martin WH, Sanders S. 21. Goldie WD, Chiappa KH, Young RR, Brooks EB.
Development of auditory function in newborn Brainstem auditory and short-latency somatosen-
infants revealed by auditory brainstem potentials. sory evoked responses in brain death. Neurology.
Pediatrics. 1977;60:831-839. 1981;31:248-256.
12. Husain AM. Evoked potentials in children and 22. Karnaze DS, Marshall LF, McCarthy CS, et al.
infants. In: Schomer DL, Lopes da Silva FH, eds. Localizing and prognostic value of auditory
evoked responses in coma after closed head injury. 26. Shannon DA, Felix JK, Krumholz A, et al. Hearing
Neurology. 1982;32:299-302. screening of high-risk newborns with brainstem
23. Uziel A, Benezech J. Auditory brain-stem responses auditory evoked potentials: a follow-up study.
in comatose patients: relationship with brain-stem Pediatrics. 1984;73:22-26.
reflexes and levels of coma. Electroencephalogr Clin 27. Johnson JL, White KR, Widen JE, et al. A multicenter
Neurophysiol. 1978;45:515-524. evaluation of how many infants with perma-
24. Committee on Practice and Ambulatory nent hearing loss pass a two-stage otoacoustic
Medicine, Bright Futures Steering Committee. emissions/automated auditory brainstem response
Recommendations for preventive pediatric health newborn hearing screening protocol. Pediatrics.
care. Pediatrics. 2007;120:1376. 2005;116:663-672.
25. Hall JW III, Smith SD, Popelka GR. Newborn hearing 28. van Straaten HL, Tibosch CH, Dorrepaal C, et al.
screening with combined otoacoustic emissions and Efficacy of automated auditory brainstem response
auditory brainstem responses. J Am Acad Audiol. hearing screening in very preterm newborns.
2004;15:414-425. J Pediatr. 2001;138:674-678.
4
Somatosensory Evoked Potentials
Somatosensory evoked potentials (SEPs) are

obtained with the electrical stimulation of periph-
Anatomy
eral nerves. Virtually any peripheral nerve can SEPs elicited after electrical stimulation of periph-
be stimulated and evoked responses are recorded eral nerves are transmitted through the dorsal
over the neuroaxis. However, in clinical prac- column pathways (1). The volley produced by
tice median, ulnar, tibial, and peroneal nerves are stimulation of a peripheral nerve enters the spinal
stimulated in most situations. SEP waveforms cord through the dorsal horns and ascends in the
are recorded from several levels of the neuroaxis, ipsilateral dorsal column tract without synapsing
which include peripheral nerve, spinal roots, spinal in the spinal cord. As they ascend in the dorsal col-
cord, brainstem, and cerebral cortex and include umn tracts, the fibers from the lower limbs are more
both near- and far-field potentials. In this chapter, mesial than the ones from the upper limbs. In the
SEP classification, anatomy, methodology, inter lower medulla, the dorsal column pathway fibers
pretation, and special considerations in pediatrics from the upper extremities synapse on the nucleus
and clinical applications are discussed. SEP exam- cuneatus, whereas fibers from the lower extremity
ples with neuroimaging correlations are explained synapse on the nucleus gracilis. From these two
following the text. nuclei arise second-order neurons that decussate
and ascend as the medial lemniscus. The medial
lemniscus ascends to the ventral posterior lateral
Classification (VPL) nucleus of the thalamus where it synapses.
The commonest classification of SEP is based on the Third order neurons arise from the VPL and ascend
peripheral nerve stimulated to obtain a response. in the internal capsule to the somatosensory cortex
Thus, as noted earlier, the commonest types of SEPs (contralateral to the side of stimulation). This path-
obtained are median, ulnar, tibial, and peroneal way is illustrated in Figure 4.1. The dorsal spino-
nerve SEP. cerebellar tract and anterolateral columns may also
Other peripheral nerves, such as the radial be involved with transmitting impulses that con-
nerve, can also be stimulated and result in radial tribute to SEP waveforms; however, their contribu-
nerve SEP. Assessment of genitourinary functions are considerably less than the dorsal column
tion is occasionally performed with pudendal pathway.
nerve SEP. Sensory cranial nerves can also be
used to elicit SEP, such as trigeminal nerve SEP.
Individual dermatomes can be stimulated and
Methodology
produce dermatomal SEP. These types of SEP are Electrical stimulation with rectangular pulses is
used only in special clinical situations and are not used most often to elicit SEP. Other stimulation tech-
discussed in this text. niques such as tactile, position change, and thermal
191
including sedative ones like p

henobarbital, do not
Somatosensory affect the SEP except when they are administered
cortex in very high doses. Cold temperature can reduce
nerve conduction velocity. If a patient’s limb has
Thalamocortical radiation
been exposed to cold for some time, the peripheral
Ventral posterior
nerve conduction may be slowed, producing a pro-
lateral nucleus longation in latency. Patients should be kept ade-
quately warm to minimize the effect of temperature
Medial lemniscus on nerve conduction.
Stimulation Parameters
Nucleus Nucleus
cuneatus gracilis A rectangular wave electrical stimulus is used to
activate peripheral nerves. Various properties of
the rectangular wave can be changed, and these
Dorsal column tract changes can alter the SEP waveforms.
Upper limb Intensity
dorsal root
ganglion Most evoked potential machines deliver electrical
stimulation in constant current, and the stimulation
intensity is measured in milliampere (mA). When
stimulating a mixed nerve, the stimulus inten-
Lower limb sity should be increased slightly above the motor
dorsal root threshold (when muscle twitch is noted). Typically,
ganglion this is about 5 to 20 mA when the nerves are healthy,
higher for diseased nerves. The higher the stimula-
Ipsilateral Contralateral tion intensity, the more the discomfort experienced
by the patient.
The motor threshold intensity is lower than the
Figure 4.1 Diagrammatic representation of the dorsal stimulation intensity used for motor nerve conduc-
column pathways arising in the upper and lower limbs. tion studies (NCS). For motor NCS, a supramaxi-
Boxes represent sites of neural cell bodies, while the lines mal stimulation intensity is used. This ensures that
represent the fiber tracts. all motor fibers of the nerve are activated. Motor
threshold intensity does not necessarily activate
stimulation can also be employed, but these are not
all motor fibers. All motor fibers of a peripheral
used in routine clinical practice.
nerve do not need to be stimulated to obtain an
SEP. Central SEP waveforms reach peak amplitude
Patient Variables before their peripheral counterparts. This principle
Because SEPs are obtained with electrical stimu- is known as “central amplification”; activation of
lation, they are associated with discomfort at the some peripheral fibers can cause activation of more
stimulation site. The patient must be told about this central fibers, resulting in activation of all central
before the test is started to reduce anxiety. Proper fibers before all peripheral fibers (Figures 4.3A–E
attention to technique can help reduce the pain and 4.4). Because of central amplification, lower
(discussed in the following). The patient should be stimulation intensity is needed for SEP than for
lying down or resting comfortably in a recliner. This motor NCS.
serves to reduce muscle activation and electromyo- When stimulating sensory nerves, the intensity is
graphic (EMG) artifact. Though every effort should increased to about 2 to 3 times the sensory thresh-
be made to keep the patient as relaxed as possible, it is old. This becomes challenging when stimulating
important that he or she remain awake. Drowsiness areas of sensory loss. When sensation cannot be felt,
and sleep can reduce the amplitude of the corti- the stimulation intensity is set at a level comparable
cal SEP waveforms, especially for tibial and pero- to mixed nerve stimulation or the intensity needed
neal nerve SEP (Figure 4.2A and B). Medications, to elicit the sensory NCS.
Chapter 4. Somatosensory Evoked Potentials ■ 193
Age: 26 years
Sex: Female
Scale: Amplitude = 0.5 μV/div; Latency = 8 ms/div
(−)↑
?P37
(A)
Side: Left
Absolute Latencies
Waveform Latency (ms) Amplitude (μV)
PF 5.6
N34 26.1 PF–P37 24.8
P37 30.4 0.02 (P37–N45) LP–P37 14.7
Figure 4.2 (A and B) Effects of sleep. (continued )

(−)↑
(B)
Side: Right
Absolute Latencies
PF 5.1
LP 14.6
Interpeak Latencies
P31 22.6
P37 36.0 0.23 (P37–N45) PF–P37 21.4
N45 42.1 LP–P37 30.9
Discussion: The amplitude of the P37 waveforms is remarkably different between the two sides, but the amplitude of the PF wave-
forms appears comparable (exact measurement not listed). This suggests that the amplitude difference in the P37 waveforms is not
due to inadequate peripheral nerve stimulation on the side that resulted in the lower amplitude waveform (left). Rather, in this case,
the patient fell asleep during electrode application and remained asleep during acquisition of the left tibial nerve SEP (A). When the
right tibial nerve SEP (B) was being acquired, she awoke and remained alert for that part of the test. The difference in the P37 wave-
form amplitude is likely due to the variance in the sleep/wake state. Amplitude asymmetry in SEP is seldom considered abnormal.
LP, lumbar potential; PF, popliteal fossa; SEP, somatosensory evoked potential.

Age: 49 years
Sex: Male
(A)
Side: Right
Stimulation intensity: 6 mA
Absolute Latencies
EP 11.0 0.55 (EP–EPn) Interpeak Latencies
N13 14.7 0.40 (N13–N13n) Waveforms Latency (ms)
P13 17.2 EP–N20 10.1
N18 18.2 0.41 (P13–N18) EP–P13 3.9
N20 21.1 0.77 (N20–P22) P13–N20 3.7
Figure 4.3 (A–E) Effects of central amplification. (continued )

(B) (C)
Side: Right Side: Right
Stimulation duration: 0.2 ms Stimulation duration: 0.2 ms
Stimulation intensity: 8 mA Stimulation intensity: 10 mA
Absolute Latencies Absolute Latencies

Waveform Latency (ms) Amplitude (μV) Waveform Latency (ms) Amplitude (μV)
EP 11.2 2.44 (EP–EPn) EP 11.2 5.26 (EP–EPn)
N13 13.5 1.31 (N13–N13n) N13 14.8 1.76 (N13–N13n)
P13 15.1 P13 14.5
N18 18.2 2.18 (P13–N18) N18 18.6 2.36 (P13–N18)
N20 20.6 2.28 (N20–P22) N20 20.2 2.79 (N20–P22)
Interpeak Latencies Interpeak Latencies

Waveforms Latency (ms) Waveforms Latency (ms)
EP–N20 9.4 EP–N20 9.0
EP–P13 2.3 EP–P13 5.7
P13–N20 5.5 P13–N20 3.6

(D) (E)
Side: Right Side: Right
Stimulation intensity: 12 mA Stimulation intensity: 14 mA

EP 11.3 6.37 (EP–EPn) EP 11.3 6.49 (EP–EPn)
N13 14.5 1.75 (N13–N13n) N13 14.5 1.74 (N13–N13n)
P13 15.5 P13 14.6
N18 18.3 2.29 (P13–N18) N18 18.8 2.56 (P13–N18)
N20 20.4 2.91 (N20–P22) N20 20.3 3.04 (N20–P22)

EP–N20 9.12 EP–N20 9.0
EP–P13 3.20 EP–P13 3.2
P13–N20 4.90 P13–N20 5.7
Discussion: As peripheral nerve stimulation intensity is increased, progressively more nerve fibers are recruited and contribute to
the various SEP waveforms. The amplitude of the EP waveform increases as progressively more nerve fibers are activated. Once the
potential reaches the central nervous system, nearly all nerve fibers in that tract are activated even with relatively low stimulation
intensity. Thus, central waveforms (P13/14, N18, and N20) reach a maximum amplitude much more quickly and at lower stimulation
intensity than the peripheral (EP) waveform. This phenomenon is referred to as “central amplification.” In the example above, the
stimulation intensity is increased from 6 ma (A) to 8 mA (B) to 10 mA (C) to 12 ma (D) to 14 mA (E). With increasing stimulation
intensity, the EP waveform’s amplitude continues to increase even when the N20 (and the P13/14–N18) waveform’s amplitude sta-
bilizes. This is presented graphically in Figure 4.4.
EP, evoked potential; SEP, somatosensory evoked potential.

EP 6.37 6.49
6
N13
N18
5 5.26
N20
Amplitude (μV)
3
2.91 3.04
2.44 2.79
2.28 2.56
2.36 2.29
2 2.18
1.76 1.75 1.74
1.31
1
0.77
0.55
0.410.4
0
6 8 10 12 14
Stimulation intensity (mA)
Figure 4.4 This graph shows central amplification from a median nerve SEP. This is a graphical representation of the data
presented in Figure 4.3A–E. The lines represent the amplitude of the EP, N13, P13/14–N18, and N20 waveforms, respectively,
with progressively increasing stimulus intensities. The N20, N18, P13/14, and N13 waveforms reach maximum amplitude much
sooner than the EP waveform, demonstrating “central amplification” of amplitude.
Duration limb SEP is 100 ms, while for upper limb stimula-
The duration of the rectangular electrical pulse can tion it is 30–50 ms). Normative data should be used
be changed between 50 and 500 μs. The longer the that corresponds to the rate of stimulation.
pulse duration, the lower the stimulus intensity Electrodes
needed to activate the peripheral nerve. Shorter Surface or needle electrodes can be used to stimu-
pulse durations activate motor fibers preferentially. late peripheral nerves. Surface electrodes can be of
Commonly, stimulus duration is set at 200 μs, and the self-adhesive type, or they may be small discs,
intensity is 5 to 20 mA. This helps activate the sen- about 5 mm in diameter, that are attached to skin
sory fibers of the mixed nerve and limits the dis- with glue or tape. They are placed about 2 to 3 cm
comfort felt by the patient. However, for diseased apart. Needle electrodes can also be used and are
nerves, a longer stimulus duration may be needed, 1 cm in length and less than 1 mm in width. They are
along with a higher intensity. placed about 1 cm apart. Since clinical SEP studies
are typically performed in awake individuals, nee-
Rate
dle electrodes are seldom used because of the dis-
The rate of stimulation can affect the SEP wave- comfort in insertion. Surface and needle electrodes
forms. For upper and lower limb stimulation, a rate are placed along the path of the nerve to be stimu-
of approximately 5/s is used. Occasionally for upper lated. The cathode is proximal and anode distal.
limb stimulation, a faster rate of approximately 30/s Before electrodes are placed on the skin, the
can be used, but the waveforms are of longer latency, site should be prepared by rubbing with a mild
lower amplitude, and less well defined (Figure 4.5A abrasive. Electrode impedance should be below
and B). The low amplitude subcortical waveforms 5 kOhm (2). This improves signal quality and also
are particularly affected by the faster rates and are reduces patient discomfort.
more difficult to resolve. For lower limb stimula-
tion, faster rates of stimulation are not used as the Unilateral/Bilateral
recording time after each stimulation is longer than Unilateral stimulation of upper and lower
for upper limb stimulation (analysis time for lower limb peripheral nerves is used most often for
clinical studies. Each dorsal column pathway can To record lower limb (tibial and peroneal nerve)
be assessed independently with unilateral stimu- SEP, the following electrodes are placed:
lation, allowing for lateralization of abnormalities.
Bilateral simultaneous stimulation can be used if • PFd and PFp: Popliteal fossa distal and proxi-
determination of laterality of abnormality is not mal, both behind the knee
important; this is seldom the case in the clinical • T12S: Thoracic 12 spine, on the posterior aspect
laboratory. Because more nerve fibers are acti- of the T12 spine
vated with simultaneous bilateral stimulation, the • IC: Iliac crest, on the IC
amplitude of the SEP is usually higher. This may • C5S: Cervical 5 spine, on the posterior aspect
be helpful when using SEP during neurophysi- of the neck at the C5 spine
ologic intraoperative monitoring (NIOM) when • CPi: Centroparietal ipsilateral to side of stimu-
waveforms after unilateral stimulation are of low lation, between C3/4 and P3/4 electrodes
amplitude. In this situation, lateralization of an • CPc: Centroparietal contralateral to side of
acute spinal cord lesion is usually less important. stimulation, between C3/4 and P3/4 electrodes
Bilateral stimulation becomes even more problem- • CPz: Centroparietal midline, between Cz
atic if the peripheral nerve conduction velocities are and Pz
different, such as when one nerve has an entrap- • Fpz: Frontopolar midline
ment neuropathy. • Ground: Vertex (or can be placed elsewhere on
the scalp)
Recording Parameters
SEP waveforms are recorded from several different The IC electrode is used as a silent reference
locations as they ascend up the dorsal column path- for recording the lumbar potential (LP). The C5S
way. By measuring latencies of various waveforms, electrode also serves as reference electrode for
and the latencies between them (interpeak latencies lower limb SEP. Though the SEP volley travels
[IPL]), conduction abnormalities in the nervous sys- under this electrode, it is so small that it cannot
tem can be detected and localized. be recorded.
Electrode placement
Recording electrodes are strategically placed along Recording montage
the neuroaxis so that appropriate SEP responses The montage for upper and lower limb SEPs should
can be recorded. have a minimum of four channels, but if avail-
To record the upper limb (median and ulnar able additional channels can be added to further
nerves) SEP, the following electrodes are placed: enhance the waveforms. Each channel can have
either a referential or bipolar derivation. The active
• EP: Erb’s point bilaterally, which is located electrode (Input 1, G1) for both types of derivation
about 2.5 cm above and lateral to the most is placed in a location optimal to record the wave-
mesial point of the clavicle form of interest. In a referential derivation, the ref-
• C5S: Cervical 5 spine, on the posterior aspect erence electrode (Input 2, G2) is placed far enough
of the neck at the C5 spine away so that it is “silent,” but not so far as to intro-
• CPi: Centroparietal ipsilateral to side of stimu- duce unwanted noise. In a bipolar derivation, both
lation, between C3/4 and P3/4 electrodes electrodes record some waveforms of the SEP. The
• CPc: Centroparietal contralateral to side of waveforms that are common to both electrodes are
stimulation, between C3/4 and P3/4 electrodes canceled (common mode rejection) by the differ-
• AA: Adam’s apple electrode, on the anterior ential amplifier, and only the unique waveforms
aspect of the neck are seen. Referential and bipolar derivations are
• Ground: Vertex (or can be placed elsewhere on included in most montages, which are constructed
the scalp) to record the ascending volley as it traverses the
nervous system.
The AA electrode is not essential, but can help The American Clinical Neurophysiology Society
visualize the P13 waveform (positive end of the (ACNS) has recommended montages for both
N13) potential. upper and lower limb SEPs (2). However, clinicians
Age: 49 years
Sex: Male
(A)
Side: Left
Absolute Latencies
P13 14.5 EP–N20 9.6
N18 17.6 1.71 (P13–N18) EP–P13 3.5
N20 20.6 1.62 (N20–P22) P13–N20 6.2
Figure 4.5 (A and B) Effects of rate of stimulation. (continued )

(B)
Side: Left
Absolute Latencies
P13 16.1 EP–N20 10.0
N18 18.3 1.67 (P13–N18) EP–P13 4.9
N20 21.3 1.26 (N20–P22) P13–N20 5.1
Discussion: Rate of stimulation of the peripheral nerve can affect the SEP in several ways. Generally, with increasing rates of stimu-
lation, there is an increase in absolute latencies, reduction in amplitude and reduced reproducibility of the waveforms. Median nerve
SEP acquired with a stimulation rate of 5.7/s (A) and 31.1/s (B) are shown. In the latter example, the SEP waveforms have a slightly
longer latency and lower amplitude. The reproducibility of the waveforms is better with the slower rate. When the SEP is abnormal,
fast rates may show a more remarkable difference in latency and amplitude than is seen here.

can expand and alter these montages as a ppropriate record the P13/14 and N18 waveforms. The com-
for their needs. posite waveform will make it difficult to identify
The recommended montage for upper limb the individual components.
(median and ulnar nerve) SEP is listed as follows The recommended montage for lower limb (tibial
(channels are numbered from the bottom up as that and peroneal nerve) SEP is (channels are numbered
is how they are typically displayed): from the bottom up as that is how they are typically
displayed) listed as follows:
Channel 4: CPc–CPi
Channel 3: CPi–EP contralateral (EPc) (reference) Channel 4: CPi–Fpz
Channel 2: C5S–EPc (reference) Channel 3: CPz–Fpz
Channel 1: EP ipsilateral (EPi)–EPc (reference) Channel 2: Fpz–C5S (reference)
Channel 1: T12S–IC (reference)
The channels labeled as reference are referen-
tial recordings, whereas the CPc–CPi derivation In the author’s laboratory, the lower limb SEP
is a bipolar one. The latter derivation records the montage is often expanded as follows:
subcortical waveforms from both electrodes but
the cortical waveform from only the CPc electrode. Channel 7: CPi–C5S (reference)
Consequently, only the cortical waveform will be Channel 6: CPc–C5S (reference)
seen in that channel. Channel 5: Fpz–C5S (reference)
In the author’s laboratory, the upper limb SEP Channel 4: CPz–Fpz
montage is often expanded as follows: Channel 3: CPz–C5S (reference)
Channel 2: T12S–IC (reference)
Channel 6: C5S–AA Channel 1: PFp–PFd
Channel 5: AA–EPc (reference)
Channel 4: CPc–CPi The extra channels allow better identification of
Channel 3: CPi–EPc (reference) the cortical P37 waveform (Figure 4.7). This wave-
Channel 2: C5S–EPc (reference) form is usually best recorded from the midline or
Channel 1: EPi–EPc (reference) ipsilateral to side of stimulation (contralateral to
somatosensory cortex activated). This is because
The extra channels allow better analysis of the this potential arises from the mesial surface of the
N13 waveform (Figure 4.6). Table 4.1 lists the chan- cortex and projects contralaterally. Sometimes,
nels of a typical montage of upper extremity SEP and it can be best seen contralateral to side of stimu-
the waveform(s) best seen in each of them. Another lation (ipsilateral to somatosensory cortex acti-
commonly used derivation is C5S–Fpz. This deriva- vated). Thus, having multiple electrodes recording
tion is discouraged as the C5S electrode will record the P37 waveform (CPi, CPc, and CPz) is useful.
the N13 waveform, while the Fpz electrode will Moreover, this montage allows evaluation of the
Table 4.1 The Channels of an Upper Extremity SEP and the P37 waveform in referential and bipolar deriva-
Waveform(s) Seen in Those Channels tions. Channels 3 and 7 (and sometimes channel
6; referential derivations) will display the P37 and
Montage Waveform(s)
the subcortical waveforms (P31, N34). Channel 4
C5S–AA N13 (bipolar derivation) will display only the P37 wave-
form as the subcortical (P31, N34) waveforms will
AA–Epc Adam’s apple
cancel. Channel 5 will display only the subcortical
CPc–CPi N20 (P31, N34) waveforms (Figure 4.8). Table 4.2 lists
the various montages of lower extremity SEP and
CPi–EPc P13/14, N18
the waveform(s) best seen in each of them. When
C5S–EPc N13 peroneal nerve SEPs are recorded, the PFp–PFd
channel is not recorded as that is the site of stimu-
Epi–Epc Erb’s point
lation (Figure 4.9).
Age: 42 years
Sex: Female
Side: Right Interpeak Latencies
Stimulation intensity: 8.8 mA Waveforms Latency (ms)
Number of repetitions: 1,500 EP–N20 8.6
Median nerve conduction velocity (median nerve to EP):
43 m/s EP–P13 3.3
P13–N20 5.3
Discussion: This is an example of a normal median nerve SEP.

A normal ulnar nerve SEP would be very similar. The montage
(−)↑ has been expanded to include six channels. It contains both
bipolar and referential channels to better define the various
waveforms. The derivations of the channels are listed to the
left of the graph. Starting from the bottom, the channels
are as follows (for generators of the various waveforms, see
text):
REP–LEP: This channel records the EP waveform. This is a ref-

erential derivation as the active electrode records the EP wave-
form and the reference electrode does not record any activity of
interest.
C5S–LEP: This channel records the N13 waveform. This is a

referential derivation as the active electrode records the N13
waveform and the reference electrode does not record any
activity of interest.
CP4–LEP: This channels records the P13/14 and N18 waveforms.

This is a referential derivation as the active electrode records the
P13/14 and N18 waveforms and the reference electrode does
not record any activity of interest.
CP3–CP4: This channel records the N20 waveform. This is a

bipolar derivation as the active electrode records the P13/24,
N18, and N20 waveforms and the reference electrode records
the P13/14 and N18 waveforms. Because of cancellation, only
the N20 waveform is displayed.
AA–LEP: This channel records the positive end of the dipole

Absolute Latencies of the N13 waveform (Adam’s apple [AA] waveform). This
is a referential derivation as the active electrode records the
AA waveform and the reference electrode does not record
EP 8.6 any activity of interest. This waveform is marked with a light
arrow.
N13 12.1
P13 11.9 C5S–AA: This channel records the N13 waveform. This is a
bipolar derivation as the active electrode records the N13
N18 15.1
waveform and the reference electrode records the AA wave-
N20 17.2 3.92 (N20–P22) form (positive end of the same dipole), resulting in a higher
amplitude waveform than seen in the C5S–LEP derivation. This
EP, evoked potential; SEP, somatosensory evoked potential. waveform is marked with a heavy arrow.
Figure 4.6 Recording montage of upper limb SEP.

Age: 15 years
Sex: Female
Interpeak Latencies
Stimulation rate: 5.7/s Waveforms Latency (ms)
Filters: 30–3,000 Hz PF–P37 27.3
Side: Left LP–P37 14.9
Discussion: This is an example of a normal tibial nerve SEP. The
montage has been expanded to include eight channels. It con-
tains both bipolar and referential channels to better define the
Tibial nerve conduction velocity (tibial nerve to T12S):
various waveforms. The derivations of the channels are listed to
46 m/s
the left of the graph. Starting from the bottom, the channels are
as follows (for generators of the various waveforms, see text):
LPd–LPp: This channel records the PF waveform. This is a bipo-
(−)↑
lar derivation as the active and reference electrodes record a
traveling waveform passing beneath them.
T12S–IC: This channel records the LP waveform. This is a ref-
erential derivation as the active electrode records the LP wave-
form and the reference electrode does not record any activity
of interest.
CPz–C5S: This channels records the P31, N34, and P37 wave-
forms. This is a referential derivation as the active electrode
records the P31, N34, and P37 waveforms and the reference
electrode does not record any activity of interest. The reference
electrode is on C5S, under which the ascending volley travels,
but it is of such low amplitude that this electrode practically
serves as a silent reference.
CPz–Fpz: This channel records the P37 waveform. This is a
bipolar derivation as the active electrode records the P31, N34,
and P37 waveforms and the reference electrode records the
P31 and N34 waveforms. Because of cancellation, only the P37
waveform is displayed.
CP3–CP4: This channel records the P37 waveform (arrow). This
is a bipolar derivation as the active electrode records the P31,
N34, and P37 waveforms and the reference electrode records
the P31 and N34 waveforms. Because of cancellation, only the
P37 waveform is displayed.
Fpz–C5S: This channels records the P31 and N34 waveforms
(arrow). This is a referential derivation as the active electrode
records the P31 and N34 waveforms and the reference elec-
trode does not record any activity of interest. This derivation is
similar to the CPz–C5S except that it does not include the P37
Absolute Latencies and makes identification of the P31, N34 easier.
CP4–C5S: This channels records the P31 and N34 waveforms
PF 7.9 (arrow). This is a referential derivation as the active electrode
LP 20.3 records the P31 and N34 waveforms and the reference elec-
trode does not record any activity of interest.
P31 26.6
CP3–C5S: This channels records the P31, N34, and P37
N34 30.2 waveforms (arrows). This is a referential derivation as the
P37 35.2 7.35 (P37–N45) active electrode records the P31, N34, and P37 waveforms
and the reference electrode does not record any activity of
N45 43.8
interest. In some patients, the P37 is seen better ipsilateral
LP, lumbar potential; PF, popliteal fossa; SEP, somatosensory
to the site of waveform generation, contralateral to the site
evoked potential. of stimulation.
Figure 4.7 Recording montage of lower limb SEP.

Age: 16 years
Sex: Female
(−↑)
(A)
Side: Left
Absolute Latencies
PF 8.2
N34 36.6 PF–P37 33.4
P37 41.4 1.09 (P37–N45) LP–P37 20.0
Figure 4.8 (A and B) Atypical P37 waveform projection. (continued )

(−↑)
(B)
Side: Right
Absolute Latencies
PF 8.6
P31 29.2
N34 33.1
P37 39.6 2.02 (P37–N45) PF–P37 31.0
N45 48.1 LP–P37 17.6
Discussion: The projection of the P37 waveform vector can be in different directions. Since this waveform arises from the mesial pari-
etal surface, it most often projects to the midline or to the contralateral side. Thus, the P37 waveform is best recorded in the midline
or on the side ipsilateral to side of stimulation. Sometimes the vector may project differently, and the P37 waveform is recorded best
ipsilateral to side of generation, contralateral to side of stimulation. In this example, after left tibial nerve stimulation (A), the best P37
waveform is seen in the CP4–C5S derivation, ipsilateral to side of generation and contralateral to side of stimulation. Conversely, it is
not seen well in the CP3–C5S and CPz–Fpz derivations. Similarly, after right tibial nerve stimulation (B), the best P37 waveforms are
seen in the CP3–C5S and CPz–Fpz derivations, ipsilateral to side of generation (and midline) and contralateral to side of stimulation.
In the CP4–C5S derivation, the P37 waveform is not seen. This atypical projection of the P37 waveform vector is not unusual.
LP, lumbar potential; PF, popliteal fossa.

Age: 51 years
Sex: Female
Scale: Amplitude = 0.5 μV/div; Latency = 8 ms/div ↑(−)
Side: Left
Peroneal nerve conduction velocity (peroneal nerve to
T12S): 39 m/s
Absolute Latencies
LP 18.9
P31 26.9
N34 30.0 Interpeak Latencies
P37 33.0 1.07 (P37–N45) Waveforms Latency (ms)
N45 40.8 LP–P37 14.1
Discussion: The waveforms recorded after peroneal nerve stimulation are similar to those obtained after tibial nerve stimulation.
There are a few key differences. The stimulation is performed behind the knee (PF). Thus, the PF waveform is not recorded, as the
stimulating electrodes are placed near that location. The first waveform recorded is the LP waveform. Thereafter all the subsequent
waveforms recorded (subcortical and cortical) are similar to that of the tibial nerve SEP. However, the absolute latency of these wave-
forms is 10 ms shorter because the stimulating electrodes are closer to the recording electrodes. As such, technically, the P31, N34,
and P37 waveforms should be called the P21, N24, and P27 waveforms, but to avoid confusion, the original (tibial nerve) nomencla-
ture is kept. The normative date for peroneal nerve SEP should be acquired in the same way as tibial nerve SEP; alternatively, if that
is not available, absolute latency values 10 ms shorter than tibial nerve SEP can be used as approximates. For example, the upper
limit of normal for the absolute latency of the P37 (P27) waveform would be 37 ms (per the author’s laboratory, the upper limit of
normal for the P37 waveform after tibial nerve stimulation is 47 ms). Note that the IPLs will not be different than the tibial nerve
SEP, as that is measured between two peaks.
LP, lumbar potential; PF, popliteal fossa; SEP, somatosensory evoked potential.
Figure 4.9 Peroneal nerve stimulation.

Table 4.2 The Channels of a Lower Extremity SEP and the over the neural structure transmitting the impulse;
Waveform(s) Seen in Those Channels in this situation, a traveling wave is recorded.
Montage Waveform(s)
Comparable waveforms, with comparable genera-
tors, are recorded after upper and lower limb stim-
CPi–C5S P31, N34, P37 ulation. The various waveforms and their potential
CPc–C5S P31, N34 neural generators are listed in Table 4.3.
Fpz–C5S P31, N34
Upper limb
CPz–Fpz P37 After stimulation of median and ulnar nerves, sim-
CPz–C5S P31, N34, P37 ilar waveforms are seen (2,3) (Table 4.3). The first
T12S–IC LP waveform seen is the EP waveform, indicating pas-
sage of the impulse through the brachial plexus.
PFp–PFd PF
This is a negative potential occurring at about 9 ms,
LP, lumbar potential; PF, popliteal fossa; SEP, somatosensory evoked hence it is also called the N9 waveform. The N13
potential.
waveform (negative potential at 13 ms) is thought
to arise from the cervical spinal cord as the sensory
Amplifier settings
fibers transition from cervical dorsal roots into the
The low- and high-frequency filter settings for dorsal column. The next potential is referred to as
SEP are typically set to 30 to 3,000 Hz. Reducing the P13 or P14 (or P13/14) waveforms and most
the high-frequency filter to 1,000 Hz can produce likely arises as the dorsal column fibers synapse at
a phase lag in the latency of the waveforms. Since the nucleus cuneatus in the lower medulla. These
many of the waveforms are of lower frequency, are positive waveforms occurring at approximately
increasing the low-frequency filter to 200 Hz will 13 to 14 ms. Both the P13 and P14 waveforms
greatly reduce the amplitude of the waveforms occur in close proximity to each other, and both
(Figure 4.10A–C). are thought to arise from the region of the nucleus
The analysis time for upper limb SEP is 30 to cuneatus. In this text, they will be referred to as the
50 ms and for lower limb it is 80 to 100 ms. This P13/14 waveform. The N18 waveform is the next
allows adequate display of all the waveforms even potential that is labeled and is a negative waveform
if they have prolonged latency. The number of repe- occurring at about 18 ms. It arises from the VPL of
titions needed to average depends on the amplitude the thalamus as the ascending fibers synapse in this
of the responses, but typically 1,000 to 3,000 repeti- structure. Finally, once the dorsal column pathway
tions are needed. Every SEP should be replicated at fibers synapse in the somatosensory cortex, the N20
least once to ensure reproducibility of the response. waveform is produced. This is a negative potential
occurring at about 20 ms. See Figure 4.6 for an illus-
Interpretation tration of the various waveforms.
The first step in interpretation of SEPs is verification Lower limb

of presence of the various expected waveforms. The The first waveform recorded after tibial nerve stim-
latencies should be compared to normative data ulation is the popliteal fossa (PF) waveform as the
and to the contralateral limb to check for asymme- signal traverses the posterior aspect of the knee. This
try. Additionally, when available, the upper and waveform is a negative potential occurring at about
lower limb SEPs should be evaluated together to 8 ms, and is also known as the N8 waveform. The LP
better define an abnormality. waveform arises from the cauda equine and lower
spinal cord as the sensory fibers transition from the
Expected Waveforms lumbosacral roots into the dorsal column. It is a neg-
The various SEP waveforms that are recorded in ative potential and occurs at about 22 ms (also called
routine clinical use are produced by the ascend- the N22 waveform). A cervical potential is not record-
ing volley reaching a synapse or the fiber pathway able from the neck electrodes as the potential is of
changing direction (tract turning from a horizontal very low amplitude. The P31 waveform is recorded
position to a vertical one). Additionally, waveforms when the dorsal column tract arising from the tib-
can be recorded if electrodes are placed directly ial nerve synapses with the nucleus gracilis, while
Age: 49 years
Sex: Male
Stimulation rate: 4.10
(A)
Side: Right
Absolute Latencies
P13 15.1 EP–N20 9.4
N18 18.2 2.19 (P13–N18) EP–P13 2.3
N20 20.4 2.30 (N20–P22) P13–N20 5.5
Figure 4.10 (A–C) Effects of changing filter settings. (continued )

(B)
Side: Right
Absolute Latencies
P13 15.9 EP–N20 9.1
N18 19.4 2.15 (P13–N18) EP–P13 3.7
N20 20.6 2.31 (N20–P22) P13–N20 4.6

(C)
Side: Right
Absolute Latencies
Interpeak Latencies
EP 10.8
N13 13.5
P13 14.8 EP–N20 N/A
N18 N/A N/A (P13–N18) EP–P13 2.7
N20 N/A N/A (N20–P22) P13–N20 N/A
Discussion: Filters can have a profound effect on the SEP. Data should always be acquired with the same filter settings as the
normative data. The first figure (A) is acquired with usual filter settings (30–3,000 Hz). When the high-frequency filter is reduced
to 1,000 Hz (B), the morphology of the waveforms changes. Moreover, a “phase delay” is introduced into the latency, making
the waveforms appear longer in latency. Notice the longer latency in especially the N13, P13/14, and N18 waveforms. If the low-
frequency waveform is increased to 200 Hz (C), there is a remarkable reduction in the amplitude of the waveforms. The N13 and
P13/14 waveforms are difficult to identify, and the N18 and N20 waveforms cannot be reliably seen (though they are tagged on
the graph). The higher low-frequency filter introduces a “phase advance,” and the waveforms appear shorter in latency (note the
EP and N13 waveforms).

Table 4.3 Upper and Lower Limb SEPs Waveform of the waveforms are not a reliable indicator of
Neural Generators central conduction as they include conduction
Upper Extremity Lower Extremity of the impulse along the peripheral nerve as well
EP Brachial plexus PF Peripheral nerve as along the central nervous system pathways.
N13 Cervical spinal cord LP Cauda equina, lower Consequently, IPL between various components of
spinal cord the SEP are evaluated to assess conduction along
P13/14 Nucleus cuneatus P31 Nucleus gracilis various parts of the nervous system. The IPLs are
N18 Thalamus N34 Thalamus compared to normative data. They should also be
N20 Somatosensory P37 Somatosensory cortex compared side-to-side to ensure symmetry. Even if
cortex the IPL is within normal limits but there is a greater
than normal side-to-side asymmetry, the study
EP, evoked potential; LP, lumbar potential; PF, popliteal fossa; SEP,
somatosensory evoked potential. is considered abnormal. Reduced or asymmetric
amplitude of the waveforms is not considered an
abnormality.
the N34 waveform results from the synapse of the Ideally, normative data should be obtained in
dorsal column pathway at the VPL of the thalamus. each laboratory performing evoked potentials stud-
These are positive and negative potentials occurring ies. However, if this cannot be done, published
at 31 ms and 34 ms, respectively. Thereafter, the P37 normative data can be used. However, the exact
waveform (positive potential occurring at 37 ms) is methodology used in the reference laboratory must
produced when the dorsal column pathway syn- be followed.
apses in the somatosensory cortex. See Figure 4.7 for
an illustration of the various waveforms. The pero- Upper limbs
neal nerve SEP is very similar to the tibial nerve SEP Obligate Waveforms
except for two major differences. The PF waveform
When evaluating upper extremity SEP, the o bligate
cannot be recorded as the stimulation is performed
waveforms are EP, N13, P13/14, N18, and N20
at the PF and the latencies of the waveforms are
waveforms (2). As noted earlier, the absence of
about 10 ms shorter due to the more proximal site of
any of these obligate waveforms is considered
stimulation (Figure 4.9).
abnormal. However, if excessive artifact prevents
While the possible generators of various wave-
adequate averaging and resolution of the low ampli-
forms are listed, it should be noted that these
tude waveforms (i.e., subcortical potentials), their
are only approximations. Most waveforms have
absence may not necessarily imply an abnormality.
more than one generator, and most neural struc-
tures provide contribution to more than one SEP Absolute Latency
waveform.
The absolute latency of the EP waveform should be
The cortical waveforms of the upper and lower
noted. A prolonged latency may indicate a periph-
limb SEPs are of different polarities. This is because
eral nerve lesion. The latency of the EP waveform
the N20 waveform obtained after upper limb stim-
and the distance between the cathode and record-
ulation is recorded off the lateral cortical surface,
ing electrode can be used to calculate the peripheral
recording the negative end of the dipole. The P37
conduction velocity. A slowed conduction veloc-
waveform obtained after tibial nerve stimulation,
ity may help confirm a peripheral nerve lesion.
however, is recorded from the mesial surface and
However, a slowed conduction velocity should be
the positive end of its dipole projects to the scalp.
interpreted with caution. When obtaining an SEP,
threshold, not supramaximal, stimulation is used to
Parameters Measured activate the peripheral nerve fibers. As such, all or
The first step in evaluation of latencies and ampli- the fastest fibers may not have been activated, thus
tudes of various waveforms is to evaluate the reducing the conduction velocity.
waveforms and ensure that they are reliable, repro-
ducible, and correctly labeled. The presence of Interpeak Latencies
the various obligate waveforms (discussed in the The most common IPLs measured for upper limb
following) should be noted. The latencies of the var- SEP are the EP–N20, EP–P13/14, and P13/14–
ious waveforms are noted. The absolute latencies N20 IPLs. The EP–N20 IPL assesses the dorsal
column pathway between the brachial plexus and is not obtained in the author’s laboratory, but
the somatosensory cortex, while the EP–P13/14 normative data can be found elsewhere (4).
assesses conduction up to the nucleus cuneatus
ipsilateral to the side of stimulation. The P13/14– Lower limbs
N20 IPL evaluates the pathway from the ipsilateral Obligate Waveforms
nucleus cuneatus to the contralateral somato- The obligate waveforms for lower extremity SEP
sensory cortex. The normative data used in the are the LP, N34, and P37 waveforms (2). Absence
author’s laboratory are presented in Table 4.4. Also of any of these waveforms is considered an abnor-
included is normative data for a fast stimulation mality, with the same caveat as noted earlier;
rate (approximately 30/s). recordings must be reliable with little artifact
Some laboratories report the N13–N20 IPL. This obstructing the ability to record the waveforms.
is the conduction time from C5S to the contralateral This is particularly an issue with the LP waveform,
cortex. This central conduction time (CCT) time can as due to its low amplitude and excessive back-
be determined by dividing the N13–N20 IPL by the ground EMG activity, it can be a difficult response
distance from C5S to the contralateral scalp. It is to obtain.
argued that this measurement is a true measure of
central conduction and does not include conduction Absolute Latencies
through the peripheral nerves. This measurement The PF waveform is not considered obligate; how-
ever when present, its absolute latency and the
Table 4.4 Normative Data for Upper and Lower Limb SEPs peripheral nerve conduction velocity should be
from Author’s Laboratory evaluated as prolongation could suggest a periph-
Median/Ulnar eral nerve lesion. However, the PF waveform latency
Stimulation Rate ≈ 5/s prolongation and slowed conduction velocity have
Latency, IPL, or Absolute Max.
the same caveats as with EP waveform prolonga-
Wave Latency (ms)* Asymmetry tion. Threshold versus supramaximal stimulation
of the peripheral nerve may not result in activation
EP–P14 4.9 0.5
of all or the fastest fibers.
P14–N20 7.8 1.3
At times, the PF and LP waveforms cannot be
EP–N20 11.5 1.2
recorded, and in these cases, the absolute latency
EP 12.4 0.6
of the P37 waveform is evaluated. This latency is
N20 22.5 1.4
a measure of the sensory impulse conduction from
Stimulation Rate ≈ 31/s the site of stimulation to the contralateral somato-
Latency, IPL, or Absolute Max. sensory cortex. A large segment of the peripheral
Wave Latency (ms)* Asymmetry nerve is included in this latency assessment, and
EP–P14 5.2 1.4 abnormality should not necessarily be ascribed to
P14–N20 9.0 1.8 the central nervous system.
EP–N20 13.1 1.3
EP 12.9 0.6 Interpeak Latencies
N20 24.5 1.6 The IPLs commonly evaluated after lower limb
Tibial Stimulation Rate ≈ 5/s (tibial nerve) stimulation include the LP–P37 and
Latency, IPL, or Absolute Max.
PF–P37 IPLs. The LP–P37 IPL evaluates the dorsal
Wave Latency (ms)* Asymmetry column pathway from the lower spinal cord to the
contralateral somatosensory cortex. Since it does
LP–P37 20.2 3
not involve conduction through the peripheral
PF–P37 35.0** 4
nerve, it is a good measure of central conduction.
P37 47.0** 5
However, the LP waveform cannot be resolved in
* Latency values mean + 3SD. all patients, and in those situations, the PF–P37 IPL
** Values may need to be height adjusted. is assessed. The PF–P37 IPL measures the conduc-
For peroneal nerve stimulation, subtract 10 ms from all values. tion of sensory impulse from the ipsilateral PF to
EP, evoked potential; IPL, interpeak latency; LP, lumbar potential; the contralateral somatosensory cortex. This latency
PF, popliteal fossa; SEP, somatosensory evoked potential. includes a segment of the peripheral nerve, so it is
not as reliable as the LP–P37 IPL in assessment of Upper limbs

central conduction. The normative data used in the In the upper limb, median nerve SEP are performed
author’s laboratory are presented in Table 4.4. most often. However, ulnar nerve SEPs are inter-
Peroneal nerve SEP measurements are similar to preted in the same manner and the same wave-
the tibial nerve SEP described earlier. However, the forms are recorded as the site of stimulation for
PF response is not recorded as that is the site of stim- both nerves is a comparable distance away from the
ulation. The absolute latency of the P37 waveform recording electrodes.
(at times referred to as the P27) is 10 ms shorter.
However, the LP–P37 IPL is the same as that of tib- Loss of Obligate Waveforms
ial nerve SEP. The obligate waveforms after upper extremity
stimulation are the EP, N13, P13/14, N18, and N20
waveforms (2). The absence of an obligate wave-
Abnormalities form suggests a lesion of its neural generator or of
SEP abnormalities can be of three main types: the dorsal column pathway distal to it. However,
loss of obligate waveforms, prolongation of IPL, as noted earlier, technical issues regarding reliabil-
or prolongation of absolute latencies. Complete ity of the loss of waveform must be considered as
loss of an obligate waveform is the most signifi- well. The following can be used as an interpretation
cant abnormality and implies that there is a lesion guide to loss of obligate waveforms:
of the generator of that waveform or of the dorsal
column pathway more distal to it. Before absence • Loss of all obligate waveforms: Technical
of a waveform is considered abnormal, however, it issues must be ruled out; lesion in peripheral
must be confirmed that the loss of waveform is not nerve distal to brachial plexus.
due to technical reasons or excessive artifact mak- • Loss of N13, P13/14, N18, and N20 waveforms;
ing it difficult to visualize. IPL prolongations imply preserved EP waveform with normal latency:
a lesion in the dorsal column pathway between Lesion in the sensory/dorsal column pathway
the generators of the respective waveforms. These between the brachial plexus and cervical spi-
measures are more specific in identifying a lesion nal cord, ipsilateral to side of stimulation.
in the central nervous system than absolute latency • Loss of P13/14, N18, and N20 waveforms;
prolongations. The latter include peripheral nerve preserved EP and N13 waveforms with normal
conduction, and lesions of the peripheral nervous latencies: Lesion in the dorsal column pathways
system (such as a neuropathy) may make the abso- between the cervical spinal cord and lower
lute latencies prolonged. IPLs and absolute laten- medulla, ipsilateral to side of stimulation.
cies are considered abnormal if they exceed the • Loss of N18 and N20 waveforms; preserved
upper limit of the normative data or if they are EP, N13, and P13/14 waveforms with normal
asymmetrically prolonged compared to the oppo- latencies: Lesion in the dorsal column path-
site side. The degree of asymmetry is also specified ways between lower medulla and thalamus,
in the normative data (4). contralateral to side of stimulation.
Amplitude (other than complete loss of waveform) • Loss of N20 waveform; preserved EP, N13,
is seldom used as a criterion for abnormality, even P13/14, and N18 waveforms with normal
when there is an amplitude asymmetry between the latencies: Lesion in the dorsal column path-
two sides, which may be due to asymmetric intensity way between the thalamus and somatosensory
of stimulation, the proximity of the stimulating elec- cortex, contralateral to side of stimulation.
trodes to the nerve, degree of alertness of the patient, • Loss of N13, P13/14, or N18 waveform;
or some other nonpathological factor. Amplitude is preserved EP and N20 waveforms with nor-
usually measured from the peak of interest to the mal latencies: Uncertain clinical significance
following peak of opposite polarity, that is, for the and the absence of one or more obligate wave-
N20 waveform, it is from the peak of the N20 waveforms may be technical.
form to the trough of the following P22 waveform.
Similarly, for the P37 waveform, the amplitude is Prolongation of Interpeak Latencies
measured from the peak (trough) of the P37 wave- Several IPL are evaluated in upper extremity SEP.
form to the following peak of the N45 waveform. The most important ones include EP–P13/14,
P13/14–N20, and EP–N20 IPLs. The N13–N20 IPL should be made to equalize the amplitude of the
is sometimes also evaluated as a measure of CCT. EP waveforms between the two sides; this suggests
The following can be used as an interpretation that a comparable number of nerve fibers have been
guide to prolongation of IPLs: activated. This is more accurate than equalizing the
stimulus intensity between the two sides.
• Prolongation of the EP–P13/14 IPL: Lesion
between the brachial plexus and lower Lower limbs
medulla, ipsilateral to the side of stimulation. Tibial nerve SEP is the most common type of lower
• Prolongation of the P13/14–N20 IPL: Lesion limb SEP performed. Rarely peroneal nerve SEP will
between the lower medulla and the somato- be recorded. However, peroneal nerve stimulation
sensory cortex, contralateral to the side of is performed at the PF. This prevents the recording
stimulation. of the PF response, and all other waveforms have a
• Prolongation of the EP–N20 IPL: Lesion latency that is 10 ms shorter than tibial nerve SEPs.
between the lower medulla and the contralat-
Loss of Obligate Waveforms
eral somatosensory cortex. Prolongation of the
EP–P13/14 and P13/14–N20 IPLs can be used The obligate waveforms after lower extremity stimu-
to further localize the lesion. If neither of these lation are the LP, N34, and P37 waveforms (2). Loss of
two IPLs is prolonged, a prolonged EP–N20 any of these waveforms suggests a lesion of its neu-
IPL is still considered abnormal. ral generator or the dorsal column pathway distal to
• Prolongation of the N13–N20 IPL: Lesion it. However, technical adequacy of the study must
between the cervical spinal cord and contralat- be confirmed before declaring the absence of a wave-
eral somatosensory cortex. form pathological. The following can be used as an
interpretation guide to loss of obligate waveforms:
Prolongation of Absolute Latencies
• Loss of all obligate waveforms: Technical
With upper limb SEP, only the absolute latency of
issues must be ruled out; lesion in peripheral
the EP waveform is evaluated. If it is prolonged
nerve distal to PF.
and IPLs are normal, this may indicate a peripheral
• Loss of LP, N34, and P37 waveforms; pre-
nerve lesion. A slowed conduction velocity can help
served PF waveform with normal latency:
confirm this. However, as noted earlier, when evalu-
Lesion in the sensory/dorsal column pathway
ating conduction velocity, it is important to consider
between the PF and lower spinal cord, ipsilat-
that the fastest conduction velocity is likely not deter-
eral to side of stimulation.
mined as only threshold, not supramaximal, stimu-
• Loss of N34 and P37 waveforms; preserved PF
lation intensity is used for SEP. This may leave some
and LP waveforms with normal latencies: Lesion
of the fastest conducting fibers unstimulated and
in the dorsal column pathways between the
not contributing to the conduction velocity. Another
lower spinal cord and contralateral thalamus.
caveat to consider when evaluating the absolute
• Loss of P37 waveforms; preserved PF, LP, and
latency of the EP waveform (and the conduction
N34 waveforms with normal latencies: Lesion
velocity) is limb temperature. Cold slows conduc-
in the dorsal column pathways between the
tion velocity. When the ambient temperature is low
thalamus and somatosensory cortex, contra-
and the limb is uncovered, the peripheral nerves
lateral to side of stimulation.
may become cold and have slowed conduction.
• Loss of PF, LP, or N34 waveforms; preserved
Reduced Amplitude P37 waveform with normal latency: Usually
A reduced amplitude of an obligate waveform is of interpreted as normal.
uncertain clinical significance and usually not con-
sidered abnormal. Amplitudes are not compared to Prolongation of Interpeak Latencies
normative data; rather, they may be compared to With lower limb SEP, the most important IPL to
the opposite side. At times the EP and N20 wave- consider is the LP–P37 IPL. When the LP waveform
forms may have asymmetric amplitudes. The pos- cannot be identified, the PF–P37 IPL is evaluated.
sible nonpathological reasons for this have been Since the PF–P37 IPL includes a segment of periph-
discussed earlier. When acquiring SEPs, an attempt eral nerve conduction, analysis of the LP–P37 IPL is
preferred. The following can be used as an interpre- As an example, if after tibial nerve stimulation,
tation guide to prolongation of IPLs: the LP–P37 IPL is prolonged, but after median
nerve stimulation all IPLs are normal, the lesion
• Prolongation of the LP–P37 IPL: Lesion is most likely between the lower spinal cord and
between the lower spinal cord and the contra- the cervical spinal cord, ipsilateral to the side of
lateral somatosensory cortex. stimulation. On the other hand, if the LP–P37 IPL
• Prolongation of the PF–P37 IPL: Lesion between after tibial nerve stimulation is prolonged and the
the PF and the contralateral somatosensory P13/14–N20 IPL after median nerve stimulation is
cortex. A lesion of the ipsilateral lumbosacral also prolonged, the site of the lesion is most likely
plexus could cause this type of abnormality. between the lower medulla and the somatosensory
cortex, contralateral to the side of stimulation. In
Prolongation of Absolute Latencies the latter scenario, an additional lesion in the spinal
Two absolute latencies are important to evaluate with cord below the cervical level cannot be excluded.
lower limb SEP. If the latency of the PF waveform Many permutations of how both upper and lower
is prolonged and the PF–P37 and LP–P37 IPLs are extremities SEPs can be used to localize a lesion are
normal, this may indicate a peripheral nerve lesion. demonstrated in the examples later in this chapter.
A slowed conduction velocity can help confirm this.
It should, however, be remembered that slowed con-
duction velocity may be due to nonsupramaximal
Pediatric Considerations
stimulation. Cold temperature of the lower limb can Sensory function, particularly vibration and posi-
also affect the absolute latency of the PF waveform. tion sense, is difficult to assess in young children.
The absolute latency of the P37 waveform should SEP can be a very useful test in this age group as
be evaluated if the LP and PF waveforms are not it assesses the dorsal column pathways. However,
seen (and consequently the LP–P37 and PF–P37 SEPs are performed sparingly in this age group as
IPLs cannot be evaluated). In this situation, a pro- they are more uncomfortable than other types of
longed absolute latency of P37 is indicative of evoked potentials.
a lesion between the site of stimulation and the
contralateral somatosensory cortex. Maturational Issues
Three types of maturational changes affect SEP:
Reduced Amplitude increasing myelination of nerves and fiber tracts,
As with upper extremity SEP, a reduced amplitude maturation of synapses, and increase in limb length.
of an obligate waveform is of uncertain clinical sig Peripheral nervous system myelination reaches adult
nificance and usually not considered abnormal. levels by age 3 to 4 years, while central nervous sys-
Amplitudes may be compared to the opposite side. tem myelination and maturation of synapses are not
At times the P37 waveform amplitude is asymmet- complete until ages 5 to 7 years (5). Thus, peripheral
ric between the two sides. Many of the possible non- conduction velocity reaches adult values faster than
pathological reasons for this have been discussed central conduction. Limb length increases until age
earlier, but change in attention (such as drowsiness 14 to 18 years, and that can result in a progressive
or sleep) can also affect the P37 waveform ampli- increase in latency of the waveforms.
tude. If a patient is awake during stimulation of one With upper limb SEP, the EP waveform latency
side but falls asleep during the acquisition of data stabilizes by about age 2 to 3 years, as peripheral
on the opposite side, there may be a remarkable myelination is almost complete. Thereafter, all the
difference in amplitude. When acquiring SEPs, an waveforms increase in latency as the limb length
attempt should be made to equalize the amplitude increases. The P13/14–N20 and N13–N20 IPLs are
of the PF waveforms between the two sides and the not affected by limb length, and decrease in latency
patient’s level of alertness should be kept consistent. initially due to central myelination, but reach adult
values by age 10 years (6). The EP and N13 wave-
Combining upper and lower limbs forms reach adult values by about age 14 to 18 years.
Upper and lower limb SEPs are commonly per- The lower limb SEP maturation is similar to that
formed together. The combined SEPs can be used of the upper limb. The latency of the LP waveform is
to further localize a lesion in the nervous system. about 16 ms preterm and decreases to 8 ms at birth
with increasing myelination. With limb growth, the

latency increases again to about 14 ms by age 8 years
Clinical Correlates
and continues to increases slightly until age 14 to 18 SEP provides a valuable adjunct to the neurological
years. The P37 waveform follows a similar course, examination. Sensory function is difficult to assess
decreasing in latency with improved peripheral and is very subjective; SEP provides an objective
and central myelination until age 8 years, and then measure to an otherwise very subjective sensory
increasing in latency until adult height is achieved. assessment. Median nerve SEP can be helpful in
Central conduction, as measured by the LP–P37 predicting outcome of anoxic coma. SEP of all types
IPL, improves (decreasing LP–P37 IPL) until age can be used during NIOM as well.
8 years when peripheral and central myelination is
complete. Increase in height does not affect the LP– Multiple Sclerosis
P37 IPL very much as the growth is predominantly
in the limbs and less so in the trunk (7). Evoked potentials have been used for many years in
There is a morphological evolution in the cortical the diagnosis and assessment of multiple sclerosis
SEP waveform shape. In young children, the N20 (MS). With sophisticated MRI techniques, evoked
and P37 waveforms after upper and lower limb potentials are used less often. However, evoked
stimulation have a broad shape. By age 5 to 8 years, potentials evaluate function, while MRI evaluates
the waveforms become more peaked and better anatomy, and they are often complementary. SEPs
defined. These cortical waveforms are present in are used to identify objective findings for subjec-
85% and 50% of term infants after upper and lower tive sensory complaints, identify silent lesions, and
limb stimulation, respectively. By age 3 years, most follow progression of the disease. They may detect
children have cortical waveforms. lesions that are not seen with MRI and vice versa.
SEPs have found to be abnormal in 60% of patients
with MS, more often than other types of evoked
Methodology potentials (8). The most common abnormality in
The basic methodology of obtaining upper and SEP noted in patients with MS is loss of an obligate
lower extremity SEP is the same in children as in waveform. Various IPLs may also be prolonged,
adults, with minor differences. Often the space depending on the site of the lesion.
along the wrist is limited, so only the cathode is
placed on the volar surface of the wrist, and the Transverse Myelitis
anode is placed on the dorsum of hand. The stimu- Transverse myelitis (TM) results in a myelopathy
lation rate may need to be slowed to approximately that is more severe than in MS. Often there are sen-
1 Hz, particularly for lower extremity stimulation. sory and motor symptoms and signs. Depending on
In very young patients whose waveforms may be the site of the lesion, upper and lower limb SEPs may
delayed, the time window of acquisition may need be abnormal. Often the abnormality is prolongation
to be increased. The stimulation intensity needed to of IPLs that encompass the site of injury. In more
activate the peripheral nerves is usually less than severe cases, there may be loss of obligate wave-
that needed for adults. In young children, the laten- forms above the site of the lesion. As an example, if
cies of the various waveforms are shorter, but their the TM occurs at a level of thoracic 8 (T8), median
names do not change. For example, the N20 wave- nerve SEP will be normal, but the tibial nerve SEP
form may occur at 15 ms, but it is still called the N20 LP–P37 IPL may be prolonged. In more severe cases,
waveform. there may be loss of the P37 waveform, while the PF,
LP, and upper extremity waveforms are present.
Interpretation
The principles of interpretation of pediatric stud-
Brainstem Lesions
ies are the same as adult studies. However, age- Strokes and tumors in the brainstem that affect the
adjusted normative data must be used to interpret dorsal column pathways can cause SEP abnormali-
these studies. Pediatric SEP normative data are ties. However, if these fiber tracts are not involved,
available elsewhere (5–7). Analysis of obligate the SEP will be normal. The abnormality seen most
waveforms, IPLs, and absolute latencies is done in often will be a prolongation of the P13/14–N20 and
a manner similar to that used for adults. EP–N20 IPLs after median nerve stimulation and
LP–P37 IPL after tibial nerve stimulation. In more of N20 waveforms bilaterally after median nerve
severe cases, there may be loss of one of these obli- stimulation is highly associated with a poor o utcome
gate waveforms. in patients in anoxic encephalopathy. In one study,
all patients with bilaterally absent N20 waveforms
Neurodegenerative Diseases after median nerve stimulation had a poor out-
come (17). Some patients may have N20 wave-
Neurodegenerative diseases that affect the dor-
forms present soon after cardiac arrest, but these
sal column pathways will result in abnormal SEP.
waveforms are not recordable after 72 hours. The
Friedreich’s ataxia is one such condition in which
American Academy of Neurology (AAN) recom-
there is selective loss of vibration and position sense.
mends use of median nerve SEP for aid in progno-
Median nerve SEPs are affected early in this disease
sis of anoxic encephalopathy after 24 to 72 hours of
(9). Abnormal SEPs have been reported in amy-
cardiac arrest (18). While the absence of N20 wave-
otrophic lateral sclerosis (ALS), a predominantly
forms bilaterally is helpful in predicting a poor out-
motor system disease. This has been taken as evi-
come, when these waveforms are present, they are
dence that there is sensory involvement as well in
not helpful prognostically. Thus, in a patient with
ALS, while others have argued that the SEP changes
anoxic encephalopathy, if median nerve SEP shows
are due to concomitant cervical spondylosis (10,11).
retained N20 waveforms unilaterally or bilaterally,
SEP may be abnormal in late stages of many other
the study should not be interpreted as suggesting
neurodegenerative diseases, but they are rarely used
good prognosis; rather such a study should be con-
in diagnosis or management of these conditions.
sidered not helpful in prognosis.
Peripheral Neuropathies In children suffering from anoxic coma, median
nerve SEPs have also been found to be useful.
Hereditary demyelinating peripheral neuropathies,
Bilaterally absent N20 waveforms have a 100%
like Charcot–Marie–Tooth disease type I, cause pro-
positive predictive value for poor outcome (19).
longation of SEP absolute latencies and slowed con-
However, if the median nerve SEP is performed
duction velocities of peripheral nerves. However,
within 24 hours of onset of anoxic injury in the very
the IPLs are usually normal. These findings con-
young patient, the N20 waveforms may initially be
firm the predominantly peripheral demyelinating
present, but disappear within 72 hours. In pediatric
nature of this disease. Mild SEP abnormalities may
patients, repeating median nerve SEP after a few days
be seen in other chronic neuropathies, such as those
is recommended before declaring poor prognosis.
seen with diabetes mellitus, chronic renal failure,
Coma from traumatic and other nonanoxic,
and Guillain–Barre syndrome (12). The abnormal-
nontraumatic causes has also been evaluated with
ities seen in these conditions are prolongation of
median nerve SEP. For most cases of coma, regard-
absolute latencies with normal IPLs, indicating pre-
less of etiology, the bilateral absence of N20 wave-
dominantly peripheral nerve involvement.
forms is suggestive of poor prognosis (20). The data
Leukodystrophies for this is not as robust as that for anoxic enceph
alopathy. However, in some unusual etiologies,
Leukodystrophies involve dysmyelination of the cen-
such as being struck by lightning, the N20 wave-
tral nervous system and the dorsal column pathways
forms may be absent initially, but reappear in a few
may be affected. In adrenoleukodystrophy, meta-
days (21). Repeating the median nerve SEP study in
chromatic leukodystrophy, Krabbe’s disease, and
these situations helps clarify the prognosis.
Pelizaeus–Merzbacher disease, abnormal SEP with
prolonged IPLs have been reported (13,14). In some of
these conditions, such as Krabbe’s disease, the periph-
eral nervous system may be involved as well, making
Examples
absolute latencies and conduction velocities abnor- In this section, examples of upper and lower limb SEPs
mal in addition to prolonged IPLs (15,16). As these are presented (Figures 4.11–4.26, Examples 1–16.)
diseases progress, the SEP abnormalities worsen. Technical issues and various types of pathologies
are highlighted. SEP abnormalities are nonspecific,
Coma so the abnormality and how the study should be
Median nerve SEPs have been used in the prog- interpreted are discussed. Whenever possible, the
nostic evaluation of patients in coma. Absence abnormalities are correlated with MRI findings.
Age: 47 years
Sex: Female
Median Nerve SEPs
(A) (B)
Stimulation intensity: 17.3 mA Stimulation intensity: 21.2 mA
Median nerve conduction velocity (median nerve to Median nerve conduction velocity (median nerve to
EP): 51 m/s EP): 49 m/s
EP 9.1 EP 9.2
N13 12.3 N13 12.7
P13 13.4 P13 13.8
N18 16.0 2.1 (P13–N18) N18 16.7 2.4 (P13–N18)
N20 19.3 0.8 (N20–P22) N20 20.0 0.7 (N20–P22)

EP–N20 10.1 EP–N20 10.9
EP–P13 4.3 EP–P13 4.6
P13–N20 5.9 P13–N20 6.2

Tibial Nerve SEPs

(C) (D)
Tibial nerve conduction velocity (tibial nerve to T12S): Tibial nerve conduction velocity (tibial nerve to T12S):
42 m/s 37 m/s
Absolute Latencies
Waveform Latency (ms) Amplitude (μV) Absolute Latencies
PF 8.5
PF 8.5
LP 21.2
LP 21.6
P31 29.2 P31 29.9
N34 33.1 N34 33.4
P37 38.5 1.19 (P37–N45) P37 37.6 1.16 (P37–N45)

PF–P37 30.0 PF–P37 29.1
LP–P37 17.3 LP–P37 16.0

History: This is a 47-year-old female with complaints of waxing and waning numbness and prickly sensations on the left side of her
body for the last 2 years. An EMG and abdominal ultrasound were normal.
Discussion: This is a normal median and tibial nerve SEP study. When evaluating the left (A) and right (B) median nerve SEP, it is
important to first look at all the channels and confirm that the appropriate waveforms are present. In the first channel from the
bottom, the EP waveform is seen. The second channel shows a reproducible N13 waveform. The subcortical waveforms are in the
third channel, labeled P14 and N18. As explained in the text, at times the P13 waveform is tagged instead of the P14 waveform.
In the fourth channel is the N20 waveform. In the fifth and sixth channels, the N13 waveform is seen again in various derivations.
After confirming the presence of all the waveforms, the absolute latencies are evaluated; however, in the author’s laboratory, these
are not compared to normative data. Thereafter the IPLs are evaluated: EP–N20, EP–P13/14, and P13/14–N20. In this example, IPLs
after stimulation of the left and right median nerves are normal.
The left (C) and right (D) tibial nerve SEPs are evaluated in a manner similar to the median nerve SEP. First, the channels should
be reviewed to confirm that all waveforms are present. Starting from the bottom, the first channel has the PF waveform and the
second channel has the LP waveform. The third channel shows the subcortical (P31, N34) and cortical (P37) waveforms. The forth
channel has only the P37 waveform and the subcortical waveforms cancel (see text for details). The fifth, sixth, and seventh channels
all show the P31 and N34 waveforms. Additionally, the CP ipsilateral–C5S (CP3–C5S after left-sided stimulation, CP4–C5S after
right-sided stimulation) also show the P37 waveform. Once the waveforms have been confirmed, the absolute latency of the P37
waveform is evaluated. Thereafter the IPLs are evaluated. If the LP waveform is present, as in this case, the LP–P37 IPL is relied upon
as it is a better measure of central conduction than the PF–P37 IPL. If the LP waveform is absent, the PF–P37 IPL is evaluated. In this
example, the absolute latencies and the IPLs after stimulation of the left and right tibial nerves are normal.
This patient’s brain and spinal cord MRIs were normal.
As the absolute and IPLs of the SEPs obtained after left and right median and tibial nerve stimulation are within normal limits, this
is a normal study. It does not help explain the etiology of her symptoms.
EMG, electromyographic; EP, evoked potential; IPL, interpeak latency; LP, lumbar potential; PF, popliteal fossa; SEP, somatosensory evoked
potential.

Age: 31 years
Sex: Female
Median Nerve SEPs
(−↑)
(−↑)
(A) (B)
Median nerve conduction velocity (median nerve to EP): Median nerve conduction velocity (median nerve to EP):
63 m/s 63 m/s

EP 10.6 EP 10.4
N13 13.3 N13 13.8
P13 15.1 P13 14.2
N18 16.8 1.38 (P13–N18) N18 17.3 1.47 (P13–N18)
N20 20.0 0.88 (N20–P22) N20 19.9 1.84 (N20–P22)

EP–N20 9.4 EP–N20 9.4
EP–P13 4.4 EP–P13 3.8
P13–N20 5.0 P13–N20 5.6

Tibial Nerve SEPs

(−↑)
(−↑)
(C) (D)
43 m/s 41 m/s

PF 10.2 PF 11.1
LP 26.6
LP 24.9
P31 34.8
P31 34.3
N34 38.5
N34 38.5 P37 43.0 1.82 (P37–N45)
P37 43.4 0.29 (P37–N45)

PF–P37 33.2 PF–P37 31.9
LP–P37 18.5 LP–P37 16.4

History: This is a 31-year-old female with a history of fibromyalgia complaining of very uncomfortable sensation on the bottoms of
her feet. This is persistent and she is concerned about having MS.
Discussion: This is a normal left (A) and right (B) median and tibial nerve SEP study. Evaluation of the median nerve SEP demon-
strates presence of all obligate waveforms, and the IPLs are within normal limits. The nerve conduction velocities also are normal.
The tibial left (C) and right (D) nerve SEPs show higher amplitude waveforms after right-sided stimulation. Important to notice is
that the amplitude of the left tibial PF waveform is much smaller than the right tibial PF waveform. Consequently, the P37 waveform
after left tibial nerve stimulation is also of lower amplitude than its counterpart. Since the stimulation intensities were comparable
(28.6 mA and 29.0 mA, after left and right stimulation), it is likely that the right tibial nerve stimulating electrodes were closer
to the nerve than the left-sided ones. The absolute latencies of the P37 waveforms and the LP–P37 IPL after left- and right-sided
stimulation were normal.
This patient’s brain MRI scan is normal.
The tibial nerve SEP amplitude asymmetry is not an abnormality in this case. The low amplitude left PF waveform suggests inade-
quate stimulation on that side. Because of the normal absolute latencies and IPLs after median and tibial nerve stimulation, this is a
normal study. This study does not reveal any findings that might be consistent with MS.
EP, evoked potential; IPL, interpeak latency; LP, lumbar potential; MS, multiple sclerosis; PF, popliteal fossa; SEP, somatosensory evoked
potential.

Age: 49 years
Sex: Female
Median Nerve SEPs
(−↑) (−↑)
(A) (B)
59 m/s 60 m/s
EP 10.0 EP 10.0
N13 13.1 N13 13.3
P13 13.4 P13 13.6
N18 17.7 1.27 (P13–N18) N18 18.3 1.43 (P13–N18)
N20 20.1 0.55 (N20–P22) N20 20.4 0.50 (N20–P22)

EP–N20 10.1 EP–N20 10.4
EP–P13 3.4 EP–P13 3.6
P13–N20 6.7 P13–N20 6.8

Tibial Nerve SEPs

(−↑) (−↑)
(−↑)
(C) (D)
Tibial nerve conduction velocity (tibial nerve to T12S): 39 Tibial nerve conduction velocity (tibial nerve to T12S):
m/s 41 m/s

PF 9.4 PF 9.6
LP 23.7
LP 24.5
P31 N/A
P31
N34 N/A
N34 P37 39.8 2.27 (P37–N45)
P37 40.3 1.47 (P37–N45)

PF–P37 30.9 PF–P37 30.2
LP–P37 15.8 LP–P37 16.1

History: This is a 49-year-old female with a history of idiopathic scoliosis who presents with neuropathic pain in her rib cage, neck,
and both arms. She also complains about immobility of both ankles. This study is being done to evaluate her pain complaints.
Discussion: This is a normal median and tibial nerve SEP study. After left (A) and right (B) median nerve stimulation, reliable and
reproducible SEPs are seen. All waveforms are appropriately tagged. The subcortical waveforms (P13/14 and N18) after right median
nerve stimulation are noisy, but still reliable waveforms are noted. All IPLs are within normal limits.
The left (C) and right (D) tibial nerve SEPs show reliable PF, LP, and P37 waveforms, but subcortical (p31 and N34) waveforms are
not seen. Though the LP, N34, and P37 waveforms are considered obligate, at times the LP and N34 may be absent as they are low
amplitude waveforms and there may be too much noise to clearly average them. This is the case after left-sided (C) stimulation
especially, and the channels that would show the subcortical waveforms are noisy. If the other absolute latencies and IPLs are within
normal limits, the absence of LP or N34 waveform should be interpreted with caution. In this patient, the absolute latency of the
P37 waveform and the LP–P37 and PF–P37 IPLs are normal after left- and right-sided stimulation.
The brain and cervicothoracic MRI scans of this patient were normal.
The absent P31 and N34 waveforms of the left and right tibial nerve SEP study are not an abnormality as the absolute latency of
the P37 waveforms and the LP–P37 and PF–P37 IPLs are normal. The left and right median nerve SEP is also normal. This study does
not help explain the patient’s symptoms.
EP, evoked potential; IPL, interpeak latency; LP, lumbar potential; PF, popliteal fossa; SEP, somatosensory evoked potential.

Age: 41 years
Sex: Male
Median Nerve SEPs
Scale: Amplitude = 0.05–0.1 μV/div; Latency = 4 ms/div
↑(−) ↑(−)
(A) (B)
55 m/s 56 m/s

EP 10.1 EP 9.8
N13 13.2 N13 13.0
P13 13.1 P13 12.7
N18 15.7 0.06 (P13–N18) N18 16.5 0.25 (P13–N18)
N20 N/A N/A (N20–P22) N20 N/A (N20–P22)

EP–N20 N/A EP–N20 N/A
EP–P13 3.0 EP–P13 3.0
P13–N20 N/A P13–N20 N/A
Figure 4.14 (A–G) Example 4. (continued )

Tibial Nerve SEPs

↑(−) (−↑)
(C) (D)
42 m/s 40 m/s

PF 9.6 PF 9.9
LP 22.9 LP 23.5
P31 28.8 P31 28.8
N34 N/A N34 N/A
P37 39.1 0.23 (P37–N45) P37 38.8 0.35 (P37–N45)

PF–P37 29.5 PF–P37 28.9
LP–P37 16.2 LP–P37 15.3
Figure 4.14 (A–G) (continued )

(E) (F)
(G)
History: This is a 41-year-old male who fell 20 feet off a ladder while painting a house. He was unconscious at the scene, and
upon arrival to the hospital was intubated. He is now sedated and this study is being done to evaluate for spinal cord and brain
dysfunction.
Discussion: This is an abnormal median nerve SEP and normal tibial nerve SEP study. After left (A) and right (B) median nerve stim-
ulation, the N20 waveform is absent. The EP, N13, P13/14, and N18 waveforms are present. The EP–P13/14 IPLs are normal bilat-
erally, and the other IPLs cannot be reported as the N20 waveform is absent. The median nerve conduction velocity is also normal
bilaterally. Note that the AA electrode was not placed, so the top two derivations are not available.
The left (C) and right (D) tibial nerve SEP shows reliable PF, LP, P31, and P37 waveforms. After right tibial nerve stimulation, there
may be a N34 in the CPz–C5S channel; however, it has not been tagged. The absolute latencies of both P37 waveforms and the
LP–P37 and PF–P37 IPLs are within normal limits.
The brain CT scans shows mild diffuse cerebral edema (E). CT of the cervical spine shows bilateral C2 laminar fractures (F). MRI of
the cervical spine shows some increased signal on a T2-weighted sagittal image, but there is not clear spinal cord compression (G).
The absent N20 waveforms with preserved N13, P13/14, and N18 waveforms on the median nerve SEPs is suggestive of an injury
to the dorsal column pathways above the VPL nucleus of the thalamus bilaterally. Since the tibial nerve SEP (in particular the P37
waveforms) is normal, the lesion is likely to be closer to the cortex than the thalamus, bilaterally. This is because there is selective
injury to the median nerve but not tibial nerve fibers, and they separate when they are closer to the cortex. Such a finding may be
seen with cerebral contusions or edema. There does not appear to be an injury to the dorsal column pathways in the spinal cord due
to the normal absolute latencies of the P37 waveforms and the normal LP–P37 and PF–P37 IPLs bilaterally. The C2 laminar injury is
not affecting the spinal cord, and the increased signal seen on MRI likely is not severe enough to affect the dorsal column pathways.
AA, Adam’s apple; EP, evoked potential; IPL, interpeak latency; LP, lumbar potential; PF, popliteal fossa; SEP, somatosensory evoked potential,
VPL, ventral posterior lateral.
Figure 4.14 (A–G) (continued )

Age: 31 years
Sex: Female
Median Nerve SEPs
↑(−)
↑(−)
(A) (B)
Median nerve conduction velocity (median nerve Median nerve conduction velocity (median nerve
to EP): 64 m/s to EP): 58 m/s

EP 9.8 EP 10.6
N13 13.8 N13 14.5
P13 15.6 P13 15.4
N18 19.2 1.30 (P13–N18) N18 19.0 1.59 (P13–N18)
N20 N/A N/A (N20–P22) N20 19.7 1.60 (N20–P22)

EP–N20 N/A EP–N20 9.0
EP–P13 5.9 EP–P13 4.8
P13–N20 N/A P13–N20 4.2

(C) (D)
History: This is a 31-year-old female who started complaining of left arm and leg weakness in the third trimester of her pregnancy.
Examination showed almost complete left-sided hemiparesis and sensory loss. This SEP was obtained as preoperative workup.
Discussion: This is abnormal median nerve SEP due to absence of the N18 and N20 waveforms after left-sided stimulation (A). The
EP and N13 waveforms are clearly present. The subcortical channel is noisy; however, the P13/14 waveform is reliable. The N18 and
N20 waveforms, though tagged, are not reproducible and reliable. The EP–P13/14 IPL is prolonged as well, and the EP–N20 and
P13/14–N20 IPLs cannot be measured due to lack of the N20 waveform.
After right median nerve stimulation (B), the SEP is normal. All waveforms are present, and all IPLs are within normal limits. Tibial
nerve SEPs were not performed.
Brain MRI showed a lesion in the right pons and midbrain, as shown in the axial image (C). This lesion was contrast enhancing, as
shown in the coronal image (D).
In the left median nerve SEP, the N18 and N20 waveforms were tagged. During interpretation, it is important to verify if the wave-
forms are reliable and correctly tagged. In this case, they were not. The dorsal column pathways arising from the left median nerve
decussate in the lower brainstem and ascend on the right side in the pons and midbrain. As would be expected, the N18 (thalamic)
and N20 (cortical) waveforms were absent due to this lesion. The EP–P13/14 IPL was also prolonged, and this suggests that the
lesion was also having an effect lower in the brainstem as well. A normal right median nerve SEP suggests that the lesion was not
affecting the dorsal column pathways of that side.
EP, evoked potential; IPL, interpeak latency; SEP, somatosensory evoked potential.

Age: 25 years
Sex: Female
Median Nerve SEPs
(A) (B)
63 m/s 60 m/s

EP 9.4 EP 9.6
N13 12.6 N13 12.8
P13 13.6 P13 13.9
N18 16.1 1.83 (P13–N18) N18 16.1 1.92 (P13–N18)
N20 17.9 4.58 (N20–P22) N20 18.2 4.36 (N20–P22)

EP–N20 8.6 EP–N20 8.6
EP–P13 4.3 EP–P13 4.3
P13–N20 4.3 P13–N20 4.3

Tibial Nerve SEPs

(C) (D)
42 m/s 44 m/s

PF 7.6 PF 7.4
LP 18.4 LP 20.1
P31 N/A P31 N/A
N34 N/A N34 N/A
P37 N/A N/A (P37–N45) P37 N/A N/A (P37–N45)

PF–P37 N/A PF–P37 N/A
LP–P37 N/A LP–P37 N/A

(E) (F)
History: This is a 25-year-old female with a history lipomeningocele repair in childhood and a Chiari I malformation, presenting with
ongoing lower extremity weakness and numbness below the knees bilaterally. On examination, she has 4/5 power in both lower
extremities and sensory loss to all modalities up to the mid-shins. This study is being done to evaluate for spinal cord dysfunction.
Discussion: In this study, the median nerve SEP is normal, but the tibial nerve SEP is abnormal bilaterally. The left (A) and right (B)
median nerve SEP demonstrates all waveforms with good reproducibility. The latencies of the waveforms and the various IPLs are
within normal limits bilaterally.
After left (C) and right (D) tibial nerve stimulation, however, there is no recordable response above the LP waveform bilaterally. The
PF and LP waveforms are present, indicating that the peripheral nerves were adequately stimulated and that the stimulus traveled
at least up to the level of the lower spinal cord (T12). Thus, the lesion can be localized to between the lower spinal cord and the
somatosensory cortex bilaterally as all responses above the LP waveforms are absent.
The spinal MRI scan of this patient showed an ongoing tethered spinal cord, as is evident on the sagittal T2-weighted image (E).
A sagittal image of the cervical and thoracic MRI shows a small syrinx as well (F).
Though the lesion localization of the abnormal tibial nerve SEP is anywhere above the lower spinal cord, when the median nerve SEP
is interpreted in conjunction with the tibial nerve SEP study, further localization is possible. Since all IPLs of the bilateral median nerve
SEP are normal, the lesion is likely below the level at which the median nerve fibers enter the cervical spinal cord (lower cervical spi-
nal cord). Thus, when interpreting the two studies together, the lesion is most likely between the lower spinal cord and the cervical
spinal cord. This is consistent with the patient’s imaging findings showing persisting tethered spinal cord. Moreover, the syrinx in the
cervicothoracic spinal cord can also affect the tibial nerve SEP and cause loss of waveforms above the LP waveform.

Age: 57 years
Sex: Male
Median Nerve SEPs
3ms 4ms
(A) (B)
63 m/s 61 m/s

EP 9.8 EP 10.2
N13 13.3 N13 13.4
P13 14.2 P13 14.9
N18 16.2 1.29 (P13–N18) N18 17.9 1.19 (P13–N18)
N20 20.2 0.87 (N20–P22) N20 N/A N/A (N20–P22)

EP–N20 10.4 EP–N20 N/A
EP–P13 4.4 EP–P13 4.7
P13–N20 6.0 P13–N20 N/A

Tibial Nerve SEPs

8ms 8ms
(C) (D)
Tibial nerve conduction velocity Tibial nerve conduction velocity (tibial nerve to T12S):
(tibial nerve to T12S): 44 m/s 43 m/s

PF 8.8 PF 8.5
LP 21.9 LP 22.3
P31 32.5 P31 32.9
N34 35.2 N34 36.7
P37 42.6 1.00 (P37–N45) P37 N/A N/A (P37–N45)

PF–P37 33.8 PF–P37 N/A
LP–P37 20.0 LP–P37 N/A

(E)
History: This is a 57-year-old male who presents with a history of right face and arm numbness that started about 1 year ago and
lasted for 1 month. He has had no visual complaints. His examination showed a sluggishly reactive left pupil, and his motor, sensory,
and reflex examinations were normal. He is referred for evaluation of MS.
Discussion: The right median and tibial nerve SEPs are abnormal, while the left median and tibial nerve SEPs are normal. The median
nerve SEPs show reliable EP, N13, P13/14, and N18 waveforms after (A) left- and (B) right-sided stimulation. However, the N20
waveform is absent with right-sided stimulation and present with left-sided stimulation. All left-sided IPLs and the EP–P13/14 IPL
are normal.
Similarly, the tibial nerve SEPs show reliable PF, LP, P31, and N34 waveforms after stimulation of the left (C) and right (D) sides.
However, after right-sided stimulation, the P37 waveform is not seen, while it is present after left-sided stimulation. The LP–P37 and
PF–P37 IPLs are normal after left tibial stimulation, and the contralateral IPLs cannot be determined due to absent P37 waveform.
The brain MRI scan of this patient showed multiple, bilateral lesions in the white matter on the T2-weighted scan (E). On the left side
of the brain, periventricular and parietal lobe lesions were noted. A spinal MRI did not show any spinal cord lesions.
Corresponding to the patient’s complaint of right-sided numbness the patient had several months ago, the right median and tibial
nerve SEP showed absent cortical waveforms. Additionally, both had normal subcortical waveforms (N18, N34), suggesting that
the signal is ascending normally to the level of the thalamus. This would make the lesion most likely between the right thalamus
and somatosensory cortex. This corresponds to at least some of the lesions seen on the patient’s MRI scan. The MRI scan, however,
had bilateral lesions, while the patient’s SEPs from the left side were normal. This is likely because the right-sided lesions have not
affected the dorsal column pathways whereas the left-sided ones have. This SEP finding supports the possible diagnosis of MS.

Age: 33 years
Sex: Female
Median Nerve SEPs
3ms 3ms
(−↑)
(−↑)
NP 12.5
(A) (B)
59 m/s 56 m/s
EP 9.4 EP 9.4
N13 12.6 N13 N/A
P13 12.7 P13 12.5
N18 16.1 2.91 (P13–N18) N18 14.9 0.77 (P13–N18)
N20 17.4 2.43 (N20–P22) N20 23.8 0.79 (N20–P22)

EP–N20 8.0 EP–N20 14.3
EP–P13 3.3 EP–P13 3.1
P13–N20 4.7 P13–N20 11.2

Tibial Nerve SEPs

8ms 8ms
(−↑) (−↑)
(C) (D)
41 m/s 41 m/s
PF 8.7 PF 8.8
LP 22.0 LP 22.0
P31 27.3 P31 27.9
N34 32.2 N34 32.3
P37 35.2 0.75 (P37–N45) P37 36.7 0.45 (P37–N45)

PF–P37 26.5 PF–P37 27.9
LP–P37 13.2 LP–P37 14.7

(E) (F)
History: This is a 33-year-old female who presents with a 3-month history of right arm and leg numbness that started after she
was lifting firewood. There was some right leg weakness as well. The right leg symptoms resolved after a few weeks, but the right
arm numbness has persisted. Her examination shows right upper extremity loss of pinprick and vibration sense but is otherwise
normal. An initial MRI showed a possible corpus callosum lesion, but no other abnormalities. She is being referred for evaluation of
MS versus cervical spondylosis.
Discussion: The right median nerve SEP is abnormal, and the left median and bilateral tibial nerve SEPs are normal. The left median
nerve SEP (A) has all waveforms present and IPLs within normal limits. The right median nerve SEP (B) shows the presence of all
waveforms, but the N20 waveform is prolonged. Consequently, the EP–N20 and the P13/14–N20 IPLs are also prolonged. The
EP–P13/14 IPL is normal. This suggests that the lesion is between the lower medulla and the left somatosensory cortex. Moreover,
even though normative data are not available for this in the author’s laboratory, the N18–N20 IPL looks remarkably prolonged, more
so than the P13/14–N18 IPL. This helps further localize the lesion to between the left thalamus and somatosensory cortex. The N13
waveform is not clearly identified, though it has been hand tagged. This finding is of unclear significance as the channel is noisy and
there is a suggestion that the N13 waveform may have been present with more repetitions in the average.
The left (C) and right (D) tibial nerve SEPs are normal. Though the subcortical and cortical channels are noisy, the waveforms are
clearly reproducible and reliable. All waveforms are present and absolute latencies and IPLs are normal.
The brain MRI scan of this patient showed multiple, bilateral lesions in the white matter. In the FLAIR MRI axial image shown (E),
there is a high T2 signal in the white matter on the left side (arrow). The T2-weighted image (F) shows a lesion in the right pons
(arrow). A cervical spine MRI was normal.
This SEP study showed a localized abnormality in the P13/14–N20 IPL after right median nerve stimulation. Further, the abnormality
appeared to be between the N18 and N20 waveforms, further localizing it closer to the somatosensory cortex. This explains why the
right tibial nerve SEP was not abnormal, as often lesions that affect the dorsal column pathways from the median nerve also affect
the same pathways from the tibial nerve. In this case, the lesion was closer to the cortex, where the fibers of the dorsal column
separate, and both can be affected independently. This finding also makes cervical spondylosis less likely and MS more likely as a
cause of this patient’s complaints.
EP, evoked potential; FLAIR, fluid-attenuated inversion recovery; IPL, interpeak latency; LP, lumbar potential; PF, popliteal fossa; SEP, somato-
sensory evoked potential.

Age: 33 years
Sex: Male
Median Nerve SEPs
(A) (B)
61 m/s 60 m/s

EP 10.1 EP 10.3
N13 14.3 N13 13.8
P13 14.9 P13 15.4
N18 17.5 1.14 (P13–N18) N18 18.2 1.41 (P13–N18)
N20 21.3 0.54 (N20–P22) N20 21.3 0.35 (N20–P22)

EP–N20 11.2 EP–N20 11.0
EP–P13 4.8 EP–P13 5.1
P13–N20 6.4 P13–N20 5.9

Tibial Nerve SEPs

(C) (D)
40 m/s 39 m/s

PF 9.9 PF 10.4
LP 20.4 LP 20.9
P31 34.4 P31 36.0
N34 41.4 N34 42.2
P37 47.5 0.17 (P37–N45) P37 47.1 0.72 (P37–N45)

PF–P37 37.6 PF–P37 36.7
LP–P37 27.1 LP–P37 26.2

(E)
History: This 33-year-old male has a history of TM 1 year ago and presents with numbness of all limbs, blurring of vision, and
episodes of confusion. His examination is remarkable for mild bilateral lower extremity weakness. This SEP is requested to see if
there is a lesion to explain his symptoms.
Discussion: The left median nerve SEP (A) is normal, but the right median nerve SEP (B) is abnormal. Both left (C) and right (D) tibial
nerve SEPs are abnormal. Both median nerve SEPs have well-defined waveforms that are correctly tagged. The IPLs after left-sided
stimulation are normal; however, after right-sided stimulation, the EP–P13/14 IPL is slightly prolonged. The EP–N20 and P13/14–N20
IPLs are normal. This suggests a possible lesion between the right brachial plexus and lower medulla.
The tibial nerve SEPs also have well-defined waveforms. After left-sided stimulation, the P37 waveform is of low amplitude. However,
it is correctly tagged; this is confirmed by noting a waveform of comparable latency in the CP2–C5S derivation. The P37 waveform
absolute latencies after left- and right-sided stimulation are prolonged. Additionally, the LP–P37 and PF–P37 IPLs are also prolonged
after stimulation of both sides. This puts the lesion between the level of the lower spinal cord and the somatosensory cortices on
both sides.
An MRI of the spine revealed no abnormalities in spinal cord signal. The brain MRI was not available for review, but was reported
to be normal.
If both median nerve SEPs had been normal and both tibial nerve SEPs showed a prolonged LP–P37 IPL, the most likely localization
for the lesion would have been in the spinal cord below the cervical spine. However, since the EP–P13/14 IPL after right median
nerve stimulation is also prolonged, the lesion could also be higher in the cervical spinal cord or lower medulla. Absence of any
abnormalities after the left median nerve stimulation makes this less likely. It is possible that the right median nerve SEP abnormality
is not related to the tibial nerve SEP abnormalities and may even be due to peripheral nerve or root problems. The MRI did not show
signal change in the spinal cord. However, the previous TM may have resulted in the tibial nerve SEP abnormalities (even though
the MRI evidence has resolved). This patient’s lower extremity complaints might be explained by the abnormality seen on the tibial
nerve SEPs. The right median nerve SEP abnormality might account for the symptoms on that side, but the left-sided complaints do
not have a neurophysiologic correlate to this study.

Age: 42 years
Sex: Male
Median Nerve SEPs
(−↑) (−↑)
(A) (B)
Median nerve conduction velocity Median nerve conduction velocity
(median nerve to EP): 60 m/s (median nerve to EP): 59 m/s

EP 11.5 EP 11.6
N13 N/A N13 15.7
P13 16.1 P13 16.6
N18 19.1 1.86 (P13–N18) N18 19.2 1.74 (P13–N18)
N20 21.1 0.86 (N20–P22) N20 21.7 1.12 (N20–P22)

EP–N20 9.7 EP–N20 10.1
EP–P13 4.7 EP–P13 5.0
P13–N20 5.0 P13–N20 5.2

Tibial Nerve SEPs

(−↑)
(−↑)
(C) (D)
Tibial nerve conduction Tibial nerve conduction
velocity (tibial nerve to T12S): 38 m/s velocity (tibial nerve to T12S): 40 m/s

PF 11.4 PF 10.8
LP N/A LP N/A
P31 40.0 P31 40.9
N34 42.8 N34 43.6
P37 47.8 0.20 (P37–N45) P37 47.2 0.90 (P37–N45)

PF–P37 36.4 PF–P37 36.4

(E)
History: This is a 42-year-old male who presents for evaluation of syncopal spells followed by persistent dizziness. He also has
chronic face pain. He was in a motor vehicle accident about 10 years ago in which he received multiple injuries, including knees,
neck, and head. This study is being done to evaluate for MS.
Discussion: The median nerve SEPs are normal and tibial nerve SEPs are abnormal. After left (A) and right (B) median nerve stimula-
tion, the EP, P13/14, N18, and N20 waveforms are seen. The N13 waveform is seen after right-sided stimulation. All IPLs are normal.
Because the IPLs are normal, the absent N13 waveform on the left side is of uncertain significance.
The left (C) and right (D) tibial nerve SEPs show the PF, P31, N34, and P37 waveforms, but the LP waveform is absent bilaterally. The
absolute latencies of the P37 waveforms are prolonged bilaterally. Because of the absence of the LP waveform, the only IPL that
can be determined is the PF–P37 IPL, both of which are also prolonged. This suggests a lesion between the PF bilaterally and the
somatosensory cortices.
Brain MRI shows punctate white matter abnormalities (E; arrow), and the cervical spine MRI is normal.
Whereas the tibial nerve SEP abnormalities could localize anywhere in the neuroaxis from the PF to the somatosensory cortices bilat-
erally, the normal median nerve SEP makes the abnormality most likely below the cervical spinal cord. However, since the PF–P37
IPL is abnormal and the LP–P37 IPL could not be determined, the lesion(s) could be in the peripheral nerves or lumbosacral plexus
bilaterally. Given the patient’s history of motor vehicle trauma, bilateral peripheral nerve lesions could be possible. Overall, this study
is not helpful in the diagnosis of MS in this patient.
EP, evoked potential; IPL, interpeak latency; LP, lumbar potential; MS, multiple sclerosis; PF, popliteal fossa; SEP, somatosensory evoked
potential.

Age: 47 years
Sex: Male
Median Nerve SEPs
↑(−)
↑(−)
(A) (B)
56 m/s 54 m/s
EP 11.6 EP 12.0
N13 15.6 N13 17.7
P13 16.2 P13 16.0
N18 20.5 0.72 (P13–N18) N18 22.0 0.70 (P13–N18)
N20 24.3 2.04 (N20–P22) N20 23.9 1.48 (N20–P22)

EP–N20 12.7 EP–N20 11.9
EP–P13 4.6 EP–P13 4.0
P13–N20 8.1 P13–N20 8.0

Tibial Nerve SEPs

↑(−) (−)↑
(C) (D)
41 m/s 41 m/s
PF 10.7 PF 10.7
LP 26.2 LP 26.4
P31 N/A P31 N/A
N34 N/A N34 N/A
P37 55.1 0.30 (P37–N45) P37 47.5 0.30 (P37–N45)

PF–P37 44.4 PF–P37 36.8
LP–P37 28.9 LP–P37 21.1

(E) (F)
History: This is a 47-year-old male who presents with a history of low back pain and left lower extremity numbness, weakness,
and gait instability. He has also noticed decreased hearing from his right ear, but he denies tinnitus. His examination is remarkable
for mild weakness in many muscle groups and reduction of vibration and pinprick on his left leg. This SEP study is being done to
evaluate for a lesion that may be causing his symptoms.
Discussion: The median and tibial nerve SEPs are abnormal. After left (A) and right (B) median nerve stimulation, all waveforms are
seen, but the P13/14–N20 and EP–N20 IPLs show prolongation. This suggests a lesion from the caudal medulla to the contralateral
somatosensory cortex.
Left (C) and right (D) tibial nerve SEPs have PF, LP, and P37 waveforms, but do not have anything reproducible in the subcortical
derivation. Both the LP–P37 and the PF–P37 IPLs are prolonged, but the LP–P37 IPL prolongation is more meaningful. The PF–P37
IPL can be prolonged with a peripheral nerve lesion(s), but the LP–P37 cannot. The lesion in this case is most likely between the LP
waveform and the somatosensory cortices.
The brain MRI showed areas of juxtacortical (E) and periventricular (F) hyperintensities on FLAIR images on the right side. There were
similar abnormalities on the left side as well. There were no clear brainstem abnormalities. Thoracic and lumbar spine MRIs were
normal.
The median nerve SEP suggests a lesion(s) between the lower medulla and both somatosensory cortices. The tibial nerve SEP indi-
cates that the lesion(s) is most likely between the lower spinal cord and the somatosensory cortices. Taken together, the median and
tibial nerve SEPs suggest that the lesion is proximal to the nucleus cuneatus and gracilis, likely bilaterally. The MRI findings of bilateral
punctate lesions do not necessarily relate to the SEP findings. These findings suggest that there is a lesion(s) proximal to the lower
medulla that may account for some (i.e., gait instability) but not all of his symptoms.
EP, evoked potential; FLAIR, fluid-attenuated inversion recovery; IPL, interpeak latency; LP, lumbar potential; PF, popliteal fossa; SEP, somato-
sensory evoked potential.

Age: 46 years
Sex: Female
Median Nerve SEPs
(−↑) (−↑)
(A) (B)
59 m/s 59 m/s
EP 10.9 EP 11.2
N13 13.8 N13 14.6
P13 14.5 P13 14.5
N18 18.3 2.09 (P13–N18) N18 18.4 2.05 (P13–N18)
N20 20.5 1.25 (N20–P22) N20 20.5 2.09 (N20–P22)
EP–N20 9.5 EP–N20 9.2
EP–P13 3.5 EP–P13 3.3
P13–N20 6.0 P13–N20 5.9

Tibial Nerve SEPs

(−↑) (−↑)
(C) (D)
43 m/s 43 m/s
PF 8.8 PF 9.1
LP 23.7 LP 23.5
P31 30.9 P31 31.3
N34 34.5 N34 33.8
P37 43.0 0.96 (P37–N45) P37 39.6 1.13 (P37–N45)

PF–P37 34.2 PF–P37 30.5
LP–P37 19.3 LP–P37 16.1

(E)
History: This is a 46-year-old female who presents with episodes of right arm numbness and weakness. She has had previous similar
episodes in the past. Previous brain MRI scans have shown bilateral foci of T2 hyperintensities. This study is requested to further
evaluate her symptoms.
Discussion: Bilateral median and left tibial nerve SEPs are normal, and the right tibial nerve SEP is normal. Both left (A) and right (B)
median nerve SEPs show reproducible waveforms and all IPLs are within normal limits.
Left (C) and right (D) tibial nerve SEPs also show reproducible waveforms. When the absolute latencies of the P37 waveforms and
the LP–P37 and PF–P37 IPLs are compared to normative data, they are all within normal limits. However, when the LP–P37 IPLs are
compared between the two sides, there is a greater than normal asymmetry between the two sides (refer to normative data), with
right-sided stimulation resulting in a longer latency. Thus, despite the overall latencies being within normal limits, this is an abnor-
mal right tibial nerve SEP due to the asymmetry of the LP–P37 IPL. This abnormality is suggestive of a lesion in the dorsal column
pathway between the lower spinal cord and the left somatosensory cortex.
A repeat brain MRI scan revealed the T2 hyperintensities in bilateral periventricular white matter. A left parietal periventricular abnor-
mality is prominent (E, arrow). A cervical spine MRI was normal.
Because both median nerve SEPs are normal, the abnormality of the right tibial nerve SEP is most likely below the cervical spinal cord
on the right side. However, a small lesion on the left side of the brain, particularly in the white matter affecting the dorsal column
pathway, cannot be excluded. The latter localization may be the correct one for this patient given the MRI findings. This abnormality,
however, does not help explain the patient’s symptoms of right arm numbness and weakness.

Age: 40 years
Sex: Female
Median Nerve SEPs
(−↑) (−↑)
(A) (B)
Median nerve conduction velocity (median nerve to Median nerve conduction velocity (median nerve to
EP): N/A EP): N/A
EP N/A EP N/A
N13 N/A N13 N/A
P13 N/A P13 N/A
N18 N/A N/A (P13–N18) N18 N/A N/A (P13–N18)
N20 N/A N/A (N20–P22) N20 N/A N/A (N20–P22)

EP–P13 N/A EP–P13 N/A
P13–N20 N/A P13–N20 N/A

Tibial Nerve SEPs

(−↑)
24.4
(C) (D)
43 m/s 43 m/s
LP N/A LP N/A
P31 N/A P31 N/A
N34 N/A N34 N/A
P37 N/A N/A (P37–N45) P37 N/A N/A (P37–N45)

(E)
History: This 40-year-old female with a history of alcohol and cocaine abuse presents with progressive weakness and numbness
of both lower extremities. She has had instances when she fell because her legs did not support her. This study is being done to
help localize her symptoms.
Discussion: Bilateral median and peroneal nerve SEPs are abnormal. Left (A) and right (B) median nerve SEPs show very little noise
and no reproducible waveforms except for the artifact occurring at about 5 ms. Similarly, left (C) and right (D) peroneal nerve
SEPs show no reproducible waveforms. Initially, tibial nerve SEPs were attempted, but since no response was obtained (waveforms
not shown), peroneal nerve SEPs were performed instead. Notice the long stimulus pulse duration and high stimulation intensity.
Whereas severe spinal cord or brain lesions could cause complete loss of SEPs, this is unlikely in this case as the EP waveforms are
absent after median nerve stimulation. These should be present in a spinal cord lesion. A more likely etiology is a severe peripheral
neuropathy, making it impossible to adequately activate the dorsal column pathways.
A brain MRI (E) showed a left-sided white matter lesion that enhanced with contrast. An EMG showed a severe peripheral neurop-
athy with demyelinating features.
When there is complete absence of all waveforms of the SEP, technical issues must be considered. It must be verified that stimulation
is being delivered, and there is not a failure of current delivery to the patient. Once that has been verified, absence of waveforms
can be attributed to patient pathology. In this case, complete loss of all waveforms is likely due to a severe neuropathy, making it
impossible to adequately elicit an SEP. The presence of the artifact in the median nerve SEP helps verify that stimulation was being
delivered to the patient. The MRI finding in this patient is not related to the SEP abnormalities.
EMG, electromyographic; EP, evoked potential; LP, lumbar potential; SEP, somatosensory evoked potential.

Age: 68 years
Sex: Male
(A)
Side: Left
Absolute Latencies
N13 N/A Waveforms Latency (ms)
P13 18.5 EP–N20 N/A
N18 20.7 0.58 (P13–N18) EP–P13 6.5
N20 N/A N/A (N20–P22) P13–N20 N/A

(B)
Side: Right
(−↑) Number of repetitions: 1,000
Absolute Latencies
N13 N/A Waveforms Latency (ms)
P13 18.4 EP–N20 N/A
N18 20.2 0.70 (P13–N18) EP–P13 7.0
N20 N/A N/A (N20–P22) P13–N20 N/A
History: This is a 68-year-old male who was discovered unconscious at home. The rescue squad was called and he was noted to be
pulseless. He was resuscitated and brought to the hospital. A therapeutic hypothermia protocol was instituted and after rewarm-
ing he remained unresponsive. Two days after the rewarming, this study was requested for prognostic assessment. He was not on
sedating medication and was normothermic.
Discussion: Both left (A) and right (B) median nerve SEPs are abnormal. After stimulation of either median nerve, a cortical N20
waveform is not seen despite little noise in the cortical channel. Notice the long stimulus pulse duration and high stimulus intensity.
In anoxic encephalopathy, the absence of the N20 waveforms of the median nerve SEP is highly consistent with a poor outcome
(death or persistent vegetative state). The presence of the EP, P13/14, and N18 waveforms is important as it confirms technical
adequacy of stimulation.

Age: 66 years
Sex: Male
(A)
Side: Left
Absolute Latencies
P13 18.2 EP–N20 10.5
N18 21.5 0.98 (P13–N18) EP–P13 4.8
N20 23.8 0.75 (N20–P22) P13–N20 5.6

(B)
Side: Right
↑(−) Number of repetitions: 2,400
Absolute Latencies
P13 17.8 EP–N20 10.8
N18 21.8 1.66 (P13–N18) EP–P13 4.7
N20 23.8 0.70 (N20–P22) P13–N20 6.1
History: This is a 66-year-old male who recently underwent a coronary artery bypass grafting surgery and was recovering at a reha-
bilitation facility. He was noted to be unresponsive and pulseless; cardiopulmonary resuscitation was initiated and was successful.
Therapeutic hypothermia was not initiated. This study was requested 3 days after resuscitation. He was not on sedating medication
and was normothermic.
Discussion: Both left (A) and right (B) median nerve SEPs are normal. All waveforms are present. Most importantly, the N20 wave-
forms are present after stimulation of both left and right median nerves.
While the absence of N20 waveforms after median nerve stimulation is interpreted as consistent with poor prognosis, the presence
of N20 waveforms should not be interpreted as consistent with good prognosis. Rather the interpretation of the presence of these
waveforms should be that the SEP is not helpful for prognosis.

Age: 58 years
Sex: Male
(↑−) (−↑)
(A) (B)
EP 12.3 EP 13.1
N13 N/A N13 N/A
P13 N/A P13 N/A
N18 N/A N/A (P13–N18) N18 N/A N/A (P13–N18)
N20 N/A N/A (N20–P22) N20 N/A N/A (N20–P22)

P13–N20 N/A P13–N20 N/A

Repeat Median Nerve SEPs performed 6 days later

(−↑)
(−↑)
(C) (D)
EP N/A EP N/A
N13 N/A N13 N/A
P13 16.8 P13 16.2
N18 19.2 1.24 (P13–N18) N18 19.0 0.96 (P13–N18)
N20 24.6 0.25 (N20–P22) N20 24.0 0.24 (N20–P22)

P13–N20 7.8 P13–N20 7.8

History: This is a 58-year-old male with coronary artery disease who suffered a cardiac arrest at home. He was resuscitated in the
field and transported to the hospital. He underwent therapeutic hypothermia protocol. On day 3 of admission he was comatose,
and this prognostic median nerve SEP was performed. At the time of the study (A and B), he was not on sedating medications and
his temperature was 35.6°C. Both upper limbs were reported as very swollen. Six days after the first study, median nerve SEPs were
repeated (C and D) as the patient remained comatose despite not receiving sedating medication. He was normothermic.
Discussion: The first set of median nerve SEPs do not show clear N20 waveforms after stimulation of either the left or the right
side. However, in the second study, repeated 6 days later, N20 waveforms are seen.
The first prognostic left (A) and right (B) median nerve SEP study was obtained 3 days after the cardiac arrest and immediately after
rewarming from a therapeutic hypothermia protocol. Clear EP waveforms are noted on both sides. After right-sided stimulation,
there are clear P13/14 and N18 waveforms, though they have not been tagged. After left-sided stimulation, these waveforms are
also seen, though not as clearly as after right-sided stimulation. The cortical channel is noisy after stimulation on both sides, and
there may be a questionable N20 waveform (arrow) after left-sided stimulation (though it is not tagged). Because the cortical chan-
nel has considerable artifact and is noisy, it is important to interpret this prognostic median nerve SEP study with caution. With these
concerns, it is best to interpret this study as “indeterminate” for prognosis (as opposed to absent N20 waveforms and consistent
with poor prognosis), and a suggestion should be made to consider repeating it in a few days.
The left (C) and right (D) median nerve SEP study was repeated 6 days later. In this study, clear N20 waveforms are seen with both
left- and right-sided stimulation. The P13/14 and N18 waveforms are also seen, but clear EP and N13 waveforms are not. The latter
are not important in this study as it is being done for prognosis, and the significant finding here is the presence of N20 waveforms.
Thus, this study is now interpreted as not being helpful for prognosis. Whenever unclear or unreliable waveforms are obtained in a
prognostic median nerve SEP study, a repeat study at a later date should be considered.
Conclusions
SEPs can be obtained by electrically stimulating of the neuroaxis. Some of the commonest uses of
almost any peripheral (or cranial) nerve. Median, SEPs include detecting lesions in the nervous sys-
ulnar, tibial, and peroneal nerve SEPs are performed tem that may not be clinically obvious, neurophysi-
most often. The montages used for recording SEPs ologically following various progressive neurologic
track the depolarization as it ascends in the dorsal diseases, assessing prognosis of coma, and pre-
column pathway. By combining upper and lower venting injury to neural structures in the operating
limb SEPs, lesions can be localized to various parts room.
References
1. Cusick JF, Myklebust J, Larson SJ, Sances A Jr. Spinal 5. Levy SR. Somatosensory evoked potentials in pediat-
evoked potentials in the primate: neural substrate. rics. In: Chiappa KH, ed. Evoked Potentials in Clinical
J Neurosurg. 1978;49:551-557. Medicine. 3rd ed. Philadelphia, PA: Lippincott-Raven
2. ACNS. Guideline 9D: Guidelines on short-latency Publishers; 1997:453-469.
somatosensory evoked potentials. J Clin Neurophysiol. 6. George SR, Taylor MJ. Somatosensory evoked poten-
2006;23:168-179. tials in neonates and infants: developmental and
3. Yamada T. Neuroanatomic substrates of lower normative data. Electroencephalogr Clin Neurophysiol.
extremity somatosensory evoked potentials. J Clin 1991;80:94-102.
Neurophysiol. 2000;17:269-279. 7. Gilmore R, Brock J, Hermansen MC, Baumann R.
4. Moncho D, Poca MA, Minoves T, et al. Interside Development of lumbar spinal cord and cortical
latency differences in brainstem auditory and evoked potentials after tibial nerve stimulation in
somatosensory evoked potentials. Defining upper the pre-term newborns: effects of gestational age and
limits to determine asymmetry. J Clin Neurophysiol. other factors. Electroencephalogr Clin Neurophysiol.
2015;32:424-427. 1987;68:28-39.
8. Chiappa KH, Hill RA. Short-latency somatosen- 16. Tobimatsu S, Fukui R, Kato M, et al. Multimodality
sory evoked potentials: interpretation. In: Chiappa evoked potentials in patients and carriers with
KH, ed. Evoked Potentials in Clinical Medicine. 3rd ed. adrenoleukodystrophy and adrenomyeloneu-
Philadelphia, PA: Lippincott-Raven Publishers; ropathy. Electroencephalogr Clin Neurophysiol.
1997:341-400. 1985;62:18-24.
9. Sival DA, du Marchie Sarvaas GJ, Brouwer OF, et 17. Young GB. Clinical practice. Neurologic prog-
al. Neurophysiological evaluation in children with nosis after cardiac arrest. N Engl J Med. 2009;361:
Friedreich’s ataxia. Early Hum Dev. 2009;85:647-651. 605-611.
10. Cascino GD, Ring SR, King PJ, et al. Evoked potentials 18. Wijdicks EF, Hijdra A, Young GB, et al. Practice
in motor system diseases. Neurology. 1988;38:231-238. parameter: prediction of outcome in comatose
11. Matheson JK, Harrington HJ, Hallett M. survivors after cardiopulmonary resuscitation
Abnormalities of multimodality evoked poten- (an evidence-based review): report of the Quality
tials in amyotrophic lateral sclerosis. Arch Neurol. Standards Subcommittee of the American Academy
1986;43:338-340. of Neurology. Neurology. 2006;67:203-210.
12. Misulis KE. Spehlmann’s Evoked Potential Primer. 19. Mandel R, Martinot A, Delepoulle F, et al. Prediction
2nd ed. Boston, MA: Butterworth-Heinemann; 1994. of outcome after hypoxic-ischemic encephalopathy:
13. De Meirleir LJ, Taylor MJ, Logan WJ. Multimodal a prospective clinical and electrophysiologic study.
evoked potential studies in leukodystrophies of chil- J Pediatr. 2002;141:45-50.
dren. Can J Neurol Sci. 1988;15:26-31. 20. Chiappa KH. Short-latency somatosensory evoked
14. Markand ON, Garg BP, DeMyer WE, et al. Brain potentials: methodology. In: Chiappa KH, ed.
stem auditory, visual and somatosensory evoked Evoked Potentials in Clinical Medicine. 3rd ed.
potentials in leukodystrophies. Electroencephalogr Philadelphia, PA: Lippincott-Raven Publishers;
Clin Neurophysiol. 1982;54:39-48. 1997:283-339.
15. Husain AM, Altuwaijri M, Aldosari M. Krabbe dis- 21. Haque IU, Udassi JP, Zaritsky AL. Outcome follow-
ease: neurophysiologic studies and MRI correla- ing cardiopulmonary arrest. Pediatr Clin North Am.
tions. Neurology. 2004;63:617-620. 2008;55:969-987, xi.
5
Evoked Potentials During Surgery
Over the last several decades, evoked potentials the clinical laboratory are presented in Chapter 3.
have been increasingly used to monitor the integ- In this chapter, the ways in which BAEP acquisition
rity of the nervous system during various types of and analysis differ when performed for monitoring
surgery. Initially, brainstem auditory and somato- neural function in the OR will be presented.
sensory evoked potentials (BAEPs and SEPs) were
used for various types of surgeries involving the
brainstem and spine. In the last 15 to 20 years, Anatomy and Physiology
motor evoked potentials (MEPs) have also been Of the various types of AEPs, BAEPs are the type
used to assess the motor pathways during surgery. used most commonly to monitor auditory and
The BAEPs and SEPs performed in the operating brainstem function during surgery. Middle- and
room (OR) are similar in many ways to those done long-latency AEPs are greatly affected by anes-

in the clinical laboratory setting. MEPs performed thetic and cannot be reliably obtained in the OR.
in the OR are obtained with high-intensity electrical Moreover, their generators are less clearly identi-
stimulation applied to the scalp and are not done fied. The electrocochleogram (ECochG), in partic-
in a clinical laboratory setting. When injury occurs ular the vestibulocochlear nerve action potential,
to neural tissue during surgery, evoked potential which occurs in the first few ms after auditory
amplitudes decrease and latencies increase. For stimulation, can be recorded in the OR with spe-
each modality, “warning criteria” represent the cial electrodes, but this is also not done routinely.
extent of change in the evoked potential that should The focus of discussion in this section will be on
prompt a warning to the surgeon of impending neu- BAEPs.
ral compromise (1). In this chapter, a brief overview The auditory pathway from the ear to the
will be presented about the use of BAEP, SEP, and brainstem and cortex has been discussed in
MEP during surgery. For a more in-depth review of detail in Chapter 3. It is important to remember
evoked potentials and other types of neurophysio- that BAEP waveforms are generated from the
logical tests used in the OR, the reader is referred to vestibulocochlear nerve and brainstem s tructures.
dedicated texts in this subject (2–5). There are five BAEP waveforms that are t ypically
recorded. They are numbered sequentially as
Brainstem Auditory Evoked waves I to V. The exact generators of the various
BAEP waveforms are uncertain. It is unlikely
Potentials that a single neural generator produces a single
BAEPs are a type of auditory evoked potential waveform. However, it is useful to consider

(AEP) recorded after aural stimulation. The certain neural structures as generators of various
BAEP waveforms occur within 10 ms of stimu- waveforms to help in localizing lesions. The var-
lation and arise from various structures within ious waveforms and their putative generators are
the brainstem. Details about BAEPs acquired in presented in Table 5.1.
265
Table 5.1 BAEP Waveforms and Their Neural Generators

Wave I Distal vestibulocochlear nerve
Wave II Proximal vestibulocochlear nerve and cochlear nuclei
Wave III Superior olivary complex
Wave IV Lateral lemniscus
Wave V Inferior colliculus
Methodology prevent seepage of fluid into the ear as that would

affect stimulus delivery. The entire ear is then cov-
The methods for obtaining a BAEP study in the ered with a dressing to completely secure the setup.
OR are similar to the methods used in the clinical The tubing must also be protected so that it is not
laboratory, with a few important differences. In this clamped during patient positioning and draping.
section, these differences will be highlighted.
Recording techniques
Stimulation techniques
Recording electrode placement for BAEPs in the
In the OR, BAEPs are obtained with the same type of OR is similar to that used in the c linical laboratory:
stimulus used in the clinical laboratory, broadband a vertex (Cz) and two ear (or mastoid) electrodes
clicks. These clicks are square wave pulses of sound are used. A ground electrode is also used, but when
with a duration of 100 μs that contain many frequen- doing multimodality monitoring, a single ground is
cies. The high-frequency sounds that activate the used for all modalities. The ground electrode can be
basal part of the cochlea contribute most to the BAEP. placed where convenient and not disruptive to the
Rarefaction and condensation clicks can be used surgery. Unlike in the clinical laboratory, in the OR
in the OR. Rarefaction clicks often produce a better needle electrodes are used more often than surface
morphology in the waveforms, particularly wave I, electrodes. These can be applied quickly and can be
whereas condensation clicks often show a clearer secured easily. Needle electrodes, however, must
wave V. However, since the sound intensity used in be applied after the patient has been anesthetized.
the OR is higher than that used in the clinical labora- A two-channel montage, Cz–Ai (ipsilateral ear)
tory, typically 90 dBnHL, the stimulus artifact can be and Cz–Ac (contralateral ear), is sufficient for most
significant. To eliminate the stimulus artifact, alter- cases of monitoring neural function in the OR.
nating clicks are used most often. This eliminates A third channel, Ac–Ai, can also be added if needed.
the stimulus artifact and often makes the earlier The various derivations allow clearer identification
waves of the BAEP easier to recognize. However, if of the various waveforms. A complete discussion
the identification of wave I is difficult, using rarefac- on how to identify various waveforms is presented
tion (and sometimes condensation) clicks may help. in Chapter 3.
The auditory stimulus is delivered to the external The amplifier filter settings are a little different
auditory canals by ear inserts that are connected than in the clinical laboratory. In the OR, filter set-
to the transducer. In the clinical laboratory, head- tings of 150 to 1,500 Hz are often used. In the clinical
phones on the ears can be used to deliver the stimu- laboratory setting, 10 to 3,000 Hz is recommended by
lus, but their bulk makes it impossible to use in the the American Clinical Neurophysiology Society (6).
OR. Adding ear inserts increases the latency of the Increasing the low-frequency filter to 150 Hz
BAEP waves, but this is not clinically significant as reduces the baseline sway and slow frequency activ-
BAEPs obtained in the OR are not compared to nor- ity. Reducing the high-frequency filter to 1,500 Hz
mative data; rather they are compared to baseline reduces a lot of the fast noise activity seen in the
data acquired at the start of the case. The ear inserts OR without appreciably changing the morphology
must be applied in a manner that they fit snugly of the waveform (7). However, since filter changes,
in the external ear canal and do not move during particularly increasing the low-frequency filter, can
the surgery. A waterproof seal must be applied to change the latency of the waveforms, the filters
Chapter 5. Evoked Potentials During Surgery ■ 267
should not be changed in the middle of a surgical The warning criteria may need to be amended
case. Rather all filter settings should be adjusted at depending on the type of surgery. For microvascu-
the start of a case and not altered thereafter. Often lar decompression surgeries for conditions such as
1,000 to 3,000 repetitions are needed to obtain a trigeminal neuralgia and hemifacial spasms, hearing
reliable and reproducible waveform. does not seem to be affected unless there is complete
loss of wave V (10). On the other hand, in patients
Anesthetic Considerations with cerebellopontine angle tumors, a much smaller
Anesthetic agents can reduce the amplitude and pro- latency shift and drop in amplitude of wave V may
long the latency of any evoked potential, with inhala- result in hearing loss. Thus a “sliding scale” approach
tional agents having a greater effect than intravenous depending on the pathology may be needed to deter-
ones. Generally, cortical w aveforms are affected mine the warning criteria for a given patient.
more than subcortical ones. BAEP waveforms are Although a 50% drop in amplitude is commonly
all generated by subcortical structures, so BAEP are regarded as the “alert criterion,” it must be under-
typically less affected by anesthetics than other types stood that these criteria are empirically established.
of evoked potentials. Halogenated inhalational As noted earlier, this may need to be changed
agents have no significant effect on BAEPs unless depending on pathology. Moreover, a change of less
the concentration becomes very high, at which point than 50% drop in amplitude may be significant. If
the waveforms decline in amplitude and increase the BAEP wave V has had consistent amplitude and
in latency. Nitrous oxide and intravenous agents latency, and suddenly there is a drop in the ampli-
have no significant effect on BAEPs. Neuromuscular tude by 30%, that may be a meaningful change and
blocking agents can enhance BAEPs by reducing the surgeon should be alerted. The suddenness of the
background EMG activity and making averaging change makes it significant. On the other hand, if the
easier. With neuromuscular blocking agents, fewer BAEP wave V has fluctuating amplitude at the start
repetitions are needed to obtain reliable BAEPs. of a case, a 50% change in amplitude may not be very
Though they are of very low amplitude, BAEPs meaningful. It may simply be the normal variabil-
are among the most anesthetic insensitive evoked ity of the waveform. In the author’s practice, when
potentials recorded in the OR. the wave V reaches a 0.5-ms increase in latency,
the surgeon is notified of the change. With further
Warning Criteria latency increase or amplitude decrement, the sur-
geon is kept informed. When the shift reaches 1.0 ms
Interpretation of BAEPs during surgery requires
or amplitude decrease reaches 50%, the surgeon is
identifying when a significant change has occurred
again informed of the magnitude of the findings.
to the BAEP waveforms that may indicate compro-
A complete loss of a BAEP waveform, such as
mise in neural function. It is of little consequence
wave V, should be immediately reported to the
how an individual’s BAEP compares to normative
surgeon. When BAEP waveforms are lost during
data; intraoperative BAEPs are interpreted in rela-
surgery, the probability of complete deafness in
tion to how they compare to BAEPs acquired at the
that ear is high, but not 100%. Some patients who
start of surgery (baseline).
lose the BAEP completely will still be able to hear.
The warning criteria for BAEPs is a latency pro-
Rarely, patients may lose hearing despite preserved
longation of 1.0 ms or an amplitude decrement
BAEP waveforms (11).
of 50% of wave V (7). A stricter alert criterion of
latency of wave V prolongation of 0.5 ms has also
been proposed (8). However, with a 1.0-ms wave V Mechanisms of Waveform Change
latency prolongation, the probability of hearing loss During the course of a surgical procedure, BAEP
is much more likely with few false negatives (9). waveforms can change in amplitude or latency.
This does not mean that an alert should not be raised This could occur due to injury to the auditory
until these parameters are met, rather one should pathways due to the surgical procedure itself, due
inform the surgeon of progressive changes in the to homeostatic changes, or due to technical issues
BAEP leading up to the threshold criteria noted ear- with the monitoring (12). It is important to distin-
lier. Unlike with other types of evoked potentials, a guish between these and communicate this to the
latency prolongation is more important than ampli- surgical and anesthesia team. Changes induced by
tude reduction for BAEPs. the surgical procedure are the most important, but
must be distinguished from the others to provide enough, BAEP waveforms can be affected. A drop in
effect monitoring. temperature can have a significant effect on BAEPs.
As temperatures decrease, there is prolongation
Surgery-induced changes of BAEP waveform latencies. The amplitude may
BAEP monitoring is often used when there is risk increase with a slight reduction in temperature, but
of injury to the auditory pathways in the vestib- thereafter decreases. At around 20°C, BAEP wave-
ulocochlear nerve and brainstem. Additionally, forms disappear (13). Anesthetic changes can also
BAEP monitoring can detect injury to the cochlea. affect BAEPs as discussed earlier in this chapter.
However, BAEP monitoring cannot detect injury to
auditory pathways above the level of the inferior Technical issues
colliculus. A summary of the BAEP changes that The OR is a challenging environment in which to
can occur during surgery is presented in Table 5.2. obtain evoked potentials. A lot of environmental
noise and urgency for setting up the monitoring can
Physiologic changes lead to technical problems that are not e ncountered
Changes in the physiologic milieu can affect the in the clinical laboratory. Technical issues that
BAEPs. The changes are usually bilateral and sym- cause the BAEP potential to change may be due
metric. Compared to other types of evoked poten- to ineffective stimulation or recording. Common
tials, BAEPs are relatively insensitive to p
hysiologic technical reasons that result in BAEP changes are
changes. However, if the changes are severe listed in Table 5.3.
Table 5.2 Patterns of Surgery-Induced Changes in BAEPs
Location Cause Effect on BAEP
Cochlea Trauma, ischemia ↑Latency of all waves, ↓Amplitude of wave I
Proximal CN VIII Stretch from retraction, mechanical, Wave I preserved, others ↑latency, ↓amplitude
vascular, thermal
Lower pons Compression, thermal, vascular Wave I preserved, others ↑latency, ↓amplitude
Mid pons to lower midbrain Compression, thermal, vascular Waves I and III preserved, wave V ↑latency,
↓amplitude
Rostral to lower midbrain Compression, thermal, vascular BAEP may not detect damage
Source: Ref. (12). Husain AM. Neurophysiologic intraoperative monitoring. In: Ebersole JS, Husain AM, Nordli DR, eds. Current
Practice of Clinical Electroencephalography. Philadelphia, PA: Wolters Kluwer Health; 2014: 488-542.
Table 5.3 Common Technical Issues Causing Changes in BAEPs
Ineffective Stimulation Ineffective Recording
Transducer failure Dislodgement of recording electrodes
Kinking of tubing leading to ear inserts High electrode impedance
Dislodgement of ear inserts Amplifier failure
Compromised seal around ear; fluid in ear canal Surgical equipment (drill, electrocautery, CUSA, others)
BAEP, brainstem auditory evoked potential; CUSA, cavitational ultrasonic surgical aspirator.
Clinical Applications malformations. Other types of monitoring, such as

SEP and MEP monitoring, are also used in conjunc-
BAEP monitoring in the OR can be used during tion with BAEP monitoring in these surgeries (15).
any surgery in which there is risk of injury to Surgeries on the base of the skull in which multiple
auditory pathways. Monitoring of BAEPs during cranial nerves are at risk, including the vestibuloco-
microvascular decompression surgeries for trigem- chlear, can also employ BAEP monitoring to reduce
inal neuralgia and hemifacial spasms has shown to the risk of hearing loss.
reduce the incidence of hearing loss (14).
Surgery on vestibular schwannomas and other
cerebellopontine angle tumors can also result in hear- Examples
ing loss. Several surgical approaches can be used to Examples of BAEP monitoring are provided
gain access to the tumor and with most approaches, (Figures 5.1–5.4, Examples 1–4). A stack of BAEP
BAEP monitoring has been used to reduce the risk waveforms are shown in the sequence in which they
of hearing loss (9). If the vestibulocochlear nerve is were acquired. Important surgical events are noted
to be sacrificed during the surgery, BAEP monitor- as well. The narrative provides a brief history and
ing is not helpful. a discussion of the important findings. In several of
BAEP monitoring is also useful during surgery the examples, BAEP monitoring was performed as
on other types of brainstem masses, such as v ascular part of multimodality monitoring.
Age: 52 years
Sex: Female
Montage:
Left-sided stimulation: Cz–A1, Cz–A2, A2–A1
Right-sided stimulation: Cz–A2, Cz–A1, A1–A2
Diagnosis: Right-sided inferior tentorial tumor
Procedure: Right retromastoid craniotomy for tumor resection
Modalities monitored: Bilateral BAEPs; bilateral median nerve SEP; right-sided cranial nerves V and VII
(A)
(B)

(C)
(D) (E)

History: This is a 52-year-old female who presents for numbness and tingling on the right side of her face that started about 6
months ago. Cold temperature, chewing, and brushing her teeth make her s ymptoms worse. Her examination is normal, including
facial sensations and facial movements. Her MRI shows a right-sided homogenously enhancing inferior tentorial tumor. She is under-
going a right retromastoid craniotomy for resection of the brain tumor.
Discussion: BAEP, median nerve SEP, and cranial nerves V and VII m
onitoring are requested for this surgery. The BAEP monitoring is
shown here. The maximum shift in the wave V obtained after r ight-ear stimulation is 0.46 ms. The surgery is shown in three panels
(A, B, and C). The surgeon was kept informed of the changes. At the end of surgery, the wave V was approximate to baseline.
An MRI of the brain shows a mass in the right inferior tentorial region. The mass could be seen on an axial T2 weighted image as
increased signal (D; arrow), but was more clearly seen as a homogenously enhancing mass in a contrast-enhanced coronal image (E).
In the panels in A to C, BAEPs from the left and right ear are shown. The montage labels appear at the top of each column. Responses
are obtained with good reproducibility at baseline, which is shown as the last tracing in panels A to C. After right-ear stimulation, clear
waves V and Vc are noted in the Cz-A2 and Cz-A1 derivations, respectively. Wave I and III are seen in the A1-A2 derivation, but a wave V
is not clearly seen. All waveforms are seen after contralateral (left) ear stimulation. As the surgery progresses, the wave V after right-sided
stimulation becomes slightly prolonged. At the end of panel B, note that the latency prolongation of wave V is noted at 0.38 ms. Even
though the notation appears in the panel for the BAEP obtained by stimulation of the left ear, the change is noted in the BAEP obtained
from right-sided stimulation. The n otation appears in different locations per the software. In the third panel (C), the latency shift of wave
V is noted at 0.46 ms. The surgeon is kept informed of the gradual change in the latency. As the surgery comes to an end, there is a
gradual return of the wave V latency to near baseline. No significant change was noted in the wave V amplitude. No significant change
was noted in the BAEP obtained after left-ear stimulation.
In this surgery, BAEP monitoring did not detect any irreversible injury to the auditory pathways up to the upper brainstem.
BAEP, brainstem auditory evoked potential; SEP, somatosensory evoked potential.

Age: 79 years
Sex: Female
Montage:
Left-sided stimulation: Cz–A1, Cz–A2, A2–A1
Right-sided stimulation: Cz–A2, Cz–A1, A1–A2
Diagnosis: Left-sided hemifacial spasms
Procedure: Left retromastoid craniotomy for microvascular decompression of cranial nerve VII
Modalities monitored: Bilateral BAEPs; bilateral median nerve SEP; left-sided cranial nerves V and VII
(A)
(B)

(C)
(D)

(E)
History: This is a 79-year-old female who presents for left-sided hemifacial spasms. She has tried conservative treatment and botu-
linum toxin but has not had significant relief. On examination, the spasms are evident. Her MRI is normal. She is undergoing a left
retromastoid craniotomy for microvascular decompression of the left facial nerve.
Discussion: BAEP, median nerve SEP, and cranial nerves V and VII m onitoring are requested for this surgery. The BAEP monitoring is
shown here in four panels (A–D). The maximum shift in the wave V obtained after left-ear stimulation is 1.95 ms. The surgeon was
kept informed of the changes, in particular when the latency shift reached 1.0 ms and thereafter. At the end of surgery, the wave
V latency shift started to improve, but had not reached baseline.
An MRI of the brain was normal. An axial FLAIR image is shown (E).
In the panels in A to D, BAEPs from the left and right ear are shown. The montage labels appear at the top of each column.
Responses were obtained with good reproducibility at baseline, which is shown as the last tracing in panels A to D. After left-ear
stimulation, a clear wave V is noted in the Cz–A1 derivations, and a less clear wave Vc can also be seen in the Cz–A2 derivation. All
waveforms are seen after right (contralateral) ear stimulation. Soon after start of the surgery, as the surgeon retracts the cerebellum,
there is a shift of wave V latency after left-ear stimulation (A). A little later, the wave V further prolongs in latency and amplitude
decreases as well (at approximately 10:47:02, B). By 11:21:19, the wave V amplitude starts to increase slightly, but the latency
remains prolonged (C). Toward the end of surgery (D), the wave V amplitude appears to be stable, but the latency is still prolonged,
though improving. No significant change is noted in the BAEP obtained after right-ear stimulation. Notice that fewer BAEP trials are
done with right-ear stimulation as that is not the operative ear.
In this surgery, a significant shift in latency of the wave V after left-ear stimulation occurred. This could result in postoperative hear-
ing impairment. However, since this surgery was for microvascular decompression and the wave V was present and recovering at
the end of surgery, it is possible that there will be no significant clinically detectable hearing loss (see text for details). In the case of
tumors in the cerebellopontine angle, such significant changes in wave V latency are more likely to be associated with postoperative
hearing loss.
BAEP, brainstem auditory evoked potential; FLAIR, fluid-attenuated inversion recovery; SEP, somatosensory evoked potential.

Age: 61 years
Sex: Male
Montage:
Left-sided stimulation: Cz–A1, Cz–A2
Right-sided stimulation: Cz–A2, Cz–A1
Diagnosis: Left-sided cerebellopontine angle tumor
Procedure: Left transtemporal craniotomy for tumor resection
Modalities monitored: Bilateral BAEPs
(A)

(B) (C)
History: This is a 61-year-old male who has a history of prior surgery for a supratentorial meningioma, now presenting with left
facial numbness and gait instability. His examination is significant for loss of sensation in the entire left side of his face. An MRI
scan of the brain showed a left cerebellopontine angle mass, likely a cavernoma. He is undergoing a left transtemporal craniot-
omy for resection of the cerebellopontine angle mass.
Discussion: BAEP monitoring is performed during this surgery. The maximum shift in the wave V obtained after left-ear stimulation
is 1.20 ms. Toward the end of surgery, the wave V started to return to baseline latency, but still remained slightly prolonged.
An MRI of the brain showed a left-sided cerebellopontine angle lesion that had evidence of hemorrhage within it. An axial T2 image
(B) shows this lesion. It enhanced with contrast, as seen in the axial contrast-enhanced image (C).
In this surgery, BAEP from the left (operative) side was present with good reproducibility, and is shown as the last tracing in panel A.
The montage labels appear at the top of each column. As the surgery progresses, the latency of wave V obtained after left-ear stim-
ulation clearly prolongs, reaching a maximum of 1.20 ms. There is also an amplitude drop, but it is less than 50%. The surgeon is
kept informed of the progressive change in the wave V latency. Toward the end of surgery, the wave V improves and almost reaches
baseline latency. Notice that fewer BAEP trials are done with right-ear stimulation as that is not the operative ear.
In this surgery, BAEP monitoring showed a significant shift in wave V latency of the BAEP obtained from the operative ear that
returned to near baseline by the end of the procedure. This degree of latency change may be associated with ipsilateral hearing loss,
especially in cases of tumors.

Age: 64 years
Sex: Female
Montage:
Left-sided stimulation: Cz–A1, Cz–A2
Right-sided stimulation: Cz–A2, Cz–A1
Diagnosis: Left-sided trigeminal neuralgia
Procedure: Left retromastoid craniotomy for microvascular decompression of cranial nerve V
Modalities monitored: Bilateral BAEPs, bilateral median nerve SEP, and right-sided cranial nerves V and VII
(A)

(B)
History: This is a 64-year-old female with a history of right-sided trigeminal neuralgia who has failed conservative treatment. She is
presenting for right-sided retromastoid craniotomy for microvascular decompression of the trigeminal nerve.
Discussion: BAEP, median nerve SEP, and cranial nerves V and VII monitoring are requested for this surgery. The BAEP monitoring
toward the end of surgery is shown here (A). Toward the end of surgery, there is considerable improvement of the BAEP waveforms
after right-ear stimulation. The final BAEP responses look comparable to those obtained at baseline (last trace of panel A).
An MRI of the brain showed a mild inferior deviation of the cisternal portion of the right trigeminal nerve with several adjacent small
vessels. An axial constructive interference in steady state (CISS) image with the right trigeminal nerve (arrow) is shown (B). These
findings are often noted in patients with trigeminal neuralgia.
In this surgery, BAEPs from the right (operative) ear are clearly present as baseline, shown as the last tracing in panel A. The montage
labels appear at the top of each column. As the surgery progresses, the surgeon adds skin hooks and introduces instrumentation into
the surgical field. This results in deterioration of the BAEP obtained after right-ear stimulation. These changes are evident when the
BAEP waveforms on the top of the panel (about 10:35:53) are compared to the baseline BAEP at the bottom of panel A. The surgeon
is informed of these changes. Toward the end of surgery, after the skin hooks are removed, the BAEP response improved remarkably.
At the end of surgery, the BAEP from the operative ear is comparable to baseline. Notice that fewer BAEP trials are done with left-ear
stimulation as that is not the operative ear.
Skin hookshooks are often used in these surgeries to improve exposure. They can cause a constriction of the tubing from the
transducer to the ear inserts. This reduces the auditory stimulation delivered to the patient and affects the BAEP by reducing the
amplitude and increasing the latency of the waveforms. This change was due to technical issues.
BAEP, brainstem auditory evoked potential; SEP, somatosensory evoked potential.

The blood supply of the spinal cord is important

Somatosensory Evoked Potentials to consider as vascular injuries can selectively affect
SEPs are obtained after electrical stimulation of certain spinal cord tracts. A single anterior spinal
peripheral nerves. The SEP waveforms are recorded artery (ASA) and a pair of posterior spinal arteries
along various points of the neuroaxis as the poten- (PSAs) supply the spinal cord. The ASA arises from
tial ascends up the dorsal column pathway. Details branches of both vertebral arteries. It descends the
about SEPs acquired in the clinical laboratory are length of the spinal cord from the cervicomedullary
presented in Chapter 4. In this chapter, the ways junction to the filum terminale. It receives radicu-
in which SEP acquisition and analysis differ when lar arteries from the aorta at various levels; two to
performed for monitoring neural function in the three such arteries supply the ASA at the cervical
level and one to three at the thoracic level. A large
OR will be presented.
radicular artery, the artery of Adamkiewicz, joins
the ASA between T9 and T12 levels and supplies
Anatomy and Physiology a large part of the caudal spinal cord. The ASA
SEP monitoring is performed by stimulating large supplies the anterior two thirds of the spinal cord,
peripheral nerves with electrical stimuli. Median or which includes the corticospinal tracts.
ulnar nerve stimulation is used in the upper limbs, A paired set of PSAs also arise from the vertebral
and tibial or peroneal nerve stimulation is used in arteries and descend the length of the spinal cord.
the lower limbs. SEPs are thought to be transmitted There is an extensive anastomotic network between
by the dorsal column pathway. The anatomy of this the two PSAs, and injury to one can be compensated
pathway is discussed in Chapter 4. by the other artery. Radicular arteries also join the
As the SEP signal ascends in the peripheral PSAs at various levels. Together, the PSAs supply
nerves, spinal cord, and brain, various waveforms the posterior one third of the spinal cord.
are recorded. The waveforms recorded after upper
and lower limb stimulation have a lot of similarities
Methodology
and are generated by comparable neural structures. The technique for obtaining SEP in the OR is similar
Neural generators of the various SEP waveforms are to that used in the clinical laboratory, but there are
presented in Table 5.4. The waveforms g enerated a few important differences. In this section, these
after median and ulnar nerve stimulation are the differences will be highlighted.
same and have the same latencies. SEP waveforms Stimulation techniques
generated after tibial nerve stimulation are com- Electrodes used in the OR for stimulation are often
parable to those after median nerve stimulation, different from those used in the clinical laboratory.
but with a longer latency. Peroneal nerve SEPs are Subdermal needle electrodes are used most often
similar to those of the tibial nerve, except that their for monitoring in the OR. These electrodes can
latency is about 10 ms shorter. The cortical P37 be quickly applied and require lower stimulation
waveform after tibial nerve stimulation can project intensity since they are closer to the nerves. They
in different directions, so recording it reliably in the can be secured easily and, if dislodged, can be more
OR requires using several derivations. easily reapplied than surface electrodes. However,
Table 5.4 Upper and Lower Limb SEP Waveform Neural Generators
Upper Extremity Lower Extremity
EP Brachial plexus PF Peripheral nerve
N13 Cervical spinal cord LP Cauda equina, lower spinal cord
P13/14 Nucleus cuneatus P31 Nucleus gracilis
N18 Thalamus N34 Thalamus
N20 Somatosensory cortex P37 Somatosensory cortex
EP, evoked potential; LP, lumbar potential; PF, popliteal fossa; SEP, somatosensory evoked potential.
the patient must be anesthetized before these are stimulate the nerves. If surface electrodes are used,
applied as insertion can be painful. Surface elec- a higher stimulus intensity is needed.
trodes can be applied when the patient is awake, The recommended stimulation rate is 2 to 8 per
but it takes much longer and over time their imped- s. In the upper limbs 5.7/s is a common stimula-
ance becomes unacceptably high. tion rate, while in the lower extremities, a slower
The nerves used most often for stimulation in the rate of 2.1/s is often used (7). If the waveforms are
upper limb are the median and ulnar. The median not clearly reproducible, slowing the rate to around
nerve is larger than the ulnar nerve and easier to 1.3/s can improve the morphology of the SEP.
stimulate. Because it innervates a large part of the The stimulation rate should not be an integer of 60
palm of the hand, its cortical representation is larger (50 in many parts of the world outside the United
than other nerves of the upper limb. The median States) as that is the electrical line frequency and
nerve enters the spinal cord through C6-T1 nerve will make eliminating electrical noise very difficult.
roots. Median nerve SEPs are used when surgery The electrode impedance should be kept below
involves the upper cervical spine (above the level 5,000 ohms (7).
of C6), as a control for thoracic and lumbar spinal
surgery, and for brain mapping when the somato- Recording techniques
sensory cortex needs to be identified during cranial Subdermal needle electrodes are also used for
surgery. The ulnar nerve is used most often as a recording. They share many of the advantages and
control when surgery is performed on the thoracic disadvantages mentioned previously. Corkscrew
or lumbar spine. It traverses the lower trunk of the electrodes are also occasionally used for recording
brachial plexus and enters the spinal cord through SEP, as they are easy to apply and very stable. These
C8-T1 nerve roots. When patients are positioned electrodes must also be applied after the patient has
with their arms above their head, stretching of the been anesthetized. Needle electrodes should be
lower trunk of the brachial plexus can produce used with caution in surgeries in which anticoag-
C8-T1 injury. Ulnar nerve SEP monitoring can help ulants are used as bleeding can occur at insertion
prevent this complication. sites. Surface electrodes can also be used for record-
In the lower limb, tibial nerve SEPs are mon- ing, and the advantages and disadvantages of these
itored most often. The tibial nerve enters the spi- electrodes were discussed earlier.
nal cord through L4-S2 nerve roots. This provides Recording montages used in the OR are similar
monitoring of almost the entire length of the spinal to those used in the clinical laboratory with a few
cord. Stimulation is performed behind the medial changes made to maximize the signal-to-noise ratio.
malleolus. In patients with a severe peripheral neu- This allows quicker acquisition of data and more
ropathy, the most distal nerves are affected most, timely feedback to the surgical team. The record-
and it may not be possible to stimulate the tibial ing montage is discussed in detail in Chapter 4.
nerve. In these situations (and in cases of below A recommended montage adapted from American
the knee amputation), stimulation of the peroneal Clinical Neurophysiology Society (ACNS) guide-
nerve at fibular head can be used to obtain peroneal lines is presented in Table 5.5 (12,16).
nerve SEP. The peroneal nerve enters the spinal Peripheral responses, the Erb’s point (EP) and
cord through L4-S1 nerve roots. Peroneal nerve SEP popliteal fossa (PF) waveforms, are very important
will have shorter latencies than tibial nerve SEP as to record in all cases in which SEP m onitoring
their site of stimulation is more proximal than tibial is used. Presence of these waveforms allows
nerve SEP. confirmation that the nerve has been adequately
Constant current stimulation is recommended stimulated, and if there is a loss of waveform, it is
for obtaining SEPs in the OR as it compensates for due to pathology that is occurring more proximally
changes in tissue impedance (7). A rectangular pulse than the site of the peripheral response generation.
of 100 to 300 μs is used with a stimulus intensity If these responses are not present, it becomes diffi-
of 30 to 40 mA. If possible, a visible twitch should cult to determine whether absent responses are due
be sought to ensure adequate stimulation but this to inadequate stimulation or surgical pathology.
may not be possible if the patient has already been The N13 and the lumbar potential (LP) waveforms
draped. In some patients, higher stimulus dura- are often not recorded in the OR. This is because
tion and intensity may be needed to adequately active electrodes required to record these responses
Table 5.5 Typical Montage Used for Upper and Lower Extremity SEP Monitoring
Upper Extremity Lower Extremity
Montage Waveform Montage Waveform
CPc–CPi (CP3/CP4) N20 CPz–Fz P37

Alternate: CPc–Fz
CPi–EPc P13/14, N18 CPi–CPc (CP3/CP4) P37
Alternate: C1’–C2’
Fz–EPc P13/14, N18 Fz–C5S P31, N34
Alternate: C5S–Fz
C5S–EPc N13 T12–IC LP
EPi–EPc EP PFd–PFp PF
EP, evoked potential; LP, lumbar potential; PF, popliteal fossa; SEP, somatosensory evoked potential.
are often in the surgical field. This prevents their Often, multiple channels are used to record the P37
application. waveform to ensure that the best P37 waveform is
Subcortical waveforms (P13/14, N18 and P31, recorded.
N34) are recorded from the brainstem, above SEP recording parameters in the OR have been
the surgical site in cases of spine surgery. Thus, recommended by the ACNS (7). Typically 500 to
changes in subcortical waveforms can indicate neu- 1,000 repetitions are needed to obtain reliable wave-
ral injury at the site of surgery. These responses are forms. However, if the signal-to-noise ratio is large,
also relatively insensitive to anesthetics and are not fewer responses may be needed. Filter setting of 30
affected much even when inhalational agents are to 3,000 Hz is common. A time base of 30 to 100
used. However, subcortical responses are of low ms for upper extremities and 75 to 100 ms for lower
amplitude, can be difficult to record when there is extremities is appropriate. If responses are very
a lot of background noise, and require many rep- delayed, as in patients with demyelinating neurop-
etitions (and more time) to reproduce. Because of athies, the time base may need to be increased.
these latter issues, these responses are sometimes
not recorded in favor of higher amplitude cortical
waveforms. Anesthetic Considerations
Cortical waveforms (N20, P37) are high- Anesthetic agents can have a significant effect
amplitude waveforms that are usually easy to on SEP. This effect is made more complicated by
record. However, they are affected by inhalational different waveforms being affected differently by
anesthetics, and to record them reliably, often intra- various agents. A summary of the effects of various
venous anesthetics are needed. The N20 waveform agents on the different SEP waveforms is presented
can be reliably recorded from a centroparietal con- in Table 5.6 (12). For a more detailed discussion, the
tralateral (CPc) to CP ipsilateral (CPi) derivation. reader is referred to other references (17).
The best recording derivation for the P37 waveform
is variable. The derivation used most commonly is
the CPi–CPc, but this is because those electrodes Warning Criteria
are already present for upper limb SEP record- The warning criteria for SEP are an amplitude drop
ing. However, at times, the P37 waveform vector of 50% or latency increase of 10%. An amplitude
projects ipsilaterally or to the midline. In these decrement is more important than latency prolon-
situations, the best montage may be CPc–CPi or a gation for SEP. This criterion is, however, empiric.
Cz–Fpz (vertex to frontopolar midline) derivation. As noted for BAEP, if the SEP has been very stable,
Table 5.6 Summary of Effects of Anesthetics on SEP

Cortical waveforms Subcortical waveforms Peripheral waveforms
Anesthetic (N20; P37) (P13/14, N18; P31, N34) (EP; PF)
Halogenated gases ↓ Latency Amplitude Minimal change No change

↑ Latency
Nitrous oxide ↓ Amplitude Minimal change No change
– Latency
Propofol Minimal change Minimal change No change
Opioids Minimal change Minimal change No change
Benzodiazepines Minimal change Minimal change No change
Muscle relaxants ↑ Amplitude ↑ Amplitude No change
– Latency – Latency
EP, evoked potential; PF, popliteal fossa; SEP, somatosensory evoked potential.
↓ = decrease; ↑ = increase; – = no change.
an amplitude decrease of less than 50% may be occur at many different levels. Surgeries can cause
clinically meaningful and should generate an alert. trauma, stretch, or thermal injury to the spinal
However, if the SEP waveforms have been variable cord, brain, or peripheral nerves. Manipulation of
and amplitude fluctuating, even a 50% decrease blood vessels can result in spinal cord infarction as
may not be meaningful. Thus, interpretation of the well. A summary of the SEP changes that can occur
change and the warning criteria are highly depen- during surgery is presented in Table 5.7.
dent on the individual surgery and reliability of
baseline responses. Physiologic changes
A misnomer about “warning criteria” is that an Changes in the physiologic milieu can affect
alert is only generated when the alert criteria are the SEPs. These changes affect the entire body,
met. In fact, if there is a gradual decrease of an SEP so the SEP changes are typically bilateral and
waveform amplitude, the surgeon should be alerted symmetric. However, often the lower limb (tibial
to the progressive nature of the change, even before nerve) SEPs are affected more than the upper limb
it reaches the 50% cutoff. This can allow the surgeon (ulnar nerve) SEPs. Also, one side may occasionally
to assess what is occurring in the surgical field and be affected more if there is a preexisting lesion of
make changes if appropriate before critical damage that side.
occurs. A reduction in temperature can have a pro-
found effect on SEPs. Peripheral nerve conduction
Mechanisms of Waveform Change becomes slower and amplitude increases initially
Changes can occur in the SEP waveforms due to sur- and then falls. Similarly, subcortical and cortical
gical damage to the dorsal column pathways, changes waveforms also increase in latency and drop in
in homeostatic milieu, or from technical problems. amplitude. The N20 waveform disappears by about
The changes may look similar, and it is important to 21°C, and subcortical and cervical waveforms
differentiate between them so that appropriate feed- disappear at 17°C. With rewarming, the waveforms
back can be provided to the surgical team. reappear in the reverse order (18). Blood pressure
changes can affect SEPs as well. A drop in blood
Surgery induced changes pressure results in reduction of amplitude of the
SEP monitoring can detect many types of injury SEP waveforms. Anesthesia also affects SEP wave-
to the dorsal column pathway. These injuries can forms, as discussed earlier.
Technical issues useful in reducing neurologic morbidity in scoliosis

Many technical issues can cause a change in the SEP. and other types of spinal surgery (20). SEP moni-
These must be quickly recognized and differentiated toring, in conjunction with other types of monitor-
from surgery induced changes. Technical issues can ing (so-called multimodality monitoring), is used in
be thought of as those affecting the stimulating or many types of spinal surgery, from the cervical to
recording system. Common technical reasons that the lumbosacral spine.
result in SEP changes are listed in Table 5.8. Surgery on the spinal cord can also cause injury
to the dorsal column tracts. SEP monitoring is often
used in these surgeries, often in conjunction with
Clinical Applications MEP monitoring. Brainstem and brain surgery
SEP monitoring has been used for many decades for often employ SEP monitoring as well, and it can
scoliosis surgery, and there is evidence that moni- help prevent injury to the dorsal column pathways.
toring reduces morbidity (19). An evidence-based SEP monitoring has been used in tethered cord
guideline confirmed that SEP monitoring was syndrome surgeries. The value of SEP monitoring in
Table 5.7 Patterns of Surgery-Induced Changes in SEP

Location Cause Effect on SEP
Thoracolumbar spinal cord Trauma, stretch, ischemia Ulnar nerve SEP: Preserved EP, P13/14, N18, N20 waveforms
Tibial nerve SEP: Preserved PF; loss of P31, N34, P37 waveforms
Cervical spinal cord (above C7) Trauma, stretch Ulnar nerve SEP: Preserved EP; loss of P13/14, N18, N20
waveforms
Tibial nerve SEP: Preserved PF; loss of P31, N34, P37 waveforms
Brainstem (above lower medulla Trauma, ischemia, thermal Ulnar nerve SEP: Preserved EP, P13/14; loss of N18, N20
and below thalamus) waveforms
Tibial nerve SEP: Preserved PF, P31; loss of N34, P37 waveforms
Cortical and subcortical white Trauma, ischemia, thermal Ulnar nerve SEP: Preserved EP, P13/14, N18; loss of N20
matter waveforms
Tibial nerve SEP: Preserved PF, P31, N34; loss of P37 waveforms
Lower limb Ischemia, trauma Ulnar nerve SEP: Preserved EP, P13/14, N18, N20 waveforms
Tibial nerve SEP: Loss of PF, P31, N34, P37 waveforms
Upper limb (including lower trunk Ischemia, stretch, trauma Ulnar nerve SEP: Loss of EP, P13/14, N18, N20 waveforms
of brachial plexus) Tibial nerve SEP: Preserved PF, P31, N34, P37 waveforms
EP, evoked potential; PF, popliteal fossa; SEP, somatosensory evoked potential.
Table 5.8 Technical Issues Causing Changes in SEPs

Dislodgement of stimulating electrodes Dislodgement of recording electrodes
Edema around stimulation site Edema around recording site
High electrode impedance High electrode impedance
Machine hardware problems Change of filter settings; addition of 60 Hz (notch) filter
Amplifier software and hardware problems
Surgical equipment (electrocautery, others)

these surgeries is uncertain as injury to individual from a strip electrode placed perpendicular to the
roots may not be detected by SEP monitoring (21). central sulcus, a “phase reversal” is seen. This is not
There seems to be more value to this monitoring in a true phase reversal as the negative and positive
complicated and redo surgeries. potentials are not opposite ends of the same dipole;
Many other type of surgeries on peripheral rather they are different dipoles of opposing polar-
nerves, spinal cord, brainstem, or brain in which ities. The central sulcus lies between the N20 and
dorsal column (somatosensory) pathways are at P22 waveforms.
risk can employ SEP monitoring. Examples of such
surgeries include aortic (endovascular and open),
cerebrovascular (endovascular and open), periph- Examples
eral nerve (such as brachial plexus), and movement Several examples of SEP monitoring are provided
disorder surgeries. This list is ever expanding. (Figures 5.5–5.8, Examples 5–8). SEPs are shown in
Median nerve SEP can also be used for brain sequence and important surgical events are noted.
mapping. SEPs are recorded directly from the cortex An accompanying narrative provides a brief his-
with a strip electrode that has four to eight contacts. tory and a discussion of the finds. Neuroimaging
The contacts on the strip electrode are referenced to findings that highlight the case are also presented
a distant electrode (such as the contralateral ear). An when available. It should be noted that despite the
N20 waveform is recorded from the somatosensory SEPs being shown in isolation in the examples,
cortex and a P22 waveform is recorded from the they are often performed along with other types of
motor cortex. When these waveforms are recorded monitoring.
Age: 18 years
Sex: Male
Diagnosis: Kyphoscoliosis
Procedure: Posterior spinal fusion and instrumentation
Modalities monitored: Bilateral ulnar and tibial nerve SEPs; bilateral upper and lower limb MEPs
(A) Median nerve SEPs intraoperative monitoring

Montage:
Left-sided stimulation: CP4–CP3, CP3–REP, LEP–REP
Right-sided stimulation: CP3–CP4, CP4–REP, REP–LEP
(B) Tibial nerve SEPs intraoperative monitoring

Montage:
Left-sided stimulation: CPz–Fpz, CPz–CP3, CPz–C5S, Fpz–C5S, LPFd–LPFp
Right–sided stimulation: CPz–Fpz, CPz–CP4, CPz–C5S, Fpz–C5S, LPFd–LPFp

(C)
History: This is an 18-year-old male with kyphoscoliosis of the thoracolumbar spine. He is presenting for posterior spinal fusion and
instrumentation to correct his spinal deformity.
Discussion: Ulnar and tibial nerve SEP and upper and lower limb MEP monitoring are requested for this surgery. The SEP monitoring
is shown here. The ulnar (A) and tibial (B) nerve SEPs did not show significant changes during the surgery; the critical part of the
surgery is shown. The ulnar nerve SEP montage has derivations that show a cortical (N20) waveform in the first column, subcortical
(P13/13, N18) waveforms in the second column, and a peripheral (EP) waveform in the third c olumn. The tibial nerve SEP montage
has two derivations that show the cortical (P37) waveform in the first two columns, subcortical (P31, N34) waveforms in the third
and fourth columns, and the peripheral (PF) waveform in the fifth column. The third column also shows the cortical (P37) waveform.
The PF waveform is not well formed due to excessive stimulation artifact.
A spine X-ray shows the kyphoscoliotic deformity (C).
Tibial nerve SEP monitoring has been used for decades to detect injury to the spinal cord during spinal instrumentation surgeries.
Even though they assess the sensory system (dorsal column pathways), tibial nerve SEPs serve as surrogate markers for motor system
injury. In spinal instrumentation surgery, injury to the spinal cord is usually due to manipulation of the vertebral column, and it was
assumed that when motor tract injury occurred, dorsal column injury would also occur and would be detected by tibial nerve SEP
monitoring. However, over the last few decades, it has been shown that tibial nerve monitoring cannot detect all cases of motor
system injury. Regardless, tibial nerve SEPs remain a vital part of monitoring as they provide complementary information to moni-
toring with MEPs.
Ulnar nerve SEPs were monitored to serve as a control for systemic issues that might affect the evoked potentials. Since the spinal
deformity and surgery are of the thoracolumbar spine, the ulnar nerve SEP will not be affected from surgical manipulation. Blood
pressure or anesthetic effects will be manifest on the ulnar and tibial nerve SEPs. The ulnar nerve SEP can also detect an impending
brachial plexopathy due to stretch injury that might occur from patient positioning.
In this surgery, no changes were noted with ulnar or tibial nerve SEP monitoring. No irreversible injury to the ulnar and tibial nerve
dorsal column pathways occurred during this procedure.
EP, evoked potential; MEP, motor evoked potential; PF, popliteal fossa; SEP, somatosensory evoked potential.

Age: 10 years
Sex: Female
Montage:
Left-sided stimulation: Cz–Fpz, CP3–CP4, CPz–CP3, CPz–C5S, Fpz–C5S, LPFd–LPFp
Right-sided stimulation: Cz–Fpz, CP4–CP3, CPz–CP4, CPz–C5S, Fpz–C5S, RPFd–RPFp
Diagnosis: Idiopathic scoliosis
Modalities monitored: Tibial nerve SEPs
(A)
(B)

(C)
(D)
(D)

History: This is a 10-year-old female with a history of congenital heart defects who presents for evaluation of thoracic scoliosis. She
is deemed a good candidate for posterior spinal fusion and instrumentation.
Discussion: Only tibial nerve SEP monitoring is requested for this surgery. At the start of surgery (A), good tibial nerve SEP wave-
forms are seen, with baseline traces at the bottom of the panel. During instrumentation (B), there is a sudden greater than 50%
decrease of the right tibial nerve cortical and subcortical waveforms. The surgeon is alerted, but improvement does not occur until
the end of surgery (C).
A spine X-ray shows the significant scoliosis (D).
Even though MEP monitoring is commonly used for scoliosis surgeries, some surgeons still only request SEP monitoring. In this
monitoring, six derivations are used to monitor tibial nerve SEPs. The first four columns include derivations that assess the
cortical (P37) waveform. The fourth and fifth columns have derivations that include the subcortical (P31 and N34) waveforms.
The sixth derivation has the peripheral (PF) waveform. All waveforms on both sides are very robust at the start of the proce-
dure. At 11:58:29, there is a sudden decrease of greater than 50% of all cortical and subcortical waveforms obtained after
right-sided stimulation. The PF waveform remains consistent, and left-sided responses do not change. This implies a possible
injury to the right-sided dorsal column pathway during instrumentation. This change did not improve until the end of surgery.
The notes of the surgery imply that the surgeon was alerted at 12:13:01, which would have been about 15 minutes after the
change occurred. In fact, the surgeon was notified promptly after the change, but notation in the record was made later. This is
problematic medicolegally, and the notation should be made immediately when the change is noted and surgeon alerted. Had
MEP monitoring also been done, the change might have been detected earlier and an intervention could have occurred sooner.
Tibial nerve SEP monitoring during this surgery showed possible injury to the dorsal column pathway after right tibial nerve stimu-
lation that did not recover by the end of surgery. This may be correlated with postoperative sensory loss and possibly weakness of
the right lower extremity.
MEP, motor evoked potential; PF, popliteal fossa; SEP, somatosensory evoked potential.

Age: 75 years
Sex: Male
Montage:
Left-sided stimulation: Fz–C5S, CPz–C5S, CPz–Fpz
Right-sided stimulation: Fz–C5S, CPz–C5S, CPz–Fpz
Diagnosis: Lumbar spinal stenosis and multilevel disc herniation
Procedure: Multilevel lumbar discectomy and transforaminal lumbar interbody fusion
Modalities monitored: Tibial nerve SEPs, bilateral lower limb EMG, and anal sphincter EMG
History: This is a 75-year-old male who presents with a 2-year history of lower back pain and fatigue. Walking short distances
results in increased pain and weakness of the lower extremities. Examination shows mild bilateral lower extremity weakness and
decreased reflexes. An MRI of the lumbosacral spine shows several herniated discs and spinal stenosis. The patient is scheduled to
undergo a posterior lumbar interbody fusion of multiple lower lumbar spinal segments.
Discussion: Tibial nerve SEP and bilateral lower extremity and anal sphincter EMG monitoring is requested for this surgery. Tibial
nerve SEP monitoring for the first hour of surgery is shown. The first three columns are the SEPs obtained after left tibial nerve
stimulation, and the second three columns are of the right tibial nerve SEPs. Cortical (P37) and subcortical waveforms are noted at
the start of surgery, but shortly thereafter they decrease in amplitude. The mean arterial pressure (MAP) is also noted to drop from
85 mmHg to 65 mmHg. The surgeon and anesthesiologist are alerted, and the MAP is increased to 90 mmHg. This results in SEPs
that are comparable to baseline. No further changes are noted during the surgery.
The modalities monitored during lower spinal surgery are highly variable and dependent largely on the surgeon’s preference. In this
case, tibial nerve SEP monitoring was requested, along with EMG monitoring. Three derivations are shown for each tibial nerve; the first
column shows the subcortical (P31, N34) waveforms and the third column also shows the cortical (P37) waveform. The second column
shows both. Between where the notes mention “prepping” and “normalized pressure,” the MAP decreases as noted earlier. This results
in a decrease in amplitudes of the P31, N34, and P37 waveforms of both sides, and responses obtained with left-sided stimulation were
affected more than the ones from the right. When the MAP is increased to 90 mmHg (denoted by “normalized pressure”), these wave-
forms increase in amplitude. Usually, blood pressure and other systemic changes affect the SEP symmetrically. However, at times, the
change may be slightly asymmetric as it was here. This may be due to preexisting neurologic injury on the side that has more changes.
The changes seen in this SEP monitoring were not permanent and easily reversed with increasing the MAP. There was no irreversible
injury to the dorsal column pathways of the tibial nerves during this procedure.
EMG, electromyographic; SEP, somatosensory evoked potential.
Figure 5.7 Example 7.

Age: 9 years
Sex: Female
Montage: Strip electrode contact 1–A1 (ref), 2–A1 (ref), 3–A1 (ref), 4–A1 (ref), REP–LEP
Diagnosis: Right frontal lobe brain tumor
Procedure: Right craniotomy for resection of brain tumor
Modalities monitored: Left median nerve SEPs for mapping; median nerve SEPs; direct cortical MEPs
(A)

(B) (C)
History: This is a 9-year-old female who had a seizure about 3 months ago. Thereafter, she complained of persistent headaches. An
MRI showed a right frontal lobe lesion. She presents for a craniotomy for resection of the lesion.
Discussion: Sensory and motor mapping of the cortex followed by SEP and MEP monitoring is requested. Left median nerve SEPs
are used for mapping of the central sulcus (A). A phase reversal is seen between contacts 3 and 4 (lighter waveforms). Previously
obtained SEP showed a phase reversal between contacts 2 and 3 (darker waveforms).
The brain MRI shows a mass in the right frontal region. An axial T2 weighted image (B) shows the high-intensity lesion with a cystic
area. A contrast-enhanced axial image (C) shows that the mass enhances as well.
Median nerve SEPs are used not only for monitoring the dorsal column pathways during surgery, but also for mapping the central
sulcus in cases of surgery near the sensorimotor cortex. It is difficult to identify the central sulcus anatomically, and median nerve
SEP phase reversal can help localize this landmark physiologically. In this example, the left median nerve is stimulated and recordings
obtained from a 4-contact strip electrode placed on the cortex by the surgeon. Initially, the darker colored waveforms are obtained,
showing an N20 waveform from the somatosensory cortex in channels (contacts) 3 and 4. The P22 waveform from the motor cortex
is seen in channels (contacts) 2 and 1. The “phase reversal” in this case is between channels (contacts) 2 and 3. However, the highest
amplitude N20 and P22 waveforms are not adjacent to each other. The N20 waveform is of higher amplitude in channel (contact)
4 than in channel (contact) 3, and the highest amplitude P22 waveform is in channel (contact) 2. This implies that contact 3 may be
lying on the central sulcus itself. The strip electrode is then moved by the surgeon, and the median nerve SEP acquired again. This
is shown as the light-colored waveforms. Now the N20 waveform is best seen in c hannel (contact) 4, and the P22 waveform is best
seen in channels (contacts) 3, 2, and 1. The highest amplitude N20 and P22 waveforms are adjacent to each other now, making the
“phase reversal” more reliable. The “phase reversal” is between contacts 3 and 4, implying that the central sulcus is between these
two electrodes. This monitoring is followed by motor mapping, which confirms the location of the motor cortex.
In this example, a “phase reversal” was best seen between contacts 3 and 4 (light waveforms). At a previous location, the “phase
reversal” was seen between contacts 2 and 4, with contact 3 possibly lying on the central sulcus. The strip electrode is moved by
the surgeon to identify the best “phase reversal.” Median nerve SEP and direct cortical MEPs were obtained during tumor resection
(not shown here).
MEP, motor evoked potential; SEP, somatosensory evoked potential.

Motor Evoked Potentials Pyramidal

neurons
MEPs performed in the OR are obtained after
high-intensity transcranial electrical stimulation.
This type of stimulation is very painful and can- Cortical
not be performed in awake individuals in the clin- interneurons
ical laboratory. Magnetic MEPs are painless and

can be performed in awake individuals, but they Internal Internal
are very sensitive to anesthetic and the equip- capsule capsule
ment required to produce them is very bulky; for

Midbrain Midbrain
these reasons, magnetic MEPs are not performed
in the OR. Electrically induced MEPs are the only
Pons Pons
evoked potential routinely done in the OR that is
not performed in the clinical laboratory. It will be
Medulla Medulla
discussed briefly here for completeness. For more
information, the reader is referred to more compre-
hensive texts on this subject (2–5). In this chapter,
the term MEP implies electrically induced MEP.
Spinal Spinal
In many cases, MEP and SEP are performed cord cord
together as they provide complementary informa-
tion. Although SEP monitors the dorsal column
pathway, MEP assesses the corticospinal tract. As
such, when performed together, they provide com- Muscle
plementary information.
Anatomy and Physiology Ipsilateral Contralateral

MEP travels in the corticospinal tract of the brain
and spinal cord. This pathway starts with the
Figure 5.9 Diagrammatic representation of the corticospinal
pyramidal neurons. Axons from these cells form the
pathway arising from the cortex.
corticospinal tract. The corticospinal tract descends
in the subcortical white matter and travels in the Depolarization of the pyramidal neurons pro-
anterior two-thirds of the posterior limb of the duces a descending corticospinal tract volley. This
internal capsule. It descends further through the descending volley can be recorded from the spinal
midbrain and pons to the medulla. A majority of cord with an epidural electrode. A large initial wave
the corticospinal tract decussates to the contralat- is seen, known as the “D wave” (direct wave) and is
eral side in the medulla and descends as the lateral thought to arise from direct activation of the pyra-
corticospinal tract in the lateral aspect of the spinal midal neurons. Following the D wave are multiple,
cord. The part of the corticospinal tract that does smaller “I waves” (indirect waves). The I waves are
not decussate descends as the much smaller ante- thought to arise due to interneuron reactivation
rior corticospinal tract in the anterior aspect of the of the pyramidal neurons. The D wave arrives at
spinal cord. The corticospinal tract ends by synaps- the anterior horn cells and results in their depolar-
ing at anterior horn cells at each spinal level. From ization. However, a single D wave cannot activate
the anterior horn cells arise motor nerve fibers that enough anterior horn cells to produce a muscle
travel in the anterior horn of the spinal cord, anterior contraction, or “M wave.” The M wave is recorded
spinal root, and finally to peripheral motor nerves. with an electrode in the activated muscles. The M
Motor nerves end at the neuromuscular junction wave is often referred to as the “MEP.”
and activate skeletal muscle fibers (Figure 5.9). Because a single electrical stimulus cannot acti-
A single, high-intensity electrical stimulus causes vate enough pyramidal neurons and anterior horn
depolarization of pyramidal neurons and nearby cells to activate an M wave, a series of electrical stim-
interneurons and activates the motor pathway. uli in quick succession is used. A train of electrical
stimuli produces multiple D waves, which produce 2 ms allows maximum summation for MEP in the
activation of more anterior horn cells by tempo- upper limbs, and an ISI of 3 to 4 ms is best for lower
ral summation. With a greater number of anterior extremity MEPs. It is best to start with an ISI of
horn cells activated, a M wave can be recorded. 2.1 ms and then adjust it, depending on the MEP
Increasing the intensity of stimulation increases waveforms obtained.
the amplitude of the D wave, and increasing the Stimulation for MEP is done with the anode.
number of pulses in the train of stimuli increases Anodal stimulation is more effective than cathodal
the temporal summation on the anterior horn cells. stimulation for activating the pyramidal neurons.
Both of these techniques help increase the ampli- Cathodal stimulation is more effective in activat-
tude of the M wave recorded from muscles. ing white matter fiber tracts. Needle or corkscrew
It is important to consider the blood supply of the electrodes are used to deliver effective stimulation.
spinal cord when discussing MEPs. The spinal cord Surface electrodes are not used for MEP as they
blood supply by the ASA and PSAs was discussed require much greater intensity stimulation to pro-
earlier in this chapter. The ASA is responsible for duce MEPs comparable to needle and corkscrew
supplying the anterior two thirds of the spinal cord, electrodes.
which contains both the lateral and anterior corti- MEP stimulating electrodes are placed in C1/2
cospinal tract. The PSAs supply the posterior one and C3/4 locations (per International 10/20 elec-
third of the spinal cord, which contains the dorsal trode placement system) (22). The C1/2 location
column tract. Thus, ischemia in the ASA territory is closer to the midline and best for eliciting MEPs
is more likely to affect MEP monitoring than PSA from the lower extremities. The C3/4 location is
ischemia. closer to the hand area and is better for eliciting
MEP in the upper limbs. A slightly more anterior
location, known as M1/2 and M3/4, has been advo-
Methodology cated as being more effective in eliciting MEPs (23).
There are many methods of performing MEPs.
Different stimulation techniques can be used to Recording techniques
enhance MEP, and recordings can be made from MEP recordings are made from muscles. Typically
the spinal cord, nerves, and muscles. In this review, needle electrodes are used to generate a large,
the most common methodology used for MEP polyphasic MEP. In lean patients, a subdermal
monitoring will be discussed. needle may be sufficient to get the electrode into
muscle mass. For obese individuals, longer needle
Stimulation techniques electrodes or wire electrodes are used. The record-
Transcranial electrical stimulation requires a device ing electrodes are inserted into multiple muscle
that is capable of an output of 1,000 V. Most neu- groups depending on site of surgery and risk for
rophysiologic intraoperative monitoring (NIOM) injury. In almost all cases, a sampling of upper and
machines have the capability of providing such lower extremity muscles is important. If the surgery
high-intensity electrical output. Many machines is on the thoracolumbar spine, the upper limb MEP
also provide a return current in milliamperes, serves as a control. If the surgery is on the upper cer-
which is a better measure of the actual amount of vical spinal cord, upper and lower limb MEP can be
stimulation provided to the patient. The duration affected by the surgery. A sample upper and lower
of each stimulus can be adjusted to either 50 μs or limb MEP montage is presented in Table 5.9.
75 μs. Some machines offer a long duration stim-
ulus of 500 μs. If a duration of 500 μs is used, the
maximum intensity that is allowed by the machine Anesthetic Considerations
is much lower. Most anesthetic agents affect MEPs. Inhalational
Because of the need for temporal summation, agents have the most effect on MEPs, and they often
machines used for MEP monitoring allow delivery make MEP monitoring difficult. When MEP monitor-
of three to nine pulses in a row. Not only can the ing is performed on neurologically normal individu-
duration and intensity of the impulses in the pulse als, 60% will have reliable MEP when 0.6% isoflurane
be set, their interstimulus interval (ISI) can also is used, and this number drops to 20% when 0.8% iso-
be adjusted between 2 and 8 ms. An ISI of about flurane is used (24). When patients are neurologically
Table 5.9 Typical Muscle MEP Recording Montage important for the NIOM team to monitor the TOF
testing and to test it in the muscle group of most
Upper Limb Lower Limb
interest; that is, if foot MEP is of most interest, the
APB–ADM AH–ADQ TOF should be tested in that muscle with activation
FCU–FCR AH–TA of the tibial nerve.
TA–TA
TA–MG Warning Criteria
VM–VL The amount of change in MEPs that is suggestive
of injury to the corticospinal pathway is not well
AH, abductor hallucis; ADM, abductor digiti minimi; ADQ, abductor established. This stems from the inherent variabil-
digiti quinti; APB, abductor pollicis brevis; FCR, flexor carpi radialis;
FCU, flexor carpi ulnaris; MEP, motor evoked potential; MG, medial
ity in the MEPs. With each transcranial stimulation,
gastrocnemius; TA, tibialis anterior; VL, vastus lateralis; VM, vastus a different population of anterior horn cells is acti-
medialis. vated. As a result, a slightly different population of
Source: Ref. (12). Husain AM. Neurophysiologic intraop- motor units is activated, which causes variability in
erative monitoring. In: Ebersole JS, Husain AM, Nordli the MEP. Consequently, there is disagreement as to
DR, eds. Current Practice of Clinical Electroencephalography. the warning criteria.
Philadelphia, PA: Wolters Kluwer Health; 2014: 488-542. The “all-or-none” rule is commonly used in inter-
preting MEP monitoring. This implies that an alert
compromised, inhalational agents can have an even is raised when the MEP amplitude decreases by at
more profound effect on MEPs. However, if needed, least 80% to 90% or disappears completely, and an
successful MEP monitoring can often be performed increase in stimulation intensity does not overcome
in neurologically normal patients with less than 0.5% this change (23,25). This approach compensates for
inhalational agent as part of balanced anesthesia. the variable nature of MEPs. Others have argued
Intravenous anesthetics are preferred when that this “all-or-none” approach is overly insensi-
MEP monitoring is being performed. This includes tive in detecting injury to the motor pathway. It has
agents like propofol. In doses up to 25 mg/kg/hr, been suggested that a 50% decrease in amplitude
propofol has no appreciable effect on the MEP. Even should prompt an alert. Others have suggested that
higher doses are compatible with effective MEP if the stimulation intensity needs to be increased
monitoring. At doses greater than 100 mg/kg/hr, by 100 V to maintain the amplitude of the MEP,
MEP can be affected. Compared to inhalational that should be considered significant and prompt
agents, propofol used at doses that cause a compa- an alert (26). Others have proposed more complex
rable amount of suppression of the bispectral index algorithms that incorporate the morphology of the
(BIS) causes a lot less suppression of MEPs (24). MEP waveform (27,28).
Propofol is often combined with narcotics like Because of its relative simplicity, the “all-or-
fentanyl to produce balanced anesthesia. This is none” approach is used by most practitioners. As
known as total intravenous anesthesia (TIVA), and with BAEP and SEP, this approach to interpretation
this often enables very effective MEP monitoring. must be tempered with the reliability of the MEP.
Neuromuscular blocking agents can eliminate If the MEP has been relatively invariant during the
MEPs. They are often used for intubation and expo- surgery, a 30% to 50% drop in amplitude may be
sure, and during these times, reliable MEP cannot significant. On the other hand, if the MEP amplitude
be obtained. However, neuromuscular blocking has been highly variable and intermittently present,
agents can also facilitate obtaining MEP in some an 80% to 90% decrement may not be meaningful.
situations. When there is excessive movement with
MEPs that is disruptive for surgery, using an infu-
sion of neuromuscular blocking agents can help Mechanisms of Waveform Change
reduce patient movement. In this situation, an infu- MEPs, like other types of evoked potentials, can
sion rather than bolus dosing of the neuromuscular change due to a variety of reasons. These changes
blocking agent must be used. Ideally, two of four may be induced by surgical factors, change in the
twitches must be present on train of four (TOF) homeostatic milieu, or technical problems with the
testing to maintain effective MEP monitoring. It is monitoring.
Surgery induced changes Technical issues

Surgical injury to the corticospinal tract can occur Technical issues that affect MEP can be due to
at many different levels. Often the etiology of dam- problems with the stimulating or recording system.
age is similar to dorsal column tract injury. In fact, These problems can occur at any time during the
both MEP and SEP are frequently affected together. surgery. Common technical reasons that result in
A summary of MEP changes that can occur at MEP changes are listed in Table 5.11.
various levels and their causes is presented in
Table 5.10.
Safety
Physiologic changes MEP monitoring involves delivery of trains of
Physiologic changes that affect the system milieu high-intensity electrical current repeatedly to the
can significantly affect MEPs. Usually these changes scalp. Concerns have been raised about poten-
affect both upper and lower extremity MEP, but tial complications of such high-intensity, recur-
often the lower extremity MEPs are affected more rent electrical stimulation. An analysis of over
than the uppers. As with SEP, if a preexisting lesion 50,000 cases of MEP monitoring revealed tongue
is present in the corticospinal tract, MEPs on that and lip lacerations to be the commonest injury,
side may be affected more by changes in the sys- occurring in less than 0.2% of cases. Other injuries
temic milieu. Hypotension can significantly affect such as seizures, cardiac arrhythmias, mandibular
the MEP waveform amplitudes. MEPs are affected fractures, and intraoperative awareness are much
more than SEPs by hypotension. Raising the MAP less common and occur in no greater frequency
often improves the MEPs. A drop in temperature than chance occurrence (29). Skin burns can occur
can also cause a decrease in MEP amplitudes. MEPs with very high-intensity stimulation, especially
are very sensitive to anesthetics, especially inhala- with pulse duration of 50 µs and intensity close
tional ones, as discussed earlier. to 1,500 V. In this scenario, 75 mJ are delivered
Table 5.10 Patterns of Surgery-Induced Changes in MEP
Location Cause Effect on MEP
Brain and subcortical Trauma, ischemia, thermal Upper limb: Selective loss of amplitude
white matter Lower limb: Selective loss of amplitude
Brainstem Trauma, ischemia, thermal Upper limb: Loss of amplitude

Lower limb: Loss of amplitude
Cervical spinal cord Trauma, stretch Upper limb: Selective loss of amplitude depending on level
Lower limb: Loss of amplitude
Throacolumbar Trauma, stretch, ischemia Upper limb: Preserved amplitude

spinal cord Lower limb: Loss of amplitude
Cauda equina Trauma Upper limb: Preserved amplitude

Lower limb: Selective loss of amplitude depending on level
Lower limb Trauma, ischemia Upper limb: Preserved amplitude

Lower limb: Selective loss of amplitude depending on segment
Upper limb Trauma, ischemia, stretch Upper limb: Selective loss of amplitude depending on segment
Lower limb: Preserved amplitude
MEP, motor evoked potential.

Table 5.11 Technical Issues Causing Changes in MEPs
Dislodgement of stimulating electrodes Dislodgement of recording electrodes
Edema around stimulation site High electrode impedance
High electrode impedance Amplifier software and hardware problems
Stimulator software or hardware problems Surgical equipment (electrocautery, others)
MEP, motor evoked potential.
to the scalp and can cause a burn (23). However, MEP monitoring has been used in many types
skin burns occur much more often with inadequate of surgeries on the brain and brainstem, such as
patient grounding. surgeries for tumors, masses, and vascular lesions.
Previously several conditions were thought to When a craniotomy is performed, MEP monitor-
be contraindications to MEP monitoring. These ing can become challenging as the stimulating
included history of epilepsy, pacemakers, vascu- electrodes may need to be displaced. The stimulat-
lar clips, deep brain stimulators, skull defects or ing electrodes are thus not in an optimal location
skull plates, and cochlear implants. Only a few of for activation of the motor cortex. The stimulation
these, such as cochlear implants and vascular clips, intensity also has to be reduced to prevent exces-
remain as contraindications to MEP monitoring. sive current delivery to the exposed cortex. Other
Most others require a discussion between the surgi- methods of stimulation, such as direct cortical stim-
cal and NIOM team to determine whether utility of ulation and subcortical stimulation, can be used to
MEP monitoring outweighs the risk. Appropriate activate the motor pathways in these types of sur-
adjustments, such as decreasing the intensity of
gery. Details about these types of MEP monitoring
stimulation, may need to be made. paradigms can be found elsewhere (3–5).
Endovascular cerebrovascular surgeries have
employed MEP monitoring as well. Inadvertent
Clinical Applications occlusion of blood vessels supplying the corticospi-
MEP monitoring is used for many types of surger- nal tract can be detected with MEP monitoring (15).
ies on the spine. It is used in conjunction with SEP Not only SEP, but also BAEP and electroencephalo-
monitoring to provide complementary information graphic (EEG) monitoring are often combined with
in monitoring. The MEP monitoring is for motor MEP monitoring in these cases.
pathways, while SEP monitoring is for sensory Open and endovascular aortic surgeries are often
pathways. Several studies have shown that SEP monitored with MEPs. Aortic surgery can cause
monitoring by itself cannot detect all instances of disruption of vascular supply to the spinal cord and
injury to motor pathways, and instance of false neg- ischemia of the anterior aspect of the spinal cord.
ative SEP monitoring are well documented (19,30). When such ischemia occurs, MEP changes occur
The value of MEP monitoring in reducing neuro- sooner than SEP changes (33). Use of multimodal-
logic morbidity during spine surgery has been doc- ity monitoring can improve outcomes of surgery on
umented in an evidence-based guideline (20). the descending aorta (34,35).
Spinal cord tumor surgeries are often monitored Many other types of surgeries employ MEP mon-
with MEPs. This is often combined with D wave itoring. These include brachial plexus surgery, teth-
monitoring, which can help the surgeon tailor ered cord syndrome, various joint replacements,
degree of tumor resection. A discussion of D wave and others. The value of MEP monitoring in these
monitoring in this situation is beyond the scope of surgeries remains to be defined.
this chapter, and the reader is referred to other ref- It is important to realize that MEP monitoring
erences (31,32). often results in significant patient movement. This
movement may be disruptive to the surgical team. the more sensitive it is in detecting potential injury
In this regard, MEP monitoring is different from to the corticospinal tract.
SEP and BAEP monitoring, which are “invisible” to
the surgeon. As such, before the electrical stimulus
for MEP is delivered, the surgical team should be Examples
alerted and permission sought. This is particularly Several examples of MEP monitoring are p rovided
true for surgeries in which a surgical microscope (Figures 5.10–5.13, Examples 9–12). These show the
is used. There may be times when the surgical MEPs in sequence during the surgery. Important
team declines MEP stimulation as patient move- surgical events are also listed in the graphs. The
ment cannot be tolerated. Because MEP stimulation narrative provides a brief history and an explana-
is delivered at defined intervals and only when tion of the changes seen. When available, imaging
allowed by the surgeon, MEP monitoring is not correlates to the surgery and monitoring are also
as “continuous” as SEP and BAEP monitoring, provided. It is important to remember that MEP
which are done continuously. The ideal frequency monitoring is not done in isolation; rather it is a
for MEP stimulation depends on the type of sur- part of multimodality monitoring that includes at
gery and what is occurring in the surgical field. The least SEP monitoring and often other types of mon-
more frequently the MEP monitoring is performed, itoring as well.
Age: 79 years
Sex: Male
Montage: Thenar–Hypothenar, Biceps–Triceps, Deltoid, Anterior tibialis, Abductor hallucis–Abductor digiti minimi
Diagnosis: Cervical spinal stenosis
Procedure: Posterior cervical decompression and fixation
Modalities monitored: Ulnar and tibial nerve SEPs; bilateral upper and lower limb MEPs; bilateral upper limb EMG
(A)

(B)
History: This is a 79-year-old male who presents with a 1-year history of neck pain and pain radiating down both arms. He also
complains of shoulder weakness and difficulty with raising his arms above his head. On examination, he has mild weakness of the
deltoid and bicep muscles, no sensory loss, and both biceps reflexes are reduced. He is scheduled to undergo posterior cervical
decompression and fixation.
Discussion: Ulnar and tibial nerve SEPs, upper and lower limb MEPs, and bilateral upper limb EMG monitoring are requested. The
MEP monitoring is shown here (A). No significant changes were noted in the MEP from the various muscles of the upper and lower
extremities.
An MRI of the cervical spine shows severe multilevel stenosis. A sagittal T2 weighted image (B) shows the severe stenosis that affects
the C4-C5 area the most.
This MEP monitoring does not show significant changes during the surgery. The montage includes MEP from the thenar–
hypothenar, biceps–triceps, deltoid, tibialis anterior, and foot muscles. The first three columns are for upper extremity muscles
and include two C5 muscles (deltoid and biceps). No significant change in MEP is noted in any muscle group. Unlike with BAEP
and SEP monitoring, MEP waveform morphology changes from one trial to the next. Notice that the MEP from the right tibi-
alis anterior (fourth column on the right side) is inconsistent. It is present but small in some traces and absent in others. This
response cannot be relied upon, and if absent should not prompt an alert unless other MEP waveforms change as well.
During this monitoring, no MEP changes were noted, and thus it is expected that this patient will not have worsening weakness
postoperatively.
EMG, electromyographic; MEP, motor evoked potential; SEP, somatosensory evoked potential.

Age: 39 years
Sex: Male
Montage: Abductor pollicis brevis–Abductor digiti minimi, Vastus lateralis–Vastus medialis, Anterior tibialis, Medial gastrocnemius,
Abductor hallucis–Abductor digiti minimi
Diagnosis: Retroperitoneal and abdominal mass encasing the thoracoabdominal aorta
Procedure: Laprotomy for resection of mass
Modalities monitored: Ulnar and tibial nerve SEPs; bilateral upper and lower limb MEPs
(A)
(B)

(C) (D)
History: This is a 39-year-old male with a history of metastatic testicular cancer. He has a retroperitoneal mass that encases several
organs including the descending aorta. On examination, he does not have sensory loss or weakness in his lower extremities. He is
scheduled to undergo surgery for removal of the mass.
Discussion: Ulnar and tibial nerve SEP and upper and lower limb MEP monitoring are requested as the tumor encases the aorta,
and injury to perforating arteries to the spinal cord is possible. The MEP monitoring from the left (A) and right (B) sides is shown
here. MEPs are present at baseline. During the surgery, there is sudden loss of all lower extremity MEPs. The surgeon is informed,
and corrective measures are taken. However, the MEPs do not recover.
A contrast-enhanced axial CT scan (C) of the abdomen shows the descending aorta (open arrow) surrounded by the tumor (solid
arrow). Another image showing the thorax and abdomen (D) demonstrates the aorta (open arrow) with the mass to the left of it
(solid arrow).
In this MEP monitoring, the left-sided MEPs are shown in panel A and right-sided MEPs in panel B. The first column is the
reference upper limb muscle (hand), and the next four columns are lower limb muscles (vastus group, anterior tibialis, medial
gastrocnemius, and foot muscles). At baseline, MEPs are noted in all muscle groups except the left vastus group (second column
in panel A). At 10:57:08, a sudden loss of lower extremity MEP on both sides is noted. Upper extremity (hand) MEPs are still
present. The presence of the reference (hand) MEP implies that technical and systemic changes are not the likely cause of the
change. The surgeon is immediately alerted and notes that tumor dissection is underway and he recently had a lot of bleed in
the field, suspecting injury to the artery of Adamkiewicz. This artery supplies the ASA and is responsible for perfusion to the
anterior two thirds of lower part of the spinal cord. Vascular injuries are difficult to correct intraoperatively and often result in
permanent injury. The lack of MEP improvement during the remainder of the surgery suggests that corticospinal tract injury in
the spinal cord has occurred. Tibial nerve SEPs are present during the entire case and do not change in amplitude when the
MEPs are lost.
This monitoring suggests that the patient had injury to the spinal cord corticospinal tract bilaterally. There is likely to be postopera-
tive lower extremity weakness.
ASA, anterior spinal artery; MEP, motor evoked potential; SEP, somatosensory evoked potential.

Age: 67 years
Sex: Male
Montage: Abductor pollicis brevis–Abductor digiti minimi, Anterior tibialis, Medial gastrocnemius, Abductor hallucis–Abductor
digiti minimi
Diagnosis: Descending thoracic aortic aneurysm
Procedure: Endovascular repair of descending thoracic aortic aneurysm
(A)

(B) (C)
History: This is a 67-year-old male with a history of an ascending aortic aneurysm with a concomitant descending thoracic aortic
aneurysm. He has undergone a Bentall hemi-arch repair 3 years ago and returns for surveillance of his descending thoracic aortic
aneurysm. He denies symptoms of chest pain, dyspnea, fatigue, or syncope. Imaging of the descending thoracic aortic aneurysm
reveals that it has expanded to 5.6 cm. Elective endovascular repair of the aneurysm is recommended.
Discussion: Ulnar and tibial nerve SEP and upper and lower limb MEP monitoring are requested. The MEP monitoring is shown here
(A). MEPs are present at the start of the case. Soon after the groin incision was made to insert the catheter (marked with the open
arrows), the anesthesiologist gave a bolus of propofol 100 mg for blood pressure management. The amplitude of the MEPs in the
upper and lower limb started to decrease. After about 20 minutes, the MEPs return to baseline (marked with solid arrows). At the
end of surgery, the MEPs are comparable to baseline.
A CT scan of the thorax and abdomen with contrast shows the descending thoracic aortic aneurysm that measured 5.6 cm in diam-
eter (B). A 3-D reconstruction image (C) shows the aneurysm in detail.
Anesthetics have a profound effect on MEP monitoring. Inhalational agents can greatly suppress MEPs, and intravenous agents like
propofol are preferred when MEP monitoring is needed. However, even high doses of propofol, especially if given as a bolus, can
affect MEP monitoring. In this example, four channels of MEP monitoring on each side are shown. The first channel is the upper
limb control (hand muscles) and the latter three channels are lower limb muscles (tibialis anterior, medial gastrocnemius, and foot
muscles). When propofol is administered (open arrows), notice that there is a decrease in the MEP in the upper and lower extremity
muscles. Such global involvement suggests a systemic cause for the change, especially since the surgery has just started. A decrease
in the blood pressure could also cause a similar change. About 20 minutes later, the propofol bolus is cleared, and the MEPs return
to near baseline. No further changes were noted.
Based on this monitoring, no injury to the corticospinal pathways is suspected as the change was transient and related to anesthetics.
MEP, motor evoked potential; SEP, somatosensory evoked potential.

Age: 63 years
Sex: Female
Montage: Abductor pollicis brevis–Abductor digiti m
inimi, Anterior tibialis, Medial gastrocnemius, Anterior tibialis–Abductor hallu-
cis, Abductor hallucis–Abductor digiti minimi
Diagnosis: Multiple traumatic thoracic spine burst fractures
(A)

(B) (C)
History: This is a 63-year-old female who presented after a high-speed, rollover motor-vehicle accident. She is complaining of upper
back pain that radiates to her chest. Examination shows normal sensory and motor function of her upper and lower limbs. A CT scan
reveals a T4 burst fracture, and a follow-up MRI scan shows additional T3 and T6 fractures. She is scheduled for an urgent posterior
spinal fusion and instrumentation.
Discussion: Ulnar and tibial nerve SEPs and upper and lower limb MEP monitoring are requested. The MEP monitoring is shown
here (A). Part way into the case, there were a lot of artifacts noted in the MEP channels, and the stimulating needle electrodes were
discovered to have been dislodged. After they were reinserted, MEP monitoring was comparable to baseline and no further changes
were noted.
An MRI of the spine showed the multilevel thoracic fractures. Sagittal T1 weighted (B) and T2 weighted (C) images show the verte-
bral column injury. The spinal cord does not appear to have abnormal signal.
In many emergency surgeries, the time to set up the patient for NIOM is very limited. In this patient, needle electrodes are inserted
into the scalp for MEP stimulation. After positioning, sway artifact is noted in all MEP channels, including upper limb (first column)
and lower limb (columns 2–5). This happens when the stimulating electrodes have been dislodged. This can be confirmed by check-
ing electrode impedances, which will be very high. After the needle electrodes are reinserted, the MEP monitoring is comparable to
baseline. Notice, however, that the morphology of the waveforms after the needles have been reinserted is slightly different. This is
because after reinsertion, the stimulating electrodes are probably not in exactly the same location as at the start of surgery. In many
cases, corkscrew electrodes are better for stimulation as they are more secure and less likely to be dislodged. In this case, due to
time limitations, needle electrodes were used.
The change in MEP monitoring in this surgery was due to technical issues and no injury to the corticospinal pathways was noted
throughout the surgery.
MEP, motor evoked potential; NIOM, neurophysiologic intraoperative monitoring; SEP, somatosensory evoked potential.

greatly affect evoked potential waveforms. When

Conclusions a significant change is noted in the monitoring, the
Evoked potentials are commonly used during surger- NIOM team must determine whether the change is
ies in which the nervous system is at risk. There are due to the surgical intervention or one of the many
convincing data that monitoring with evoked poten- other possible causes. Close communication with the
tials helps reduce neurologic morbidity in these sur- surgeon and anesthesiologist is imperative to ensure
geries. Anesthesia and other systemic factors can optimal monitoring.
References
1. ACNS. Guideline 11A: recommended standards 13. Husain AM, Ashton KH, Mohrhaus CA, et al. Effects
for neurophysiologic intraoperative monitoring: of cooling on brainstem auditory evoked potentials.
principles [online]. Available at: http://www.acns Clin Neurophysiol. 2006;117(suppl 1):S118.
.org/pdf/guidelines/Guideline-11A.pdf. 14. Radtke RA, Erwin CW. Intraoperative monitoring
2. Galloway GM, Nuwer MN, Lopez JR, et al. Intra- of auditory and brain-stem function. Neurol Clin.
operative Neurophysiologic Monitoring. Cambridge, 1988;6:899-915.
UK: Cambridge University Press; 2011. 15. Lopez JR. Neurophysiologic intraoperative moni-
3. Husain AM. A Practical Approach to Neurophysiologic toring of pediatric cerebrovascular surgery. J Clin
Intraoperative Monitoring. 2nd ed. New York, NY: Neurophysiol. 2009;26:85-94.
Demos Medical Publishing; 2015.
16. ACNS. Guideline 11B: recommended standards for
4. Nuwer MN. Intraoperative Monitoring of Neural interoperative monitoring of somatosensory evoked
Function. Amsterdam, Netherlands: Elsevier; 2008. potentials [online]. Available at: http://www.acns
5. Simon MV. Intraoperative Neurophysiology: A .org/pdf/guidelines/Guideline-11B.pdf.
Comprehensive Guide to Monitoring and Mapping. 17. Khan SA, James ML. Anesthetic considerations. In:
New York, NY: Demos Medical Publishing; 2010. Husain AM, ed. A Practical Approach to Neurophysiologic
6. ACNS. Guideline 9C: guidelines on short-latency Intraoperative Monitoring. 2nd ed. New York, NY:
auditory evoked potentials. J Clin Neurophysiol. Demos Medical Publishing; 2015:76-84.
2006;23:157-167. 18. Stecker MM, Cheung AT, Pochettino A, et al. Deep
7. ACNS. Guideline 11 C: recommended standards hypothermic circulatory arrest: I. Effects of cooling
for intraoperative monitoring of auditory evoked on electroencephalogram and evoked potentials.
potentials [online]. Available at: http://www.acns Ann Thorac Surg. 2001;71:14-21.
.org/pdf/guidelines/Guideline-11C.pdf. 19. Nuwer MR, Dawson EG, Carlson LG, et al.
8. Colletti V, Fiorino FG, Mocella S, et al. ECochG, Somatosensory evoked potential spinal cord mon-
CNAP and ABR monitoring during vestibular itoring reduces neurologic deficits after scolio-
Schwannoma surgery. Audiology. 1998;37:27-37. sis surgery: results of a large multicenter survey.
Electroencephalogr Clin Neurophysiol. 1995;96:6-11.
9. Simon MV. Neurophysiologic intraoperative moni-
toring of the vestibulocochlear nerve. J Clin 20. Nuwer MR, Emerson RG, Galloway G, et al. Evidence-
Neurophysiol. 2011;28:566-581. based guideline update: intraoperative spinal moni-
10. James ML, Husain AM. Brainstem auditory evoked toring with somatosensory and transcranial electrical
potential monitoring: When is change in wave V sig- motor evoked potentials: report of the Therapeutics
nificant? Neurology. 2005;65:1551-1555. and Technology Assessment Subcommittee of the
American Academy of Neurology and the American
11. Harner SG, Harper CM, Beatty CW, et al. Far-field Clinical Neurophysiology Society. Neurology.
auditory brainstem response in neurotologic sur- 2012;78:585-589.
gery. Am J Otol. 1996;17:150-153.
21. Khealani B, Husain AM. Neurophysiologic intraop-
12. Husain AM. Neurophysiologic intraoperative mon- erative monitoring during surgery for tethered cord
itoring. In: Ebersole JS, Husain AM, Nordli DR, eds. syndrome. J Clin Neurophysiol. 2009;26:76-81.
Current Practice of Clinical Electroencephalography.
Philadelphia, PA: Wolters Kluwer Health; 2014: 22. Legatt AD, Emerson RG, Epstein CM, et al. ACNS
488-542. Guideline: transcranial electrical stimulation motor
evoked potential monitoring. J Clin Neurophysiol. 30. Minahan RE, Sepkuty JP, Lesser RP, et al. Anterior spi-
2016;33:42-50. nal cord injury with preserved neurogenic “motor”
23. Macdonald DB. Intraoperative motor evoked poten- evoked potentials. Clin Neurophysiol. 2001;112:
tial monitoring: overview and update. J Clin Monit 1442-1450.
Comput. 2006;20:347-377. 31. Kothbauer KF. The interpretation of muscle motor
24. Chen Z. The effects of isoflurane and propofol on evoked potentials for spinal cord monitoring. J Clin
intraoperative neurophysiological monitoring during Neurophysiol. 2017;34:32-37.
spinal surgery. J Clin Monit Comput. 2004;18:303-308. 32. Kothbauer KF, Deletis V, Epstein FJ. Motor-
evoked potential monitoring for intramedul-
25. MacDonald DB. Overview on criteria for MEP
lary spinal cord tumor surgery: correlation of
monitoring. J Clin Neurophysiol. 2017;34:4-11.
clinical and n europhysiological data in a series
26. Calancie B. Intraoperative neuromonitoring and of 100 consecutive procedures. Neurosurg Focus.
alarm criteria for judging MEP responses to tran- 1998;4:e1.
scranial electric stimulation: the threshold-level
33. MacDonald DB, Janusz M. An approach to intra-
method. J Clin Neurophysiol. 2017;34:12-21.
operative neurophysiologic monitoring of thora-
27. Journee HL, Berends HI, Kruyt MC. The percentage of coabdominal aneurysm surgery. J Clin Neurophysiol.
amplitude decrease warning criteria for transcranial 2002;19:43-54.
MEP monitoring. J Clin Neurophysiol. 2017;34:22-31.
34. Husain AM, Swaminathan M, McCann RL, et
28. Segura MJ, Talarico ME, Noel MA. A multipara- al. Neurophysiologic intraoperative monitor-
metric alarm criterion for motor evoked potential ing during endovascular stent graft repair of the
monitoring during spine deformity surgery. J Clin descending thoracic aorta. J Clin Neurophysiol.
Neurophysiol. 2017;34:38-48. 2007;24:328-335.
29. MacDonald DB. Safety of intraoperative transcranial 35. Sloan TB, Jameson LC. Electrophysiologic monitor-
electrical stimulation motor evoked potential moni- ing during surgery to repair the thoraco-abdominal
toring. J Clin Neurophysiol. 2002;19:416-429. aorta. J Clin Neurophysiol. 2007;24:316-327.
Index
AAN. See American Academy of anterior spinal artery (ASA), 280, 295 electrode properties, 6
Neurology artifact rejection, 11 examples of, 138–187
abnormal evoked potentials ASA. See anterior spinal artery filters, 11
abnormal waveforms, 30 auditory evoked potential (AEP) intensity, 115, 124–126
absence of waveforms, 30 brainstem. See brainstem interpeak latency, 129, 134, 135
amplitude reduction, 22–30 auditory evoked potentials interpretation, 136
latency prolongation, 22, 26 auditory evoked potentials (AEPs), 1 latency/intensity series, 137,
localization, 30 brainstem. See brainstem auditory 140–145
absolute latency, 15 evoked potentials (BAEPs) masking noise, 123
ACNS. See American Clinical averaging, 11 maturational issues, 135
Neurophysiology Society methodological considerations, 136
acute demyelinating conditions, 62 multiple sclerosis, 137
AEPs. See auditory evoked potentials. BAEPs. See brainstem auditory neurophysiologic intraoperative
AION. See anterior ischemic optic evoked potentials monitoring, 146
neuropathy baseline to peak amplitude, 17 newborn hearing screening, 146
ALS. See amyotrophic lateral sclerosis behavioral audiogram, 136–137, obligate waveforms, 129, 134
amblyopia, 63 138–139 patient variables, 114, 116, 117
American Academy of Neurology bi-fed waveform, 14 peaks of, 1
(AAN), 218 binocular latency prolongation, 56 physiologic changes, 268
American Clinical Neurophysiology brainstem auditory evoked polarity effect on latencies of, 115,
Society (ACNS), 123, 199 potentials (BAEPs), 1, 265 118–122
amplifiers, properties of absence of waveforms, 31 recording montage, 123, 128
artifact rejection, 11 absolute latency, 129, 134, 135 recording techniques, 266–267
averaging, 11 amplifier settings, 123, 130–132 sample report, 31, 33
differential, 10 amplitude abnormalities, 134 stimulation intensity, 4
filters, 11 anatomy and physiology, 265–266 stimulation rates, 115, 123, 127
input board, 10 anesthetic agents effect on, 267 stimulation techniques, 266
selector switches, 10 behavioral audiogram, 136–137, surgery-induced changes, 268
amplitude 138–139 technical issues, 268
abnormalities, 134 cerebellopontine angle tumors, 137 types of stimuli, 115
ratio, 17–19 classification of, 113 vascular lesions effect on, 146
reduction, 22, 26–27 clinical applications, 269 warning criteria, 267
amyotrophic lateral sclerosis coma evaluation, 146 waveforms, 128, 133, 265, 266
(ALS), 218 definition of, 113
analog filters, 11 dysmyelinating diseases,
anterior ischemic optic neuropathy childhood, 137 cathode ray tube (CRT) monitors, 37
(AION), 62 electrode placement, 123 central amplification, 4, 192
311
312 ■ Index
cerebellopontine (CP) angle visual. See visual evoked abnormal evoked potentials, 22
tumors, 137 potentials (VEPs) brainstem auditory evoked
Charcot–Marie–Tooth disease waveform generators, 2–4 potentials, 267
type I, 218 expected waveforms, 50, 52, 54–55, latency/intensity (L/I) series, 137,
chiasmatic disorders, 63 208, 212 140–145
CMRR. See common mode LCD monitors. See liquid crystal
rejection ratio display monitors.
coma, brainstem auditory evoked far-field potentials, 2, 3 LED goggles. See light emitting
potentials, 146 filters, 11 diode (LED) goggles
common mode rejection ratio flash visual evoked potential (FVEP), leukodystrophies, 218
(CMRR), 10 1, 35, 59–60, 62, 110 LFFs. See low-frequency filters
cones, 35 full-field PRVEP, 35, 36 light emitting diode (LED)
contrast, visual evoked potentials, full-field stimulation, 45, 58 goggles, 60
37, 42–43 FVEP. See flash visual evoked liquid crystal display (LCD)
cortical evoked potentials, 2 potential monitors, 37
CP angle tumors. See L/I series. See latency/intensity
cerebellopontine (CP) angle (L/I) series
tumors hearing level (HL), 115 low-frequency filters (LFFs), 11
crossed asymmetry, 57, 63 HFFs. See high-frequency filters luminance, visual evoked potentials,
CRT. See cathode ray tube hemifield stimulation, 36–37, 48, 37, 42–43
57–59
high-frequency filters (HFFs), 11
demyelinating neuropathies, 62 HL. See hearing level MEPs. See motor evoked potentials
differential amplifier, 10 metabolic disorders, 62–63
digital filters, 11 monocular hemifield amplitude
duration of stimulus, 4 input board, amplifiers, 10 ratio, 58
dysmyelinating diseases, interhemispheric amplitude ratio, 53 monocular hemifield stimulation, 57
childhood, 137 interindividual variability, 19 monocular latency prolongation, 56
interocular hemifield amplitude motor evoked potentials (MEPs),
ratio, 58 265, 294–299
electrode(s), 4, 6–10 interpeak latency (IPL), 15, 22, “all-or-none” rule, 296
placement, 10 128–129, 134 anatomy and physiology, 294–295
properties, 6, 9 prolongations, 30, 32, 134–135, anesthetic agents effects on,
types, 9–10 214–216 295–296
evoked potentials, 1 intraindividual variability, 19, 20 clinical applications, 298–299
abnormal. See abnormal evoked intraoperative monitoring, physiologic changes, 297
potentials somatosensory evoked recording techniques, 295
amplifiers, 10 potentials, 286–287 safety, 297–298
amplitude ratio, 17, 19 IPL. See interpeak latency stimulation techniques, 295
normal. See normal evoked surgery induced changes, 297
potentials technical issues, 297, 298
pattern reversal. See pattern Krabbe disease, 63 warning criteria, 296
reversal visual evoked MS. See multiple sclerosis
potential (PRVEP) multiple sclerosis (MS), 137, 217
peak identification, 14–16 latency Myer’s loop, 36
peaks of, 1–2 absolute, 15, 129, 134, 135
P37 waveform, 3 P1/P100 waveform, 52
recording methodology, pattern reversal visual evoked NCS. See nerve conduction studies
6, 9–11 potential, 52 near-field potentials, 2
somatosensory. See somatosensory evoked nerve conduction studies (NCS), 192
somatosensory evoked potentials, 15, 17 neurodegenerative diseases, 218
potentials (SEPs) interpeak. See interpeak latency neurophysiologic intraoperative
stimulation methodology, 4, 6 (IPL) monitoring (NIOM), 146, 199
types of, 1–4 measurement, 15 newborn hearing screening, 146
variability, 19 prolongation, 22–24, 56 nHL. See normal hearing level
Index ■ 313
NIOM. See neurophysiologic selector switches, 10 sound pressure level (SPL), 115
intraoperative monitoring sensory level (SL), 115 spinal cord evoked potentials, 2
N105 waveform, 15, 52, 53, 67, 97 SEPs. See somatosensory evoked SPL. See sound pressure level
normal evoked potentials, 19, 22. potentials stimulation intensity, 4
See also abnormal evoked SL. See sensory level stimulus rate, 4, 6
potentials smoothing, digital filtering, 11 striate cortex, 36
normal hearing level (nHL), 115 somatosensory evoked potentials strobe light, 35
(SEPs), 1, 265, 280
abnormalities, 214–216
obligate waveforms, 2, 129, 134–135 absolute latency of waveform, tibial nerve, somatosensory evoked
ophthalmologic disorders, 62, 63 15, 17 potential monitoring. See
optic nerve fibers, 36 amplifier settings, 208, 209–211 somatosensory evoked
optic neuritis/multiple sclerosis, 62 amplitude reduction, 22, 26–27 potentials
anatomy and physiology, 280 TM. See transverse myelitis
anesthetic agents effects on, 282 transverse myelitis (TM), 217
paradoxical lateralization, 52 brainstem lesions, 217–218
parasagittal, 53 classification of, 191
pattern reversal visual evoked clinical applications, 284–285 uncrossed asymmetry, 57, 63, 91,
potential (PRVEP), 1, 35, coma, 218 101, 105, 108. See also crossed
59–62, 101 dorsal column pathways, asymmetry
abnormalities, 53, 58, 87, 91, 94, 97 191, 192
absolute latency, 52 duration of electrical pulse, 198
expected waveforms, 50 electrode placement, 199 vascular lesions, 146
intensity of stimulation, 4 examples of, 218–263 ventral posterior lateral (VPL), 191
maturational changes in, 60 intensity, 192, 195–197 VEPs. See visual evoked potentials
normative data, 53 intensity of stimulation, 4 visual cortex, 36, 52, 58
peak identification, 14–15, 16 interpretation, 217 visual evoked potentials (VEPs),
peak to peak amplitude, 15, 17, 18 intraindividual variability, 1. See also somatosensory
peripheral nerve evoked 19, 20 evoked potentials (SEPs)
potentials, 2 intraoperative monitoring, abnormalities, 53, 56–57
peripheral neuropathies, 218 286–287 abnormal waveforms, 28–30
peroneal nerve, somatosensory latency prolongation, 22–24 anatomy, 35–36
evoked potential monitoring. leukodystrophies, 218 chiasmatic disorders, 63
See somatosensory evoked maturational issues, 216–217 classification, 35
potentials multiple sclerosis (MS), 217 display monitor, 37, 44–45
persistent vegetative state (PVS), 146 neurodegenerative diseases, 218 electrode placement, 48, 49
P1/P100 waveform, 48, 50, 60 parameters measured, 212–214 examples of, 63–110
abnormalities, 53, 56–57 patient variables, 192–194 expected waveform, 50, 52,
absolute latency of, 52 peak identification, 15 54–55
latency prolongation, 53, 56 peaks of, 1 field of stimulation, 45, 48
postchiasmatic disorders, 63 peripheral neuropathies, 218 filters, 11, 12
posterior spinal arteries (PSAs), physiologic changes, 283 flash visual evoked potential,
280, 295 rate of stimulation, 200 59–62
prechiasmatic disorders, 62–63 recording montage, 199, 202 hemifield, 57–59
PRVEP. See pattern reversal visual recording techniques, 281–282 hysterical blindness, 63
evoked potential sample report, 33 interpretation of, 50, 52–57
PSAs. See posterior spinal arteries stimulation rate, 198 latency, 37, 52
PVS. See persistent vegetative state stimulation techniques, 280–281 luminance and contrast, 37, 42–43
surface/needle electrodes, 198 metabolic disorders, 62–63
surgery induced changes, 283, 284 montage, 50
“Queen Square” method, 48, 49 technical issues, 284 neuropathies, 62
transverse myelitis, 217 number of repetitions, 6, 8
unilateral/bilateral stimulation, ophthalmologic disorders, 62, 63
replication of evoked potential, 6 198–199 optic neuritis/multiple sclerosis, 62
rods, 35 warning criteria for, 282–283 patient variables, 36–37, 38–41
314 ■ Index
visual evoked potentials (VEPs) (cont.) tumors, 62 bi-fed, 14

peaks of, 1 visual angle, 37, 45 expected, 208, 212
peak to peak amplitude, 15, 17, 18 VPL. See ventral posterior lateral N105, 15, 52, 53, 67, 97
P100 waveform. See P1/P100 obligate, 2, 129, 134–135
waveform P1/P100, 48, 50, 60
postchiasmatic disorders, 63 waveform generators, 2–4, 280 abnormalities, 53, 56–57
prechiasmatic disorders, 62–63 waveforms, 280 absolute latency of, 52
Queen Square method, 49 abnormal, 28, 30 latency prolongation, 53, 56
sample report, 33 absence of, 30 P37, 3
stimulation rate, 37 absolute latency of, 15, 17 “W-shaped” P100 waveform, 50, 57

Potenciales Evocados Aaitif M Husain

Uploaded by

Copyright:

Available Formats

You might also like

Potenciales Evocados Aaitif M Husain

Uploaded by

Document Information

Original Title

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

Potenciales Evocados Aaitif M Husain

Uploaded by

Copyright:

Available Formats

Illustrated Manual of

Illustrated Manual of Clinical Evoked Potentials

Clinical Evoked Potentials

n Detailed review of methodology of evoked potential studies

An Imprint of Springer Publishing

Acquisitions Editor: Beth Barry

Copyright © 2018 Springer Publishing Company.

Library of Congress Cataloging-in-Publication Data

Contact us to receive discount rates on bulk purchases.

Printed in the United States of America by Publishers’ Graphics.

1. Basics of Evoked Potentials 1

SEP, somatosensory evoked potential.

Figure 1.1 Tibial nerve SEP near-field and far-field potentials.

prolonged, the site of pathology is at or distal to its Parameters

SEP, somatosensory evoked potential.

Figure 1.2 Median nerve SEP waveform latency variations.

BAEP, brainstem auditory evoked potential.

Figure 1.3 (A and B) Effect of stimulation rate on BAEP waveforms.

Age: 43 years (B)

Figure 1.4 (A–E) Effect of number of repetitions on VEP waveforms. (continued )

Figure 1.4 (A–E) (continued )

Averaging are not sine waves. It is not possible to determine

Figure 1.5 (A–C) Effect of filter changes on VEP waveforms.

Figure 1.6 (A–C) Effect of smoothing filter on BAEP waveforms. (continued )

BAEP, brainstem auditory evoked potential.

Figure 1.6 (A-C) (continued )

the recording method. An example of the latter Absolute latency

SEP, somatosensory evoked potential.

Figure 1.7 Automated tagging of peroneal nerve SEP waveforms.

IPL, interpeak latency; SEP, somatosensory evoked potential.

Figure 1.8 Latency measurements of SEP waveforms.

Absolute Latency Amplitude

P100 Amplitude Ratio

Figure 1.9 Amplitude measurements of VEP waveforms.

waveform. This is commonly used in BAEPs where Interindividual variability

SEP, somatosensory evoked potential.

Figure 1.10 (A and B) (continued )

IPL, interpeak latency; SEP, somatosensory evoked potential.

Figure 1.11 Latency prolongations of median nerve SEP waveforms.

IPL, interpeak latency; SEP, somatosensory evoked potential.

Figure 1.12 (A and B) (continued )

SEP, somatosensory evoked potential.

Figure 1.13 (A and B) (continued )

Absolute Latency Amplitude

P100 Waveform Amplitude Ratio

Figure 1.14 (A and B) W-shaped VEP waveform. (continued )

Absolute Latency Amplitude

P100 Amplitude Ratio

Figure 1.14 (A and B) (continued )

BAEP, brainstem auditory evoked potential.

Figure 1.15 Absence of BAEP waveforms.

BAEP, brainstem auditory evoked potential; IPL interpeak latency.

Figure 1.16 Prolongation of IPLs of BAEP waveforms.

INTERPRETATION: This is a normal visual evoked response study.

Figure 1.17 Sample VEP, BAEP, and SEP reports.

are available to enhance the waveforms. Guidelines

Absolute Latency Amplitude

Absolute Latency Amplitude