Original Article

Reading eye movements among homonymous hemianopia

Nayan Gupta, Ayisha Atiya, Jameel R Hussaindeen, Selvakumar V Ambika

Purpose: To measure the reading eye movement’s parameters and quantify the oculomotor dysfunction Access this article online
among subjects with homonymous hemianopia compared to age‑matched controls using ReadAlyzer. Website:
Methods: This was a prospective study carried out in the neuro‑optometry clinic of a tertiary eye care https://journals.lww.com/ijo
hospital in South India, from October 2018 to Janurary 2019. Fifty consecutive patients diagnosed with DOI:
homonymous hemianopia were enrolled in the study after obtaining their written informed consent. 10.4103/IJO.IJO_2050_23
Reading eye movements were measured using ReadAlyzer in patients with homonymous hemianopia
and age‑matched controls. Reading eye movement parameters were represented in median [interquartile
range (IQR)]. Results: Subjects with homonymous hemianopia showed increased number of fixations/100 Quick Response Code:
words‑ median (IQR) [175 (80–270)] compared to controls [91 (52.5–127.5)]; increased number of
regressions/100 words [33 (‑13–79)] compared to controls [18 (‑10–46)]; and reduced reading rate
[93 (46–140) words/min] compared to controls [186 (144–228) words/min] (Mann–Whitney U test, P ≤ 0.05).
Similar trend was observed for grade level equivalent 2 (0–4) and regression to fixation ratio 15 (1–29)
compared to controls [grade equivalent 7 (3–11) and regression to fixation ratio 11 (‑2–24)] (Mann–Whitney
U test, P ≤ 0.005). Conclusion: Reading eye movements assessed using ReadAlyzer are found to be
significantly impaired among homonymous hemianopias compared to age‑matched controls (Mann–
Whitney U test, P ≤ 0.05).

Key words: Eye movements, homonymous hemianopia, ReadAlyzer, saccades

Cerebrovascular accident is found to be the most common cause of fixations. About 10%–15% of our reading saccades are
in 70% of patients with homonymous hemianopia (HH), a visual regressive. Toward the end of the line of text, a large right to
field disorder occurring due to post‑chiasmal damage secondary left anti‑saccade eye movement is made close to the beginning
to ischemia in the lateral geniculate body, optic radiations, or the of the next line.
occipital lobe and can manifest as sectoranopias, quadrantanopias,
Eye movements during reading are systematically
or hemianopias, either congruous or incongruous.[1] A global
influenced by visual and lexical characteristics of the text
damage to the neural network of the occulomotor system an
information extracted during a fixation.[3] The region of effective
acquired brain injury leads to reading dysfunction.
processing during reading, the perceptual span, extends about
Reading is one of the complex skills which is disrupted at the three or four characters to the left and up to 15 characters to the
visual and phonological stages post‑cerebrovascular accident. right of fixation. As 1° of visual angle encompasses about three
Impaired gaze holding, saccades, and impaired smooth characters for most normal tests, these values are equivalent to
pursuits affect various levels of visual attention. Patients can 1.3° to the left and 5° to the right of fixation.[3‑6]
also develop hemianopic dyslexia if not rehabilitated at the
There exists functional plasticity of the visual, attentional,
correct time. Reading becomes so laborious that many patients
and oculomotor systems involved in text processing. Poor
give up recreational reading and if reading is an occupation,
eye movements, reduced visual fields, and poor reading
then this becomes an employment risk.
ability together contribute to prolonged fixation, reduced
The accurate, rhythmical, and spontaneously executed saccadic amplitudes, and increased regressive saccades. Foveal
sequences of the eye movements with fixational pauses are processing of fixated words enables word identification.
essential for normal reading.[2] Following spatially distributed Fixation duration is influenced by factors such as word
visual information along with maintaining comprehension frequency, predictability, and age of acquisition.[3]
requires both eyes to follow a typical scan path across the ReadAlyzer is an objective eye movement recording device
text, that is, from left to right and from top to bottom. Plotting that tracks the reading eye movements. Studies have shown
the eye position against time reveals a staircase pattern as
saccadic eye movements regularly alternate with periods
This is an open access journal, and articles are distributed under the terms of
the Creative Commons Attribution‑NonCommercial‑ShareAlike 4.0 License,
Neuro-Optometry Clinic, A Unit of Medical Research Foundation, which allows others to remix, tweak, and build upon the work non‑commercially,
as long as appropriate credit is given and the new creations are licensed under
Sankara Nethralaya, Chennai, Tamil Nadu, India
the identical terms.
Correspondence to: Dr. Nayan Gupta, Research Scholar,
Neuro‑Optometry Clinic, Unit of Medical Research Foundation, For reprints contact: [email protected]
Sankara Nethralaya, Chennai, Tamil Nadu, India. E‑mail: 1nayan.
[email protected] Cite this article as: Gupta N, Atiya A, Hussaindeen JR, Ambika SV. Reading
Received: 01‑Aug‑2023 Revision: 22‑Jul‑2024 eye movements among homonymous hemianopia. Indian J Ophthalmol
2025;73:720-4.
Accepted: 12‑Aug‑2024 Published: 19-Sep-2024

© 2024 Indian Journal of Ophthalmology | Published by Wolters Kluwer - Medknow

May 2025 Gupta, et al.: REM among HH 721

impaired eye movement parameters among various visual Reading parameters included fixations per
field defects.[7,8] Hence, it is necessary to understand the eye 100 words (progressive saccades); regressions per 100
movement parameters among HH patients. words (backward saccades); fixation duration (seconds),
which is the average length of time (in parts of a second) the
Methods eyes paused or fixated; reading rate (words per minute); grade
A prospective comparative study was conducted between level equivalent (GLE), which is the weighted average of the
October 2018 and January 2019 in the neuro‑optometry clinic of grade levels for the subject’s fixations, regressions, and reading
a tertiary eye care center in South India. The study adhered to rate, yielding a combined grade level; and comprehension (%),
which is the percentage of the correct answers. There were
the tenets of the Declaration of Helsinki, and the methodology
also large right‑to‑left oblique saccadic eye movements called
was reviewed and accepted by the Institutional Review Board
saccades in return sweep, which occur when one must shift to
and the Medical Ethics Committee.
the next line of print [Fig. 1].[12]
A total of 50 subjects with HH and 50 controls, considering
Results were analyzed using Statistical Package for Social
a 1:1 ratio between cases and controls, were included in the
Sciences (SPSS) (Version 17.0; SPSS Inc, Chicago, IL, USA).
study. Subjects diagnosed with HH based on the Humphrey
The data set did not follow normality (Shapiro–Wilk test),
visual field test 30‑2 were referred by a neuro‑ophthalmologist.
hence appropriate coding was generated for categorical
Subjects with HH aged between 18 and 50 years with best
variables. Mann–Whitney U test was used to compare the
corrected visual acuity of 20/60 for distance and N6 for near,
values between hemianopic subjects and healthy controls.
who had an onset of 6 months to 12 months were recruited
Spearman’s correlation was used to understand the relationship
for the study. All subjects with multiple disabilities, seizures,
between the variables. The median and interquartile
or any retinal pathology were excluded from the study.
range (IQR) were used to represent the data. The alpha error
Age‑matched controls with normal ocular health were recruited
was set as 5%.
through voluntary participation from those who visited
the clinic during the study duration. A duly signed, written Results
informed consent was obtained from all the study participants.
Results from 50 cases and 50 controls were included for
Testing protocol statistical analysis. The mean age of cases and controls was 46
All subjects received a comprehensive eye examination, and 41 years, respectively, with a male to female ratio of 6:1.
which included history taking, refraction, sensory and motor Twenty‑ (42%) cases were diagnosed to have right HH (RHH)
evaluation, followed by a detailed neuro‑optometry evaluation. and 29 (58%) cases with left HH (LHH).
Reading eye movement assessment In the majority of cases (70%), the etiology was found to
The developmental eye movement (DEM) test was used to be cerebrovascular accident, whereas 20% presented with
understand the horizontal and vertical time, based on which traumatic brain injury and 10% with tumor excision [Fig. 2].
the oculomotor dysfunction (OMD) was classified based on the In the total sample, 44 (88%) out of 50 subjects presented to
ratio score. Reading eye movements were assessed objectively the clinic with the complaint of difficulty in reading, among
using ReadAlyzer (Compevo AB, Stockholm, Sweden). which 17 (80%) cases were RHH cases and 27 (93%) cases were
ReadAlyzer consists of infrared emitters and detectors mounted LHH cases.
in safety goggles. It can determine the eye positions by sensing
DEM testing was done for the cases to grade OMD. The
several infrared reflections from the cornea. The measuring
median (IQR) scores are presented in Table 1. Based on the
speed of the instrument is 60 Hz. The software automatically
vertical score, horizontal score, and ratio score, OMD was
compensates for head movements.[9‑11] The near interpupillary
graded as OMD type I, II, III, and IV. Thirty‑five (70%) subjects
distance was measured with a penlight and an interpupillary
were diagnosed with OMD type IV, eight (16%) with OMD
distance scale. The subjects wore eye movement goggles,
type III, seven (14%) with OMD type II, and two (4%) with
and the near interpupillary distance was adjusted. The test
OMD type I [Fig. 3].
paragraphs were placed at 40 cm with the near correction on.
In addition, proficiency with the English language was asked The results of the reading eye movement assessment using
as they were given to read an English paragraph. ReadAlyzer were compared between cases (HH) and age‑matched
controls [Table 2]. Subjects with HH showed increased number
Eye movements were recorded for cases and controls on the
of fixations/100 words‑ median (IQR) [175 (80–270)] compared to
computer screen connected to the ReadAlyzer, while the subject
controls [91 (52.5–127.5)]. An increased number of regressions/100
read a short English paragraph aloud. The highest grade level words was noted among cases [33 (‑13–79)] compared to
paragraph (Grade 10) was used for measurement. There were
five different passages in Grade 10, which is equivalent to the
font Times New Roman and a size of 12 points. The subject Table 1: Developmental eye movement
read one practice paragraph, following which two trials were
DEM (n=50) Median (IQR)
made with different passages. The second trial was taken as
the final reading to assure a stable baseline measurement.[12] A Vertical score (Test A + Test B) 35 (21.75–48.25)
comprehension test comprising 10 “Yes” or “No” responses was Adjusted horizontal score 43.5 (25.50–61.50)
also administered to confirm the subject’s comprehension. After Ratio score (adjusted horizontal 2.23 (1.96–2.50)
the recording, the system performed an automatic analysis and time/vertical time)
provided the report [Fig 1]. DEM=developmental eye movement, IQR=interquartile range
722 Indian Journal of Ophthalmology Volume 73 Issue 5

Table 2: ReadAlyzer parameters among cases and controls

Eye movement parameters Cases Median (IQR) Controls Median (IQR) P
Fixations/100 words 175 (80–270) 91 (52.5–127.5) P≤0.05
Regressions/100 words ‑33 (‑13–79) ‑18 (‑10–46) P≤0.05
Fixation duration (sec) 0.00 0.45 (0.15–0.75) P≤0.05
Reading speed (words/min) 93 (46–140) 186 (144–228) P≤0.05
Grade equivalent level 2 (0–4) 7 (3–11) P≤0.05
Regression/fixation ratio 15 (1–29) 11 (‑2–24) P>0.05
Comprehension (percentage) 80 (45–115) 92 (64–120) P≤0.05
IQR=interquartile range

a c

b d
Figure 1: (a) Output sheet. (b). Sentence details (fixations and regressions). (c). Graph pattern among RHH. (d) Graph pattern among LHH.
LHH = left homonymous hemianopia, RHH = right homonymous hemianopia

and regression to fixation ratio 11 (‑2–24) (Mann–Whitney U

test, P ≤ 0.05).
The relationship between the number of fixations/100
words and reading rate in HH subjects showed a significantly
strong negative correlation [Fig. 4] (Spearman’s correlation).
To understand the reading eye movements, specifically the
reading speed (words/min), based on the side of visual field
defect, a comparison was done between RHH and LHH
and we found that RHH subjects had a prolonged reading
speed [78 (36–120)] compared to subjects with LHH [102 (68–
Figure 2: Oculomotor dysfunction among homonymous hemianopia 136)] (Mann–Whitney U test, P ≤ 0.05) [Fig. 5].

controls [18 (‑10–46)], and reduced reading rate was noted among A detailed analysis of the reading eye movement among
cases [93 (46–140) words/min] compared to controls [186 (144– subjects with RHH and LHH presented in Table 3 reports
228) words/min] (Mann–Whitney U test, P ≤ 0.05). Similar an increased number of fixations/100 words, regressions/100
trend was observed for GLE [2 (0–4)] and regression to fixation words [28 (2–56)], and reduced reading rate (words/min) of
ratio [15 (1–29)] compared to controls – grade equivalent 7 (3–11) 92 (44–128) [represented in median (IQR), P ≤ 0.05].
May 2025 Gupta, et al.: REM among HH 723

Table 3: ReadAlyzer parameters among RHH and LHH

Eye movement RHH LHH P
parameters Median (IQR) Median (IQR)
Fixations/100 words 74 (56.7–91.5) 108 (90.5–127.5) P≤0.05
Regressions/100 words ‑13 (−9–35) –4 (‑10–2) P>0.05
Fixation duration (sec) 0.00 0.15 (0.05–0.19) P>0.05
LHH=left homonymous hemianopia, IQR=interquartile range, RHH=right
homonymous hemianopia

Discussion
This study focuses on the reading eye movements among
HH patients compared to age‑matched controls. Reading
performance shows a markdown in patients with HH due to
their poor tracking levels. To assess this, the objective method
of reading eye movements was adopted in this study.
DEM testing was used to determine the automaticity and
the reading ability using a vertical and horizontal spatial array
Figure 3: Correlation between the fixations and reading speed of numbers. The ratio score represented a measure to evaluate
the horizontal and vertical time simultaneously. The majority
of the hemianopic subjects in our study population were
diagnosed with type IV OMD. This elicited that the vertical
time, horizontal time, and ratio scores were all abnormal,
which represents the deficiencies in both automaticity and
oculomotor skills, hence disrupting eye movements. This
evaluation provides insight into the functional integrity of
the brain. There were significantly disrupted saccades and
pursuits among hemianopic cases. Reduced visual fields,
poor eye movements, and perceptual difficulties together
contribute to poor reading ability.[13] The accuracy and ability
to make a saccadic eye movement or a smooth pursuit were
hampered.
We evaluated the reading eye movements using the
ReadAlyzer, a detailed measure of increased fixations, and the
regressions were obtained. The prolonged fixations and the
regressions together contribute to the reduced reading speed
among hemianopic patients. Right hemianopic patients show
poor reading speed compared to the left hemianopic patients.
Figure 4: Different etiologies among homonymous hemianopia Similar findings were reported by Kerkhoff G, Trauzettel and
Ong, et al in 1970.[14‑16] The fixations are significantly different
from normal ones in the spatial locations; in addition, the
hemianopic population has a strategic fixation pattern that is
distinct from the healthy controls.[17] The scan paths studied
previously among hemianopes explain the higher number of
fixations of shorter duration and longer saccadic amplitudes
due to the visual field defect, which significantly affects the
eye movements.
Reading eye movement provides essential information
about the visual perceptual field and learning effects, which are
helpful in rehabilitation. Vision training can improve reading
ability by decreasing errors and improving reading speed.
Neuro‑optometric visual rehabilitation helps in improvising
eye and head scanning, thus improving visual attention, and
increasing the number and amplitude of saccades into the
impaired hemifield. Read‑Write, NeuroCoach, and Sanet
Vision Integrator are available software that are used to teach
systematic scanning tactics in more organized and efficient
Figure 5: Reading speed among RHH and LHH. LHH = left search times.[18] Visual rehabilitation and its impact on the
homonymous hemianopia, RHH = right homonymous hemianopia quality of life needs further exploration.
724 Indian Journal of Ophthalmology Volume 73 Issue 5

Conclusion 6. Ciuffreda MA, Ciuffreda KJ, Santos D. Visagraph: Baseline analysis and
procedural guidelines. J Behav Optom 2003;14:60‑4.
Reading eye movement parameters are prolonged in subjects 7. Rayner K. The neural control of eye movements in acquired and
with HH compared to healthy controls. These findings along developmental reading disorders. J. Neurolinguistics 2013;4:97.
with intensive neuro‑optometric evaluation help clinicians 8. Rayner K, McConkie GW, Zola D. Integrating information across eye
movements. Cogn Psychol 1980;2:206‑26.
to evaluate reading deficits and to observe the improvement.
9. Colby, Dirk S, Laukkanen HR, and Yolton RL. Use of the Taylor Visagraph
These findings also serve as baseline measurements to decide
II system to evaluate eye movements made during reading. Optometry
upon neuro‑optometric vision therapy and improve the quality 1998;69:22‑32.
of life among such individuals. 10. Webber A, Wood J, Gole G, Brown B. DEM test, visagraph eye movement
recordings, and reading ability in children. Optom Vis Sci 2011;88:295‑302.
Acknowledgments
11. Reddy AV, Mani R, Selvakumar A, Hussaindeen JR. Reading eye movements
The authors would like to acknowledge the developer of in traumatic brain injury. J Optom 2019;13:155‑62.
ReadAlyzer. 12. Ciuffreda KJ, Han Y, Kapoor N, Ficarra AP. Oculomotor rehabilitation for
reading in acquired brain injury. NeuroRehabilitation 2006;21:9‑21.
Financial support and sponsorship: Nil.
13. Goodwin D. Homonymous hemianopia: Challenges and solutions. Clin
Conflicts of interest: There are no conflicts of interest. Ophthalmol 2014;8:1919‑27.
14. Kerkhoff G, Münßinger U, Haaf E, Eberle‑Strauss G, Stögerer E.
References Rehabilitation of homonymous scotomata in patients with postgeniculate
damage of the visual system: Saccadic compensation training. Restor Neurol
1. Pula JH, Yuen CA. Eyes and stroke: The visual aspects of cerebrovascular Neurosci 1992;4:245‑54.
disease. Stroke Vasc Neurol 2017;2:210‑20. 15. Trauzettel‑Klosinski S. Rehabilitation in TBI‑hemianopia. In: North American
2. Taylor SE. Eye movements in reading: Facts and fallacies. Am Educ Res J Neuro‑Ophthalmology Society Annual Meeting Syllabus, 2009. p. 67‑74.
1965;2:187‑202. 16. Ong YH, Brown MM, Robinson P, Plant GT, Husain M, Leff AP. Read‑Right:
3. Rayner K, Well AD. Effects of contextual constraint on eye movements in A “web app” that improves reading speeds in patients with hemianopia.
reading: A further examination. Psychon Bull Rev 1996;3:504‑9. J Neurol 2012;259:2611‑5.
4. McConkie GW, Rayner K. The span of the effective stimulus during a fixation 17. Pameijer JK. Reading problems in hemianopia. Ophthalmologica
in reading. Percept Psychophys 1975;17:578‑86. 1970;160:322‑5.
5. McDonald SA, Spitsyna G, Shillcock RC, Wise RJ, Leff AP. Patients with 18. Pirozzolo FJ, Rayner K. The neural control of eye movements in acquired
hemianopic alexia adopt an inefficient eye movement strategy when reading and developmental reading disorders. Studies in Neurolinguistics. Academic
text. Brain 2006;129:158‑67. Press; 1979. p. 97‑123.