Content uploaded by Gang Luo
Author content
All content in this area was uploaded by Gang Luo on May 22, 2018
Content may be subject to copyright.
Comparison of Flat and Curved Monitors:
Eyestrain Caused by Intensive Visual Search Task
Gang Luo, Ying Chen, Amy Doherty, Rui Liu, Cong Shi, Shuhang Wang, Hoai Le, Eli Peli
Schepens Eye Research Institute, Mass Eye and Ear, Harvard Medical School, Boston, MA
Abstract
Twenty-seven observers performed intensive visual search tasks
with curved and flat monitors. Based on a 22-item eyestrain
questionnaire, fewer subjects reported eyestrain, difficulty-to-
focus and blurred-vision symptoms with a curved monitor than
with a flat monitor. Fewer subjects had reduced saccade peak
speed following the curved monitor use.
Author Keywords
Eyestrain; visual search; binocular accommodative function.
1. Objective and Background
Symptoms of eyestrain are frequently reported, presumably due to
the large increase in time spent viewing electronic displays. The
various vision problems associated with using computers is
defined by the American Optometric Association as computer
vision syndrome (CVS) [1]. It has been reported that up to 90% of
computer users experience CVS symptoms [2]. The most common
symptoms of CVS are eyestrain, headaches, ocular discomfort,
dry eye, diplopia, or blurred vision [1, 3].
Due to the visual discomfort, CVS has an economic impact.
Productivity is reduced as a result of increased errors, more
frequent breaks, and workers compensation costs from work
related musculoskeletal injuries associated with computer use.
Minimizing symptoms that reduce efficiency will provide a
financial benefit [1].
CVS symptoms may be reduced by treating uncorrected refractive
error, accommodative/vergence anomalies, and dry eye [1];
allowing breaks; and optimizing ergonomic positioning [3]. Better
display designs can potentially be another approach to address
CVS. Curved TVs and monitors are available on the market, and
have been claimed to reduce visual fatigue [4]. However, there
has not been sufficient evidence to support such claims. The
purpose of this study was to determine whether symptoms of CVS
could be lessened with a curved monitor than a flat monitor. We
also measured the effects of the monitors on visual functions.
2. Methods
Participants performed an intense visual search task on curved and
flat monitors. A questionnaire, eye movement recordings, and
convergence and accommodative function tests were
administrated before and after the search task. Each participant
was tested with one of the monitors in the first visit and the other
monitor in second visit, at least one day later.
Monitors: A Samsung 34” curved monitor (S34E790C, radius 3
meters) and a Samsung 34” flat monitor were used for this study.
The brightness of the monitors was 280 cd/m2, and 260 cd/m2,
respectively. The resolution for both monitors is 3440×1440.
Participants sat 40cm from the monitors, and the monitors
spanned about 92° horizontally.
Visual Search Tasks: The participants performed 2 tasks. In
the first task, they searched for a given target within a real world
image using a mouse cursor to select the target. There were a total
of 300 images including faces, indoor scenes, and object
collections. The second task involved searching for a target image
among multi-page search results returned by a Google image
search using a keyword, such as “zebra” (Figure 1). In total,
participants searched for 12 targets among 7089 returned images.
To minimize the effect of memory, each participant was given
different targets in the 2 visits. The orders of monitor and search
targets were counter-balanced across participants. Search time
was recorded to quantify search performance.
Figure 1. On the left is 1 of the 300 search images.
Assigned search target is shown in the upper left corner. On
the right is an example of the multi-page image array
returned by Google when given the keyword, “zebra”.
Targets were given in a different window not shown here.
Questionnaires: Participants were administrated a discomfort
questionnaire immediately before and after the visual search
experiment. The questionnaire included 22 items: General
discomfort, Overall fatigue, Headache, Eyestrain/ache, Difficulty
focusing (eye), Salivation increasing, Sweating, Nausea,
Difficulty concentrating, Fullness of head, Blurred vision,
Dizziness with eyes open, Dizziness with eyes close, Vertigo,
Stomach awareness, Burping, Eyes feel tired, Pulling feeling
around eye, Double vision, See objects on screen move/jump,
Eyes feel itchy, Eyes feel scratchy. These items were categorized
into 4 categories: asthenopic (eye strain), ocular-surface related,
visual, and motion sickness.
Binocular Accommodation Tests: Binocular
accommodative functions were measured in each visit before the
pre-search questionnaire and repeated after the post-search
questionnaire. The assessment included 15 clinical tests for:
phoria, convergence, accommodative functions etc. Based on
these tests, diagnosis for 6 possible binocular deficiencies were
assigned [5]: Convergence insufficiency (CI), Convergence
excess (CE), Fusional version dysfunction (FVD),
Accommodative insufficiency (AI), Accommodative excess (AE),
and Accommodative infacility (A_infacility). Changes in these
diagnoses were used in data analysis..
Eye Movement Recording: Eye movements of each
participant were measured using a 500Hz eye tracker (EyeLink
II), before and after the visual search task. During the eye
movement recording session, participants were instructed to make
20° leftward (n=20) and downward (n=20) saccades from fixation
to a target. Saccade latency, duration, and peak speed were
measured. In an additional free viewing task, the eye tracker
counted the number of blinks as the participants viewed 6 real-
world images for 1 minute in total.
Participants: Twenty-seven normally sighted participants (age
18 to 35) completed the study. Their near visual acuities (with
habitual correction when needed) were 20/20 or better.
66-2 / G. Luo
SID 2016 DIGEST • 903
ISSN 0097-966X/16/4702-0903-$1.00 © 2016 SID
3. Results
Visual Search: On average, the visual search task took about
55 minutes to complete. Overall, there was no significant
difference in search performance between curved and flat
monitors (p=0.89, paired t-test).
Eyestrain Questionnaire: After visual search, 26 out of 27
participants reported at least one discomfort symptom for either or
both monitors. Comparing participants’ reports before and after
the visual search task, overall most of the discomfort symptoms
increased (20 out of 22) and only one decreased. The numbers of
increased cases for all symptoms are reported in Figure 2.
According to a proportion test, a difference of 3 cases (dashed line
in Figure 2) is significantly different from no change. As can be
seen, 9 of the 22 symptoms were above this level: general
discomfort, overall fatigue, eyestrain, difficulty focusing,
difficulty concentrating, blurred vision, eyes feel tired, pulling
feeling around eyes, and eyes feel scratchy. While for the other
symptoms were reported after the search task the increase was not
statistically significant.
Figure 1. Out of 22 symptoms, 9 symptoms significantly
increased (beyond the dashed line, which indicates
significance) after visual search for either or both monitors.
Based on the number of symptoms each participant reported pre-
and post-search, a paired t-test did not find a statistically
significant difference between the two monitors. Figure 3 shows
the number of participants who reported worsening symptoms
after search for the 9 symptoms mentioned above. The proportion
test found that significantly more participants reported some
worsening symptoms with the flat monitor than the curved
monitor (indicated by star markers in Figure 3): eyestrain (4 more,
p=0.019), difficulty to focus (4 more, p=0.019), blurred vision (3
more, p=0.037), and eyes feel tired (3 more, p=0.037). On the
other hand, overall fatigue was reported by more subjects for the
curved monitor than the flat monitor (3 more, p=0.037, indicated
by a cross marker in Figure 3).
Figure 2. The 4 star markers indicate symptoms that
worsened in significantly more participants with the flat
monitor than the curved one. The cross marker indicates
that overall fatigue (not eye specific) symptom worsened in
significantly more subjects with curved than the flat monitor.
Eye Movements: We did not find any significant difference
between flat and curved monitors in saccadic latency, saccadic
duration or number of blinks. The only significant difference
found was in peak speed of horizontal saccades: more participants
had slower horizontal saccadic peak speed after the search task
with the flat monitor (p=0.004).
Figure 3. Among the 9 eye movement parameters analyzed
including: latency, speed and blinks, only horizontal
saccade peak speed was significantly different between the
two monitors. More participants had slower horizontal
saccadic peak speed after visual search with the flat than
with the curved monitor.
-2 0246810
General discomfort
Overall fatigue
Headache
Eyestrain/
ache
Difficulty
focusing (eye)
Salivation
increasing
Sweating
Nausea
Difficulty
concentrating
Fullness
of head
Blurred
vision
Dizziness with
eyes open
Dizziness with
eyes closed
Vertigo
Stomach
awareness
Burping
Eyes feel tired
Pulling feeling
around eye
Double
vision
Words on screen
move/jump
Eyes feel
itchy
Eyes feel
scratchy
# of increased cases after visual search
Flat Curved
0 5 10 15
General
Discomfort
Overall fatigue
Eyestrain/ac he
Difficulty focus ing (eye)
Difficulty
concentrating
Blurred
vision
Eyes feel
tired
Pulling feeling
around eye
Eyes feel
scratchy
# of subjects who reported worsening symptoms
Flat
Curved
66-2 / G. Luo
904 • SID 2016 DIGEST
Convergence and Accommodative Functions: The
convergence and accommodative function testing examined
participants’ convergence and accommodation capabilities or
deficiencies. These tests not only check convergence and
accommodation separately, but also check combined (binocular)
convergence and accommodation performance. In order to clearly
see a target, the eyes need to turn towards the object and focus
properly for its distance.
Among the 26 participants who reported symptoms after the
search task for either or both monitors, 14 were diagnosed with at
least one binocular accommodative dysfunction either pre- or
post-search, and the other 12 were not found to have any
dysfunction. The only one participant who did not report any
symptom was not found to have any dysfunction.
Figure 5 shows the number of participants who were diagnosed
with binocular accommodative dysfunctions and also reported
eyestrain, difficulty focusing, or blurred vision symptoms after the
search task. There seems to be a pattern that they were more likely
to have accommodative dysfunctions, namely, AI, AE and
A_infacility. Interestingly, for those 3 dysfunctions, the curved
monitor seemed to be associated with fewer cases than the flat
monitor, but the difference did not reach statistical significance.
Figure 4. Accommodative dysfunctions (AI, AE, A_infacility)
might be more common in participants who reported post-
search eyestrain, difficulty focusing and blurred vision than
convergence dysfunctions. The curved monitor seemed to
be associated with fewer cases for these 3 dysfunctions
than the flat monitor, but the difference was not significant.
4. Impact
Our preliminary results showed that the visual search task used in
this study was intense enough to cause notable eye discomfort
within a short period of time (nearly one hour on average) for
almost all participants. This is consistent with numerous studies
and user complaints about CVS. Since computer users can easily
spend many hours a day engaging in intense tasks on electronic
displays, the prevalence of eye discomfort could be even greater.
Curved monitors have been suggested as one solution to address
the problem. Anecdotal evidence based on very short exposure
(0.5-2 minutes) might suggest that users felt “more comfortable”
with curved monitors than flat ones [6]. However, it is unlikely
that observers experienced CVS symptoms following such a short
exposure. Their reports probably reflect some visual appeal of
novel curved monitors. Our study evaluated the curved monitor
based on commonly used clinical tests and symptoms.
We did not find significant differences in eye discomfort
symptoms between flat and curved monitors based on comparison
of group means or medians. Based on the proportion test,
however, the questionnaire data did show that fewer observers
reported worsening eyestrain, difficulty focusing, blurred vision,
and eye tiredness symptoms following the search task with the
curved monitor than with the flat one. This could suggest that the
curved monitor might be more comfortable for some display
users.
We did not find evidence from eye movements and convergence-
accommodation tests to suggest what might account for fewer
complaints for the curved monitor, and what types of users might
benefit from curved monitors if the benefits are real. A possibly
higher percentage of accommodative dysfunctions (AI, AE,
A_infacility) among reporters of symptoms could suggest that
accommodation may be a factor for discomfort in our study. It
should be kept in mind that the difference in viewing distance at
the monitor edge between the two monitors was very small
(accounting only for about 0.1diopter in accommodation demand).
It is not clear how such a small difference might play a role. Also,
a fair number of observers reported symptoms but did not have
any dysfunction. Nevertheless, these findings warrant further
investigation.
It is unclear why the curved monitor was associated with more
overall fatigue, while some eye-specific symptoms were less at
the same time. This was inconsistent with the “general
discomfort” question, and no other items in our questionnaire
might suggest overall fatigue with the curved monitor would be
high.
It has been previously reported that greater eyestrain and more
reading errors were associated with electronic displays, when
compared to paper copies [7]. When convergence and
accommodative insufficiencies, which were found to be
associated with eyestrain symptoms [8], were successfully treated
in school-age children, their academic performance improved [9].
In this study, we did not find a difference in search performance
between the two monitors. This may be because the difference in
symptoms was not large enough to impact visual search during
the short experiment.
Taken together, the effect of a curved monitor on CVS and task
performance needs to be evaluated in larger-scale, longer-term
follow up studies.
5. Acknowledgement
Samsung Ltd (Korea) provided the flat and curved monitors tested
in this study and supported the research with an unrestricted gift.
6. References
[1] Rosenfield M. Computer vision syndrome: a review of ocular
causes and potential treatments. Ophthalmic Physiol Opt.
2011; 31(5): 502-15.
[2] Thomson WD. Eye problems and visual display terminals--
the facts and the fallacies. Ophthalmic Physiol Opt. 1998;
18(2): 111-9.
[3] Blehm C, Vishnu S, Khattak A, Mitra S, & Yee RW.
Computer vision syndrome: a review. Surv Ophthalmol.
2005; 50(3): 253-62.
[4] Samsung Curved UHD TV: A review of an optometrist.
2014 [Nov 19, 2015]; Available from:
http://www.philstar.com/gadgets/2014/10/07/1377489/samsu
ng-curved-uhd-tv-review-optometrist.
[5] Master CL, Scheiman M, Gallaway M, et al. Vision
0
1
2
3
4
5
6
CI CE FVD AI AE A_Infacility
# of cases
Flat
Curved
66-2 / G. Luo
SID 2016 DIGEST • 905
Diagnoses Are Common After Concussion in Adolescents.
Clin Pediatr (Phila). 2015; 2015(7).
[6] Herrmann H-J. Experimental subjective user test -
Comparison of a flat and curved 34 inch: TUV Rherinland
Internal report. 2015.
[7] Wilkinson RT & Robinshaw HM. Proof-reading: VDU and
paper text compared for speed, accuracy and fatigue.
Behaviour & Information Technology. 1987; 6(2): 125-133.
[8] Borsting E, Rouse MW, Deland PN, et al. Association of
symptoms and convergence and accommodative
insufficiency in school-age children. Optometry. 2003;
74(1): 25-34.
[9] Borsting E, Mitchell GL, Kulp MT, et al. Improvement in
academic behaviors after successful treatment of
convergence insufficiency. Optom Vis Sci. 2012; 89(1): 12-
18
66-2 / G. Luo
906 • SID 2016 DIGEST
