C L I N I C A L A N D E X P E R I M E N TA L

RESEARCH

Stereoacuity measurement using a phoropter combined with two 4K

smartphones

Clin Exp Optom 2018; 101: 272–275 DOI:10.1111/cxo.12614

Huang Wu MD Purpose: To evaluate stereoacuity using a combination of a phoropter and two 4K

Shu Liu MD smartphones.
Ruiqing Wang MOptom Methods: A stereopsis measurement system was established using a phoropter (Topcon
Department of Optometry, Second Hospital of Jilin VT-10) and two Sony 4K smartphones (Sony Xperia Z5 Premium Dual E6883). A total of
University, Changchun, Jilin, China 154 subjects (69 children and 85 adults) were enrolled for evaluation using this system
E-mail: wuhuang@jlu.edu.cn
and subsequent comparison with the Fly Stereo Acuity test.
Results: There was a high level of agreement between the two methods using Bland-
Altman statistical analysis (children group, 95 per cent limits of agreement
0.14 log arc-
seconds [arcsec]; adult group, 95 per cent limits of agreement
0.13 log arcsec).
Submitted: 21 April 2017 Conclusion: The closely spaced pixels of a 4K smartphone display enable measurement of
Revised: 13 July 2017 stereoacuity at a relatively short distance. The flexibility and versatility of the mobile test
Accepted for publication: 2 September 2017 system are likely to be useful in clinical practice.

Stereopsis is an important binocular func-
tion that enables subjects to judge distance
with precision, while stereoacuity measures
the detectable differences in depth; the lat-
ter can be evaluated with a variety of tools.
Some methods are used in real depth, such
as the Howard-Dolman apparatus1 and
Frisby stereotest.2 In clinical practice, bin-
ocularity is often broken using a certain
device before performing the examination,
which means that what the right eye sees
cannot be seen by the left eye, and vice
versa. The differences in the two images –
each seen by a single eye – constitute the
geometric parallax and are used to detect
the threshold of stereopsis. The smaller the
parallax one can distinguish, the better the
stereoacuity.
Binocular equipment – such as red-
green/red-blue glasses (TNO [The Nether-
lands Optical Society] Stereoacuity Test3) –
or polarised glasses (Titmus Stereoacuity
Test4) can be used to separate the two eyes.
With the development of information tech-
nology, computer-aided, three-dimensional
(3D) technology has become a reality. Kim
et al.5 developed a contour-based stereotest
using a 3D polarisation monitor that can
measure 5,000 to 20 arcseconds (00 ) of dis-
tance stereopsis. We also established a
stereoacuity measurement system with 3D
shutter glasses technology.6 Compared
with 3D computer polarisation techniques,
active shutter glasses technology can
express 3D images without reducing the
vertical resolution; however, observers still
have to wear shutter glasses, which are
uncomfortable, especially for those who
have to wear corrective spectacles. The test-
ing distance remains a problem due to the
screen resolution; hence, the test should be
performed at a relatively long distance.
Eliminating the need for glasses to
achieve a 3D effect can be realised with
the use of autostereoscopic – or naked eye
– 3D displays, which are now used in
advertising and home entertainment.
Researchers have made great efforts to use
this method to measure stereopsis in clini-
cal and research settings. Fujikado et al.7
evaluated the stereopsis of strabismus
patients using a 10-inch liquid crystal dis-
play equipped with an image splitter sys-
tem, which allowed 3D images to be seen
without glasses. Breyer et al.8 established a
random-dot stereotest – also based on use
of an autostereoscopic monitor – and
achieved high correlation with the Lang I
Stereotest in children older than three
years. However, the relatively low screen
resolution of autostereoscopic tests may
still be a limiting factor in the evaluation
of stereopsis, rendering these tests more
useful as qualitative tools than precise
quantitative tools.
The display of a 4K mobile phone
addresses this problem due to its extra-
dense dot pitch, which can produce a dis-
parity small enough to measure the
stereoacuity at a relatively short distance.
We combined two smartphones to display
pictures synchronously with specific dispar-
ities, thus producing stereo symbols when
viewed binocularly with the aid of a
phoropter.
METHODS
Combination of two 4K
smartphones and a phoropter
We set up a stereopsis measurement system
using a phoropter (Topcon VT-10, Topcon
Corp, Tokyo, Japan) and two Sony smart-
phones (Sony Xperia Z5 Premium Dual
E6883; resolution, 3840 × 2160; Sony
Mobile Communications Inc., Tokyo,
Japan). The two phones were placed side
by side using cell phone holders and sus-
pended on a near vision rod. A string was
used to support the weights of the cell
phones by connecting the end of the rod
and the top knob of the phoropter. An
acrylonitrile butadiene styrene (ABS) plas-
tic sheet (22 cm × 38 cm) with a thickness
of 0.6 mm was attached to the near vision
rod to separate completely the two eyes
(Figure 1).

Clinical and Experimental Optometry 101.2 March 2018 © 2017 Optometry Australia
272

Figure 1. Photograph showing the phorop-
ter combined with two 4K smartphones to
establish a stereoacuity measurement system Figure 2. Legend of test pages
Test symbols
We copied the Fly Stereo Acuity test
(Vision Assessment Corporation, Elk Grove
Village, Illinois, USA), which contains
10 squares, with each square containing
four circles. A stereo circle was hidden in
such a position that it could be detected
using polarised glasses only if the stereoa-
cuity of the subject was better than the dis-
parity of the specific stereo target. The
testing distance was set at 40 cm, and the
disparities of the 10 targets were 40000 ,
20000 , 16000 , 10000 , 6300 , 5000 , 4000 , 3200 , 2500
and 2000 .
The dot pitch of the display of each cell
phone was 0.0317 mm, and it could
achieve an accuracy of up to 1000 (arcse-
cond, 1 pixel disparity) at a testing distance
of 65 cm. Based on the size of the screen,
we drew on a pair of pages, each page con-
taining nine squares, with the disparities of
the stereo symbols being 40 pixels (px),
20 px, 16 px, 10 px, 6 px, 5 px, 4 px, 3 px
and 2 px; these represented 40000 , 20000 ,
16000 , 10000 , 6000 , 5000 , 4000 , 3000 and 2000 ,
respectively, at a testing distance of 65 cm
(Figure 2). Three values – 6300 , 3200 and 2500
– could not be imitated at the regular
test distance (65 cm), but changing the test
distance solved the problem. Square
5 (5 px disparity), square 8 (3 px disparity)
and square 9 (2 px disparity) of the smart-
phones could imitate square 5 (6300 ),
square 8 (3200 ) and square 9 (2500 ) of the
Fly Stereo Acuity test at testing distances of
62 cm, 61 cm and 52 cm, respectively.
Test procedure
(1) The subject was seated behind the
phoropter with the distance spectacle
© 2017 Optometry Australia
Measuring stereopsis using phoropter and smartphones Wu, Liu and Wang
correction in place. The phoropter was
adjusted to the interpupillary distance
of the patient for near.
(2) Two smartphones fixed together side
by side were hung on the near vision
rod at a testing distance of 65 cm, sep-
arated by the plastic sheet.
(3) The rotary (Risley) prisms were placed
before the eyes. Both prisms were
rotated to 5.5Δ base out to help the
subject fuse the two smartphones into
one image.
(4) The subject was asked to identify the
circle that was more protuberant than
the others in square 1 (up, down, right
or left).
Subsequently, the subject was directed to
proceed to square 2, and so on, until the
subject could not distinguish any promi-
nent circle. The square immediately pre-
ceding it was recorded as the stereoacuity
value for the subject. If a subject success-
fully identified all nine tests, stereoacuity
was recorded as 2000 . If a subject could
point out the target circle in square 4 but
failed at square 5, the phones were moved
to a testing distance of 62 cm, and the sub-
ject was evaluated for square 5 again. If the
subject identified it correctly, a value of 6300
was recorded; if incorrect, a value of 10000
was recorded. If a subject could distinguish
the circle in square 7 but failed at square
8, the phones were pushed to 61 cm, and
evaluated for square 8 again. If the answer
of the subject was correct, 3200 was
recorded; otherwise, 4000 was recorded. If a
subject could identify the target circle in
square 8, but failed at square 9, the dis-
tance of the phones was changed to 52 cm,
and the subject was asked to distinguish
square 9 again. If the subject identified it
Clinical and Experimental Optometry 101.2 March 2018
14440938, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cxo.12614 by Readcube (Labtiva Inc.), Wiley Online Library on [18/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
correctly, a value of 2500 was recorded; if
incorrect, a value of 3200 was recorded.
Using this method, the smartphones can
be used to imitate all 10 values of the Fly
Stereo Acuity test.
We examined the subjects using the two
sets randomly, with a three-minute break
between the two tests.
Subjects
A total of 154 subjects, whose stereoacuity was
at least 40000 as evaluated by the Fly Stereo Acu-
ity test, were enrolled. The children group
comprised 69 primary school students (38 boys
and 31 girls, aged 10 to 12 years); the adult
group included 85 subjects (38 men and
47 women, aged 20 to 35 years).
Before participation in the study, a writ-
ten, informed consent was obtained from
all adult participants; informed consent was
obtained from the guardian for all under-
age participants. The research protocol fol-
lowed the tenets of the Declaration of
Helsinki and was approved by the ethics
committee of the Second Hospital of Jilin
University (No. 2017-89).
Statistical analysis
All data were processed using MedCalc Sta-
tistical Software (version 17.6, MedCalc
Software bvba, Ostend, Belgium). Stereoa-
cuity values were transformed to log arcsec
for analysis.9 The Bland-Altman method
was used to evaluate the agreement
between the two tests.
RESULTS
Children group
The mean stereoacuity measured by the Fly
Stereo Acuity test was 1.66
0.21 log arc-
sec, compared with 1.66
0.22 log arcsec
measured using two 4K smartphones. The
95 per cent limits of agreement (LoA) were
−0.13 to 0.14 log arcsec. The maximum
allowed difference between methods (the
interval between the lower 95 per cent con-
fidence interval [CI] limit of the lower LoA
and the higher 95 per cent CI limit of the
higher LoA 95 per cent CI) was −0.16 to
0.17 log arcsec (Figure 3A).
Adult group
The mean stereoacuity measured by the Fly
Stereo Acuity test was 1.53
0.22 log
273
 14440938, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cxo.12614 by Readcube (Labtiva Inc.), Wiley Online Library on [18/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Measuring stereopsis using phoropter and smartphones Wu, Liu and Wang

Figure 3. Bland-Altman plots for comparison between the Fly Stereo Acuity test and use of two 4K smartphones. A: Children’s
group. Mean difference between methods was 0.0044 log arcsec, and 95% confidence interval (CI) of the mean was −0.012 to 0.021
log arcsec. The 95% limit of agreement (LoA) was −0.13 to 0.14 log arcsec; the 95% CI limit of the lower LoA was −0.16 to −0.10 log
arcsec; the 95% CI limit of the higher LoA was 0.11 to 0.17 log arcsec. B: Adult group. Mean difference between methods was
0.0035 log arcsec, and 95% CI of the mean was −0.010 to 0.017 log arcsec. The 95% LoA was −0.12 to 0.13 log arcsec; the 95% CI
limit of the lower LoA was −0.15 to −0.10 log arcsec; the 95% CI limit of the higher LoA was 0.11 to 0.15 log arcsec.

arcsec, compared with 1.53
0.23 log arc-
sec measured using two 4K smartphones.
The 95 per cent LoA was −0.12 to 0.13
log arcsec. The maximum allowed differ-
ence between methods was −0.15 to 0.15
log arcsec (Figure 3B).
In terms of the measurement accuracy of
stereopsis tests in the clinic,
5 arcsec is
small enough for routine work. This
equates with
0.7 log arcsec, which is
larger than the LoA of the two methods
(
0.14 log arcsec for children and
0.13
log arcsec for adults) and the maximum
allowed difference between methods
(
0.17 log arcsec for children and
0.15
log arcsec for adults) in our study. There-
fore, the two measurements showed a high
level of agreement.
DISCUSSION
The phoropter is a conventional tool used
by an optometrist in daily practice. Measur-
ing the stereoacuity is another common
means of evaluating binocular function.
Combining both is convenient for the
examiner. Smartphones can easily achieve
the effects doctors and researchers desire,
including contour-based figures, random-
dot stereograms, different colors and
contrasts, and diverse sizes and forms.
Moreover, separation of the two eyes in this
method is more thorough than in the
polarising technique or active liquid crystal
shutter glasses used in computer 3D
evaluations.
Although the Fly Stereo Acuity test is not
the gold standard for stereoacuity measure-
ment, it is widely used clinically because it
is simple and convenient. Therefore, we
used the Fly Stereo Acuity test as a compar-
ison; moreover, it was easy to duplicate.
The two methods showed high agreement
in our experiment. A limitation of the Fly
Stereo Acuity test is the unequal measure-
ment steps, which affect measurement
accuracy, especially in preceding squares.
This distribution characteristic affects the
interpretation of test results. For example,
a score of 16000 actually means that the
stereoacuity is in a range >10000 to 16000 .
Similarly, a score of 10000 means the
stereoacuity is in a range of >6300 to 10000 .
Therefore, the difference between 16000
and 10000 is not simply 6000 , but within an
interval from 100 (10100 to 10000 ) to 9600
(16000 to 6400 ).
Stereoacuity measurement using smart-
phones can be performed flexibly, depend-
ing on the purpose; this is a major
advantage over traditional methods of
stereoacuity measurement. We can estab-
lish an equidistant stereoacuity measure-
ment with 4K smartphones as we did using
a 3D laptop equipped with shutter glasses
technology.6 This allows us to determine
stereoacuity with an accuracy of 1000 using
three stepped choices, and it can reduce
the deviation in clinical and research
results.
With the development of computer-
aided 3D technology, mature 3D tech-
niques such as polarisation technology or
active shutter glasses technology have been
used to detect the threshold of stereopsis.
The use of polarisation glasses or shutter
glasses may affect binocular stereopsis, and
their use is different from the actual binoc-
ular fusion experience in the real world.10
Based on previous studies, Kim et al.10
designed an improved four-view parallax
barrier system for the random dot stereot-
est without the need to wear glasses. How-
ever, the autostereoscopic system also has
its limitations. Although subpixel rendering
may increase the measuring range, limita-
tions in pixel intervals still render the test
insufficient for near distance examination.
The calibration process is important in the
glasses-free test system, and the head of the
subject must stay motionless throughout
the test procedure; otherwise, crosstalk may
affect the results.
Glasses-free stereotesting does not mean
eliminating corrective spectacles, but it
means testing without the need for

Clinical and Experimental Optometry 101.2 March 2018 © 2017 Optometry Australia
274

polarisation glasses, shutter glasses or ana-
glyph glasses, while a phoropter can supply
optimal optical correction. The limitation
of the dot pitch of the 3D display was
addressed by using 4K smartphones which
have an extremely fine dot pitch. Combin-
ing a phoropter with two 4K smartphones
allowed us to measure stereoacuity at a rel-
atively short distance without additional
glasses. This method still has obvious limita-
tions, including the fact that it is unsuitable
for small children who cannot cope with
examination by phoropter.
It is possible that intelligent terminals
with smaller pixel intervals will be manufac-
tured in the near future. This will usher
in rich and fascinating opportunities for
evaluating stereoacuity, helping us develop a
deeper understanding of human stereopsis.
© 2017 Optometry Australia
Measuring stereopsis using phoropter and smartphones Wu, Liu and Wang
ACKNOWLEDGEMENTS
This study was supported by a grant from
Jilin Provincial Science & Technology
Department, China (No. 20170519004JH).
REFERENCES
1. Saladin JJ. Phorometry and Stereopsis, In
Benjamin WJ (ed). Borish’s Clinical Refraction. 2nd
edn. St. Louis: Butterworth–Heinemann Elsevier,
2006: 899–960.
2. Anketell PM, Saunders KJ, Little JA. Stereoacuity
norms for school-age children using the Frisby
stereotest. J AAPOS 2013; 17: 582–587.
3. van Doorn LL, Evans BJ, Edgar DF, Fortuin MF.
Manufacturer changes lead to clinically important
differences between two editions of the TNO
stereotest. Ophthalmic Physiol Opt 2014; 34:
243–249.
4. Tejedor J, Ogallar C. Comparative efficacy of
penalization methods in moderate to mild ambly-
opia. Am J Ophthalmol 2008; 145: 562–569.
Clinical and Experimental Optometry 101.2 March 2018
14440938, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/cxo.12614 by Readcube (Labtiva Inc.), Wiley Online Library on [18/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
5. Kim J, Yang HK, Kim Y, Lee B, Hwang JM. Dis-
tance stereotest using a 3- dimensional monitor
for adult subjects. Am J Ophthalmol 2011; 151:
1081–1086.
6. Wu H, Jin H, Sun Y et al. Evaluating stereoacuity
with 3D shutter glasses technology. BMC Ophthal-
mol 2016; 16: 45.
7. Fujikado T, Hosohata J, Ohmi G et al. Use of
dynamic and colored stereogram to measure ste-
reopsis in strabismic patients. Jpn J Ophthalmol
1998; 42: 101–107.
8. Breyer A, Jiang X, Rütsche A, Mojon DS. A new
3D monitor-based random-dot stereotest for chil-
dren. Invest Ophthalmol Vis Sci 2006; 47: 4842–4846.
9. Ma DJ, Yang HK, Hwang JM. Reliability and valid-
ity of an automated computerized visual acuity
and stereoacuity test in children using an interac-
tive video game. Am J Ophthalmol 2013; 156:
195–201.
10. Kim J, Hong JY, Hong K et al. Glasses-free randot
stereotest. J Biomed Opt 2015; 20: 065004.
275