Content uploaded by Erica Watkins
Author content
All content in this area was uploaded by Erica Watkins on Jan 21, 2014
Content may be subject to copyright.
761
JRRD
JRRD
Volume 43, Number 6, Pages 761–770
September/October 2006
Journal of Rehabilitation Research & Development
Effects of preferred retinal locus placement on text navigation and
development of advantageous trained retinal locus
Gale R. Watson, MEd, CLVT;
1
*
Ronald A. Schuchard, PhD;
1
William R. De l’Aune, PhD;
2
Erica Watkins, BA
1
1
Blind Rehabilitation Service, Rehabilitation Research and Development Center of Excellence for Aging Veterans with
Vision Loss, Atlanta Department of Veterans Affairs Medical Center, Decatur, GA;
2
Atlanta Research and Education
Foundation, Atlanta, GA
Abstract—Sixty readers were evaluated for visual function and
text-navigation ability. The visual field and preferred retinal
locus (PRL) were measured with a scanning laser ophthalmo-
scope (SLO). We found significant differences in text-navigation
ability based on scotoma and PRL placement. Readers with a
PRL to the left of or above a scotoma had significantly less text-
navigation abilities. Readers with a PRL to the left of a scotoma
tended to misread words with similar beginnings and omit the
last word on a line. Readers with a PRL above a scotoma tended
to skip a line or reread the same line twice. In a follow-up study,
seven subjects with a nonadvantageous PRL quickly developed a
trained retinal locus (TRL) during instruction with an SLO.
Although the readers developed the TRL in about 15 minutes,
they read slower with the TRL than the PRL. This TRL research
provides promising pilot data.
Key words: low vision, macular scotoma, preferred retinal
locus, reading, rehabilitation, scanning laser ophthalmoscope,
text navigation, trained retinal locus, vision rehabilitation,
visual impairment.
INTRODUCTION
The term “preferred retinal locus” (PRL) describes a
retinal area that acts as a pseudofovea for visual tasks
when a central macular scotoma affects visual perform-
ance. Individuals with macular disease who are unable to
develop new skills with a PRL cannot optimally use low-
vision devices or effectively use their remaining vision
for reading [1]. The individual’s visual system may not
develop the best retinal area for reading with the PRL; for
example, when reading, part of the word may be placed
in the scotoma. Neither the individual nor the clinician
may be aware of this misplacement except possibly
through the clinician’s observation of reading errors
when the individual is reading aloud. Even individuals
with a paracentral scotoma may have a scotoma that is
next to the fovea and impairs reading. Low-vision reha-
bilitation effectively helps individuals with macular loss
regain their reading abilities.
Vision Function
The prevalence of persons with low vision and macular
scotomas is considerably larger than previously suspected.
Fletcher et al. found a macular scotoma prevalence of
Abbreviations: ETDRS = Early Treatment Diabetic Retinopa-
thy Study, logMAR = logarithm of minimum angle of resolu-
tion, MNREAD = Minnesota Low-Vision Reading (Acuity
Charts), OR = odds ratio, PRL = preferred retinal locus, SLO =
scanning laser ophthalmoscope, TRL = trained retinal locus,
VSRT = (Pepper) Visual Skills for Reading Test.
*
Address all correspondence to Gale R. Watson, MEd,
CLVT; Blind Rehabilitation Service, Rehabilitation
Research and Development Center of Excellence for Aging
Veterans with Vision Loss, Atlanta Department of Veter-
ans Affairs Medical Center, 151-R, 1670 Clairmont Road,
Decatur, GA 30033; 404-321-6111, ext 6789; fax: 404-728-
4837. Email: Gale.Watson@va.gov
DOI: 10.1682/JRRD.2005.07.0120
762
JRRD, Volume 43, Number 6, 2006
about 83 percent in their typical low-vision rehabilitation
service [2]. Less than half of these patients with low vision
were diagnosed by their referring ophthalmologist or low-
vision specialist as having macular scotomas.
Comprehensive rehabilitation can dramatically help
persons with visual impairments achieve full lifestyles
[3–4]. However, research has indicated that accurate, pre-
cise measurement of the scotoma area relative to the PRL
for visual tasks may be important, if not critical, for
visual performance tasks such as reading [5–9]. Macular
scotomas also impair visual functions such as contrast
sensitivity [10], contrast discrimination [11], stereo-
scopic depth perception [12], and fixation precision/sta-
bility [5,13].
Even when persons are aware of a scotoma (they see
a blurry, hazy gray area), they are seldom aware of the
true extent of the scotoma [14]. Persons typically
describe new scotomas resulting from laser treatments as
“large black spots” or “boulders,” but over a few months,
these spots fade and can no longer be localized without
perimetric testing. In an eye with a central scotoma
affecting the entire fovea, one or more eccentric PRLs
naturally and reliably develop [5–7,15–17]. A shift of the
oculocentric reference from the fovea to an eccentric
PRL is possible in persons with bilateral macular disease
[11,16].
Fletcher et al. used the scanning laser ophthalmo-
scope (SLO) to study the characteristics of macular
scotomas in 742 eyes [2]. The retinal locations of fixa-
tion, foveal PRL or eccentric PRL, were recorded and
graded. Dense scotomas were mapped with a 12 arc-
minute square target with a 50,000 Td retinal illumi-
nance. Relative scotomas were also mapped with the use
of targets at the retinal illuminance level that was deter-
mined to represent the threshold sensitivity of the fovea
or eccentric PRL. With this testing method, fixation
could be monitored. Even in persons with very unsteady
fixation, macular scotomas were identified in relationship
to the anatomy of the macula. The size, location, and
density of scotomas cannot be inferred consistently from
the ophthalmoscopic appearance of the macula. Figure 1
depicts an SLO retinal map.
In Fletcher et al.’s study [2], 83 percent (616/742) of
eyes had dense scotomatous areas present in the central
visual field, 14 percent (101/742) had scotomas of less
than 5° in diameter, and 69 percent (515/742) had scoto-
mas of greater than 5° in diameter. Scotoma shape varied
widely, from round scotomas centered on a nonfunction-
ing fovea to ring scotomas surrounding a functioning
fovea to highly complex amoeboid shapes. The PRL (for
fixation) was frequently near a dense scotoma. The rela-
tionship of the PRL to surrounding dense scotomas in the
visual field was categorized by how many of the four bor-
ders (superior, inferior, right, and left) around the PRL
were dense scotomas. (Note: a visual field dense scotoma
to the left of the PRL corresponds to an area of nonfunc-
tioning retina anatomically to the right of the PRL.) The
total number of dense-scotoma borders around the PRL
were (1) no borders: 22 percent, (2) one border: 46 per-
cent, (3) two borders: 14 percent, (4) three borders: 7 per-
cent, and (5) four borders (ring scotoma): 11 percent.
Fletcher and Schuchard have developed and used a
uniform PRL scoring system using an SLO to determine
measurable qualities of the PRL [18]. The SLO directly
determines the retinal location of visual stimuli with
respect to retinal image characteristics. Their retinal-
location scoring mechanism includes PRL size, fixation
ability, saccade ability, and pursuit ability. PRL size is
recorded in degrees. PRL fixation, saccade, and pursuit
abilities are scored from 0 to 4; 0 indicates no ability and
Figure 1.
Scanning laser ophthalmoscope (SLO) retinal map of scotoma shape
and size (outlined in green) and documented position and size o
f
nonfoveal preferred retinal locus (PRL) (outlined in red). Gree
n
outline of scotoma border is artistic rendering of dense scotoma (DS).
Green area encompasses nonseeing retinal points as designated b
y
manual hybrid perimetry with fixation monitoring in SLO. Note:
anatomic area of nonfunctioning retina anatomically to right of PRL
corresponds to visual field dense scotoma to left of PRL.
763
WATSON et al. Trained retinal locus for text navigation
4 indicates abilities that approximate those of a person
using foveal vision.
Reading with Preferred Retinal Locus
Programs that teach the visual skills used in reading
have been developed and widely used [19–21]. Previous
research has shown that training individuals with macular
degeneration in visual skills and use of low-vision
devices effectively increases reading accuracy [22–23],
rate [24–25], and comprehension [26–27].
Two oral reading assessments, the Pepper Visual
Skills for Reading Test (VSRT) and the Minnesota Low-
Vision Reading (MNREAD) Acuity Charts, have been
developed and standardized for persons with low vision
and are commonly used research measures. These assess-
ments were developed for quick evaluation of reading
ability, since persons with low vision are likely to read
more slowly and fatigue more quickly than persons with-
out visual impairments.
The Pepper VSRT is a highly reliable assessment of
text-navigation ability [28] and is used in low-vision clin-
ics for diagnosis and remediation of reading problems in
persons with macular degeneration [23,29]. The test can
be administered in approximately 10 minutes and can
diagnose specific reading errors of persons with macular
scotomas. Rate and accuracy scores and diagnoses of spe-
cific errors can be derived. This test was highly valid when
compared with the Gray Oral Reading Tests, a graded
reading evaluation of continuous text [30]. Classified error
codes were developed for providing an accuracy score.
Errors include omissions, insertions, misidentifications,
connecting words, separating words, and skipping lines.
The MNREAD Acuity Charts simply and quickly esti-
mate reading acuity, maximum reading rate, and critical
print size (smallest print size at which fluent reading can be
maintained) of persons with low vision [31]. Unlike the
Pepper VSRT, which evaluates visual skills in reading, the
MNREAD Acuity Charts assess reading acuity and rate in
a controlled manner and derive critical print size for most
fluent reading. The MNREAD Acuity Charts have high
test-retest reliability and are highly correlated with silent
reading ability (as measured by comprehension).
Fletcher et al. documented the characteristics of
macular scotomas relative to the PRL by relating SLO
macular perimetry to reading performance [1]. The PRL
ability (fixation stability, pursuit ability, and saccadic
ability) was rated with the 0 to 4 scoring system just
described [18]. This SLO test, a test of visual acuity
(Early Treatment Diabetic Retinopathy Study [ETDRS]
chart), and the MNREAD Acuity Charts [31] were
administered to 40 patients during the initial visit before
rehabilitation intervention. The MNREAD Acuity Charts
were administered as recommended. As in a previous
study [9], Snellen acuity accounted for only 19 percent of
the variability in reading rate, a scotoma above and/or
below the PRL accounted for only 12 percent, and fixa-
tion stability accounted for 27 percent. However, other
studied clinical variables had much stronger relationships
to reading rate than previously studied parameters. Sac-
cadic ability accounted for 38 percent of the variability in
reading rate, pursuit ability accounted for 37 percent, and
a scotoma to the right and/or left of the PRL accounted
for 40 percent. Subjects with scotomas to both the right
and left of the PRL or subjects with scotomas only to the
right had significantly reduced reading rates compared
with subjects with scotomas to the left or no scotomas on
either side of the PRL.
Trained Retinal Locus
Nilsson reported that, since the 1970s, a low-vision
service in Sweden has been instructing patients in the use
of a trained retinal locus (TRL) in a retinal area that is
“more advantageous to reading,” defined as above or
below an absolute central scotoma as opposed to the left
or right of the scotoma [25]. In Nilsson et al.’s feasibility
study, they trained a retinal locus in a scotomatous eye;
each subject’s other eye was unimpaired [32]. Their
results showed that subjects who did not develop a PRL
in the scotomatous eye (because they had one unimpaired
eye) developed a TRL in the scotomatous eye and had
significantly improved reading rates with the scotoma-
tous eye following training.
Nilsson et al. further reported on 20 persons with
age-related macular degeneration and central scotomas in
both eyes who were trained to develop a TRL [33]. In 18
subjects, the study eye was the lesser-seeing eye, since
the researchers wanted to avoid using subjects who had
already been trained to use eccentric viewing methods.
For the remaining two subjects, one had equal acuity in
both eyes and one had a rapidly progressing lesion that
prevented training of the lesser-seeing eye. None of their
subjects had been exposed to rehabilitation, had access to
magnification devices, or had been currently reading by
using their vision. Eleven subjects had a PRL to the left
of the lesion, six had a PRL to the upper-left of the lesion,
two had a PRL to the lower-left of the lesion, and one had
764
JRRD, Volume 43, Number 6, 2006
a PRL to the right of the lesion. The subjects were trained
via SLO. Subjects were taught to look at a letter in the
center of crossed lines; the letter was slowly moved so
that the subject saw the letter projected in a “trained reti-
nal locus” area, while the crossed lines remained in the
scotoma. The subject was told to continue looking at the
letter; the crossed lines remained as a reference point and
assured that the subject understood how to continue
viewing with the TRL.
Nilsson et al. reported that they could use the SLO to
constantly check the subject’s fixation and correct it if
the subject lost the fixation from the trained area [33].
After training with the SLO, subjects were taught to use
the TRL with high-plus magnifying spectacles [33].
Exercises with similar fixation lines were used [19].
Rather than use fixation and saccadic eye movements, the
subjects were taught to scroll printed reading exercises so
that the words moved into the TRL field of view.
The authors reported that 18 of the 20 subjects devel-
oped a TRL [33]. Of the remaining two subjects, one was
unable to use a TRL after 5 hours of training and one
refused to use the magnifying spectacles. Of the 18 sub-
jects who developed a TRL, 12 were trained to use a TRL
below the visual-field scotoma and 6 were taught to use a
TRL above the visual-field scotoma. Subjects received an
average of 5.4 hours of training, with 1 week of home-
work between each training session. Mean reading rate
prior to training was 9.7 words/minute; after training,
mean reading rate was 68.3 words/minute. The authors
concluded that their subjects were able to be trained to
use a TRL and following rehabilitation, study eyes
showed significantly increased reading rates with the
TRL. However, the authors did not train the subjects’
naturally occurring PRL, so we do not know what ability
the study eyes might have acquired had their PRL been
trained. In earlier studies [25], these researchers achieved
dramatic and similar increases in reading rates (from 0 to
75.5 words/minute) through training for approximately
the same rehabilitation time period but without direct
monitoring of eye movements. One must question, then,
whether training is the most important aspect of reading
rate improvement, regardless of whether the reader uses a
PRL or TRL.
We do not know what ability Nilsson et al.’s subjects
might have gained if their better-seeing eyes had been
trained. The better-seeing eye is generally chosen for
rehabilitation in low-vision clinics unless a patient can
read with a binocular low-vision device, such as a closed-
circuit television system or a hand/stand magnifier.
Nilsson et al. also concluded that a PRL to the left or
right of a scotoma was not advantageous and a PRL
above or below a scotoma was better for maximal read-
ing ability [33]. Unfortunately, these conclusions are
based on clinical heuristics and laboratory research on
subjects without visual impairments who were provided
with artificial scotomas [34–35]. In a study of readers
with low vision with naturally occurring central scoto-
mas, Fletcher et al. did not find an advantage of any par-
ticular PRL-scotoma position for reading rate [1].
Our study investigated whether PRL and scotoma
placement were related to text-navigation ability in sub-
jects with low vision. We also investigated whether a
TRL could be developed in the better-seeing eyes of sub-
jects with low vision who were long-term PRL users.
EXPERIMENT 1
Methods
We recruited 60 subjects who had visual impairments
due to macular diseases (absolute scotoma in the visual
field as measured by the SLO perimetry), 20/800 or bet-
ter acuity in the better-seeing eye, and interest in reading
with low-vision devices. We obtained the subjects’
informed consent using the procedures required by the
Emory University Human Investigations Committee.
Following informed consent, subjects were evaluated
without low-vision devices by the following procedures:
1. ETDRS Distance Visual Acuity Charts: Distance acuity.
2. MNREAD Acuity Charts: Reading acuity, reading
rate, and critical print size.
3. Pepper VSRT (with low-vision devices): Text naviga-
tion ability, reading accuracy, and reading rate.
4. SLO testing: Macular perimetry as described by Sun-
ness et al. [36] and PRL characteristics, including fixa-
tion stability, as described by Fletcher and Schuchard
[18]. The confocal SLO with graphics capabilities
allows an examiner to directly determine the retinal
location of visual stimuli on the retinal image in real
time. The SLO with graphics capabilities continuously
obtains retinal images with a nearly invisible infrared
laser (780 nm) and scans graphics (e.g., lines, letters,
and words) on the retina with a modulated visible red-
light laser (633 nm). Thus, the subject observes the
stimuli, which the examiner sees directly on the
765
WATSON et al. Trained retinal locus for text navigation
subject’s retina. The retinal illuminance of the stimuli is
adjustable by 256 steps within the visible red-light laser
range (50–50,000 Td). The SLO provides a 32° × 22°
image of the retina with a minimum resolution of
approximately 4 arcminutes for measurement of the reti-
nal areas and positioning of targets. Macular scotomas
can be mapped on the SLO with kinetic, static, or hybrid
testing techniques.
5. Binocular perception test: Evaluation of dominant
PRL. A computer system displayed stereo pairs
sequentially on a monitor at 120 Hz. The electronic
eyewear, liquid crystal shutters synchronized with the
monitor, determined which eye saw each frame of the
stereo pair. When the left image was on the video
screen, for example, the left shutter opened while the
right shutter closed. The subject was told to fixate a 1°
star (seen as a cross by the right eye, an X by the left
eye, and a star binocularly), then report which target
was seen. If the subject reported a cross or an X, then
he or she was not seeing the visual information from
the PRL/fovea of both eyes during binocular viewing.
Results
Table 1 shows subjects’ ages; scores on vision func-
tion, reading, and visual acuity measures; and PRL per-
formance.
We determined fixation size for the PRL by measur-
ing fixation eye movements for 20 seconds; the diameter
that we recorded encompassed the area of fixation eye
movements [18]. We reported PRL performance using
the previously described method of rating from 0 to 4 the
subject’s use of the PRL in the SLO for fixation, pursuit,
and saccade abilities [18]; 0 indicated no ability and
4 indicated that the subject’s off-foveal PRL ability was
similar to that for a foveal PRL.
Rate, accuracy, and error scores on the Pepper VSRT
were recorded and matched to the PRL recorded on the
SLO. We coded text-navigation errors on the Pepper
VSRT as left directional errors, right directional errors,
or below directional errors. Left directional errors were
missing or confusing the beginning of a word or missing
the first word on a line. Right directional errors were
missing or confusing the ending of a word or missing the
last word on a line. Below directional errors were skip-
ping a line or reading the same line twice. In this article,
we report results for text-navigation errors only; other
results will be published later.
We scored scotomas by placing a matrix over the
SLO field plot. The matrix was a transparency with con-
centric circles, a center fixation dot, and eight radial lines
from the center fixation dot. The fixation dot was cen-
tered over the PRL; each of the radials was examined. By
using this overlay, we determined the presence or
absence of a dense scotoma on each radial. Scotoma con-
figuration may have included several radials. Figure 2
depictions the matrix transparency over an SLO map.
For this analysis, we used dense scotomas in the eye
with the dominant PRL; if no dominant PRL existed, we
used dense scotomas in the subject’s preferred eye.
Table 1.
Ages and scores on vision function and reading measures for
60 subjects with absolute scotoma in visual field in Experiment 1.
Measure Mean ± SD Range
Age (yr) 74.5 ± 10.6
23
Acuity (logMAR) 0.8 ± 0.3 0.1–1.5
PRL Size (° fixation) 3.9 ± 2.1 1.5–10
Fixation Ability Score 3.0 ± 0.7 0–4
Pursuit Score 2.5 ± 0.7 0–4
Saccade Score 2.3 ± 0.9 0–4
MNREAD Critical Print Size (logMAR) 1.1 ± 0.4 0.2–1.6
MNREAD Acuity (logMAR) 0.9 ± 0.4 0.1–1.7
MNREAD Rate (words/min) 78.1 ± 52.4 4–210
Pepper VSRT Rate (words/min) 29.0 ± 17.6 3.5–81.8
Pepper VSRT Accuracy (%) 83.8 ± 19.4 20.8–99.3
logMAR = logarithm of minimum angle of resolution, MNREAD = Minnesota
Low-Vision Reading (Acuity Charts), PRL = preferred retinal locus, SD =
standard deviation, VSRT = Visual Skills for Reading Test.
Figure 2.
Transparency overlay for scoring scotomas. Matrix was transparency
with concentric circles, center fixation dot (outlined in red), and eight
radial lines from center fixation dot. Scotoma is outlined in green.
766
JRRD, Volume 43, Number 6, 2006
We coded subjects’ reading errors on the Pepper
VSRT as specific directional errors; “right” errors were
mistaking the ending of a word or omitting the last word
on a line, “left” errors were mistaking the beginning of a
word or omitting the first word on a line, and “below”
errors were skipping or rereading a line.
We performed a simple analysis of the presence or
absence of directional reading errors (right, left, below)
from the Pepper VSRT and the presence or absence of
scotomas in each radial from the PRL. In each case, we
constructed a fourfold table and performed a chi-square
analysis. A total of 24 fourfold tables were constructed
and analyzed. For each table, we evaluated the signifi-
cance of the resultant chi-square
value, and if it was sta-
tistically significant, we computed and reported an odds
ratio (OR). We obtained the following results:
• A dense scotoma to the right of the PRL (radial 3)
was associated with more right errors (
χ
2
= 4.1, p =
0.05). The OR indicated that a scotoma in this radial
was associated with 2.033 times the likelihood of
right errors. A dense scotoma to the left of the PRL
(radial 7) was associated with one-third fewer right
errors (OR = 0.348;
χ
2
= 3.9, p = 0.048).
• Left errors were not associated with any particular
PRL configuration.
• A dense scotoma below the PRL (radial 1) was asso-
ciated with below errors (
χ
2
= 4.2, p = 0.04). The OR
indicated that a scotoma in this radial was associated
with 3.368 times the likelihood of below errors. A
dense scotoma below and to the right of the PRL
(radial 2) was also associated with below errors (
χ
2
=
5.9, p = 0.015). The OR indicated that a scotoma on
this radial was associated with 4.474 times the likeli-
hood of below errors.
EXPERIMENT 2
Methods
In the second experiment, we recruited subjects
whose naturally occurring PRL was to the left of their
scotoma and trained them to develop a TRL that was
below their scotoma. We chose the PRL location based
on our and others’ research that found this location to be
disadvantageous for reading, and we chose the TRL loca-
tion based on our and others’ research that found this
location to be advantageous for reading. To locate sub-
jects whose PRL was to the left of their scotoma, we
screened the records of the 60 subjects from Experiment
1 for whom we already had documented SLO visual
fields. Of these 60 subjects, 13 fit our criterion but 3 had
either moved or were ill. The remaining 10 potential sub-
jects provided informed consent according to the require-
ments of the Emory University Human Investigations
Committee and were pretested for acuity and central
visual-field perimetry via SLO. Three subjects had
changed PRL locations (two had continued macular loss
and one deliberately changed his PRL, as explained in
the “Discussion” section). This resulted in seven subjects
who agreed to participate. Because of the very small sam-
ple size, this experiment was exploratory.
We administered the retinal locus characteristics and
ability scales using the SLO [18], the Pepper VSRT
[28,30], and the MNREAD Acuity Charts [31]. We
instructed subjects in development of a TRL below the
scotoma and in use of this TRL for reading words pre-
sented on the SLO. Following this instruction, we read-
ministered all tests.
Results
We could not completely ascertain whether subjects
were reading with the PRL or TRL when they were view-
ing the reading charts in posttesting; however, subjects
verbalized their understanding and monitored their own
eye movements after training. They demonstrated appro-
priate eccentric viewing techniques when asked to view
the experimenter’s face and other near targets, showing
that they understood and could use the TRL, rather than
the PRL, in posttesting.
The average age of the subjects was 75.7, with a range
from 68 to 82. The average subject acuity was right eye =
1.0 logarithm of minimum angle of resolution (logMAR)
(range 0.64–1.26), left eye = 0.98 logMAR (range 0.48–
1.52), both eyes = 0.91 logMAR (range 0.64–1.12), and
dominant PRL = 0.95 logMAR (range 0.64–1.18).
In our SLO protocol, we used a fixation target (cross)
in the PRL as a “tag” and another target (letter) for sec-
ondary fixation with the TRL. We gave the subject verbal
feedback while viewing his or her performance in the
SLO to assist development of the TRL. The subject was
told to “look at” the cross in the PRL and “notice” the let-
ter in the TRL. We used this method to help the subject
understand the difference between the PRL and TRL and
avoid a situation in which the subject was tempted in the
early training stage to “look” with the PRL at the TRL
target. Targets in the TRL were initially single letters and
767
WATSON et al. Trained retinal locus for text navigation
then two-letter words, three-letter words, etc. Figure 3
depicts the SLO map during TRL training.
When the subject consistently fixated, used the TRL
for eye movements, and read several successive words
with the TRL, we extinguished the PRL fixation target
(cross). The TRL was subsequently used for further eye
movement and reading instruction. We gave the subjects
a short break, then had them read again in the SLO. All
subjects were able to read with the TRL after the break;
none reverted to the PRL to read words in the SLO.
Table 2 provides the subjects’ PRL and TRL charac-
teristics. All subjects developed a TRL after an approxi-
mately 15-minute training session. This training session
with the SLO resulted in a TRL that closely matched the
PRL in size and in fixation, saccade, and pursuit abilities.
The TRL and PRL abilities did not significantly differ.
Subjects read significantly slower and with less text-
navigation ability with the TRL than the PRL. However,
considering that the training session was short, provided
exclusively in the SLO, and did not include practice fol-
lowing the SLO intervention, subject performance was
surprisingly responsive.
DISCUSSION
Experiment 1 showed very strong associations
between scotoma placement and reading errors on the
Pepper VSRT. We found that a scotoma to the right of the
PRL was related to reading errors on the right (missing
the end of words or the last word on the line of print). A
scotoma below the PRL was strongly related to reading
performance (both errors of skipping or rereading a line
of print and slower reading rate), a finding that has never
been reported in the literature. We expected to find that
scotoma to the left of the PRL would be related to reading
errors on the left (missing or confusing the beginning of a
Figure 3.
Scanning laser ophthalmoscope (SLO) image during trained retinal
locus (TRL) instruction of subject with central dense scotoma (DS)
(outlined in green). Subject looked at cross (+) with preferred retinal
locus (PRL) (outlined in red). Subject was asked to look at + and
notice V shown in TRL. V and + are superimposed on SLO image.
Once subject understood and complied with TRL training, + was
extinguished. SLO image vertically inverted as would be seen by
experimenter. Subject would see V below scotoma, while
experimenter would see V above scotoma.
Table 2.
Preferred retinal locus (PRL) and trained retinal locus (TRL) characteristics for seven subjects in Experiment 2.
Characteristic
PRL TRL
Mean Minimum Maximum Mean Minimum Maximum
Fixation Ability Score 3 3 3 3 3 3
PRL Size (° diameter) 5.1 3.5 6.0 5.1 4.0 7.0
Saccade Ability Score 2.4 2.0 3.0 2.0 1.0 3.0
Pursuit Ability Score 3.0 3.0 3.0 2.2 1.0 3.0
Reading Ability
MNREAD Rate (words/min) 74.4 19 135 34.1 8 92
MNREAD Acuity (logMAR) 1.14 0.71 1.42 1.34 1.02 1.66
Critical Print Size (logMAR) 1.33 0.80 1.60 1.45 1.30 1.60
Pepper VSRT Rate (words/min) 27.1 12.3 41.2 17.1 11.3 29.6
Pepper VSRT Accuracy (%) 94.4 87.9 99.3 72.6 54.6 94.7
logMAR = logarithm of minimum angle of resolution, MNREAD = Minnesota Low Vision Reading (Acuity Charts), VSRT = Visual Skills for Reading Test.
768
JRRD, Volume 43, Number 6, 2006
word or missing the first word on a line), but this proved
not to be the case.
Other low-vision clinicians and clinical researchers
have suggested that a PRL above or below a scotoma is
the most suitable PRL for reading because the scotoma
does not interfere with scanning across the line of print
[19,25,33,37]. Our data show that a PRL above the
scotoma is strongly associated with rereading lines or
skipping lines, which severely disrupts the reading pro-
cess. Our findings of the deleterious effects of a PRL
above a central scotoma make intuitive sense. Because
we read English from left to right and top to bottom, a
scotoma in this position always obscures that which is
below fixation. Because most languages are read from
the top of a page to the bottom, even non-English readers
with such a scotoma may have this difficulty.
Other low-vision clinicians and researchers have also
stated that a PRL to the right of a scotoma is a disadvan-
tageous reading position because a scotoma in this posi-
tion would ostensibly cause errors in scanning a line of
print [19,25,33,37]. Although problematic in the labora-
tory for readers without visual impairments using artifi-
cial scotomas, our research does not show that readers
with macular scotomas are disadvantaged. Our readers
with scotomas in this position actually showed an advan-
tage in reading the ends of words and the last word on a
line correctly. A PRL to the right of the scotoma may
provide suitable reading ability.
The results of Experiment 1 challenge former think-
ing about the relationship between scotoma placement and
reading performance. Our data suggest that the position of
the PRL relative to the scotoma may be important for
readers with macular loss but not as important as was pre-
viously thought. Prevailing clinical heuristics that a
scotoma to the right of the PRL is deleterious and a
scotoma above the PRL is advantageous appear to have
been validated. However, a scotoma below the PRL may
impair performance more than previously thought and a
scotoma to the left of the PRL may be more benign. In
clinics providing PRL training, low-vision clinicians can
use this information to better understand their patients’
reading performance.
Another result of our research is that error scores on
the Pepper VSRT are useful for determining the effect of
PRL placement on text navigation. The error scores as
coded and used in this research may be a quick, expedi-
ent proxy for determining PRL position in the absence of
an SLO. A fixation target in a fundus camera [32] or
visuoscope for determining retinal fixation position can
provide this information as well and has been used suc-
cessfully as an SLO proxy.
Although only a small number of subjects who were
screened qualified for Experiment 2, all subjects in this
experiment were able to develop a TRL in their better-
seeing eye with a well-developed PRL after approxi-
mately 15 minutes of SLO training. This SLO training
session resulted in a TRL that closely matched the PRL in
size and in fixation, saccade, and pursuit abilities. We did
not measure the distance of the PRL and TRL from the
fovea, so we do not know whether fixation stability was
similar because of this reason. The fixation scoring scale
may be a relatively coarse measure, and more precise
measures might have shown more differences in fixation
stability, pursuits, and saccades.
Subjects read significantly slower and with less text-
navigation ability with the TRL than the PRL. However,
considering that the training session was very short, pro-
vided exclusively in the SLO environment, and did not
include practice following the SLO intervention, subject
performance was surprisingly good. These subjects were
regular readers with a longstanding PRL; therefore, their
uniform ability to develop and use a TRL in the better-
seeing eye after a 15-minute SLO training intervention is
very encouraging. The results, while suggestive, were not
subjected to inferential statistical testing and must be
interpreted in terms of their exploratory nature. However,
we conclude, as do other researchers, that TRL develop-
ment is an area that deserves more research and clinical
attention.
Our subjects had long-standing macular loss (dura-
tion of at least 1 year) and were experienced low-vision
device users. They had been subjects in various experi-
ments at our center and volunteered eagerly for various
reasons: to gain more information about their visual
impairments, to enjoy a novel experience, to understand
whether something new might assist them in daily life,
etc. Some personality mechanism or other psychosocial
characteristic may have motivated these low-vision read-
ers to become research subjects and therefore respond
willingly and easily to our TRL training.
On the other hand, this is the first report of training a
TRL in the better-seeing eye of patients with macular loss
who were using their better-seeing eye with low-vision
devices for reading. These readers developed a TRL with
relative ease and speed. Whether it is possible for the gen-
eral clinical population in a low vision service to do so
769
WATSON et al. Trained retinal locus for text navigation
and find the TRL as useful as a PRL remains to be seen.
We do not know how long the effect of this short-term
TRL development lasted because we did not follow the
subjects.
One subject screened for Experiment 2 reported that
he had deliberately taught himself to use a TRL. Whereas
he had earlier demonstrated a PRL to the left of the
scotoma, at the recall visit, he demonstrated a retinal locus
below the scotoma. When the experimenter remarked on
this change, the subject stated that he had requested and
received his Pepper VSRT results from Experiment 1,
which demonstrated right directional errors. Upon return-
ing home, he taught himself to change his fixation and
found that reading was easier. Although he was dismissed
as a subject, the results of his enterprising self-training and
our study are in agreement and are encouraging.
CONCLUSIONS
Contrary to clinical heuristics, we found that the text-
navigation performance of our readers with macular
degeneration who had a PRL to the right of a scotoma did
not suffer unduly. We further found that text-navigation
of readers with a PRL above a scotoma was more prob-
lematic than previously believed. This research questions
some of the commonly used assessment and training
techniques developed for this population. Although clini-
cal heuristics are prevalent in medicine and rehabilitation
as quick “rules-of-thumb,” for practice, they must be
carefully evaluated.
Our subset of readers was able to quickly and reliably
develop a TRL in a position that was advantageous for
scanning text. Promising lines of future research are to
discover whether other advantageous TRL positions
(such as to the right of the scotoma) exist, whether a
longer training and practice session would improve per-
formance, and whether persons with macular loss can be
taught this technique without use of an SLO.
ACKNOWLEDGMENTS
This material was based on work supported by the
Department of Veterans Affairs, Rehabilitation Research
and Development Service (VA RR&D), grant C849-RA
(Gale R. Watson); and the Atlanta VA RR&D Center of
Excellence for Aging Veterans with Vision Loss (Gale R.
Watson).
The authors have declared that no competing inter-
ests exist.
REFERENCES
1. Fletcher DC, Schuchard RA, Watson GR. Relative loca-
tions of macular scotomas near the PRL: Effect on low
vision reading. J Rehabil Res Dev. 1999;36(4):56–64.
[PMID: 10678458]
2. Fletcher DC, Schuchard RA, Livingstone CL, Crane WG,
Hu SY. Scanning laser ophthalmoscope (SLO) macular
perimetry and applications for low vision rehabilitation cli-
nicians. Ophthalmol Clin N Am. 1994;7:257–65.
3. Maino JH, Carty RE. Part II. VICTORS: A model for the
provision of low vision services to the partially sighted vet-
eran. J Am Optom Assoc. 1983;54(11):991–93.
[PMID: 6643902]
4. Watson GR, De l’Aune WR, Stelmack JA, Maino JH, Long
S. National survey of the impact of low vision device use
among veterans. Optom Vis Sci. 1997;74(5):249–59.
[PMID: 9219282]
5. Timberlake GT, Mainster MA, Peli E, Augliere RA,
Essock EA, Arend LE. Reading with a macular scotoma. I.
Retinal location of scotoma and fixation area. Invest Oph-
thalmol Vis Sci. 1986;27(7):1137–47. [PMID: 3721792]
6. Timberlake GT, Peli E, Essock EA, Augliere RA. Reading
with a macular scotoma. II. Retinal locus for scanning text.
Invest Ophthalmol Vis Sci. 1987;28(8):1268–74.
[PMID: 3610545]
7. Cummings RW, Whittaker SG, Watson GR, Budd JM.
Scanning characters and reading with a central scotoma.
Am J Optom Physiol Opt. 1985;62(12):833–43.
[PMID: 4083327]
8. Legge GE, Rubin GS, Pelli DG, Schleske MM. Psycho-
physics of reading—II. Low vision. Vision Res. 1985;
25(2):253–66. [PMID: 4013092]
9. Legge GE, Ross JA, Isenberg LM, LaMay JM. Psycho-
physics of reading. Clinical predictors of low-vision read-
ing speed. Invest Ophthalmol Vis Sci. 1992;33(3):677–87.
[PMID: 1544792]
10. Mitra S. Spatial contrast sensitivity in macular disorder.
Doc Ophthalmol. 1985;59(3):247–67. [PMID: 4006670]
11. Schuchard RA, Raasch TW. Retinal locus for fixation:
Pericentral fixation targets. Clin Vis Sci. 1992;7:511–620.
12. Raasch TW. Method for assessing stereopsis and positional
sensitivity in normally-sighted and low vision observers.
Opt Soc Am Tech Dig Ser. 1991;1:109–12.
770
JRRD, Volume 43, Number 6, 2006
13. Schuchard RA, Fletcher DC, Maino JH. A scanning laser
ophthalmoscope low vision rehabilitation system. Eye Clin
Vis Care. 1994;6:101–7.
14. Schuchard RA. Validity and interpretation of Amsler grid
reports. Arch Ophthalmol. 1993;111(6):776–80.
[PMID: 8512478]
15. Whittaker SG, Budd JM, Cummings RW. Eccentric fixa-
tion with macular scotoma. Invest Ophthalmol Vis Sci.
1988;29(2):268–78. [PMID: 3338884]
16. White JM, Bedell HE. The oculomotor reference in
humans with bilateral macular disease. Invest Ophthalmol
Vis Sci. 1990;31(6):1149–61. [PMID: 2354915]
17. Schuchard RA, Fletcher DC. Preferred retinal locus. A review
with applications in low vision rehabilitation. Ophthalmol
Clin N Am. 1994;7:243–56.
18. Fletcher DC, Schuchard RA. Preferred retinal loci relation-
ship to macular scotomas in a low-vision population. Oph-
thalmology. 1997;104(4):632–38. [PMID: 9111255]
19. Bäckman O, Inde K. Low vision training. Malmö (Swe-
den): LiberHermods; 1979.
20. Watson GR, Berg RV. Near training techniques. In: Jose
RT, editor. Understanding low vision. New York (NY):
American Foundation for the Blind; 1983. p. 317–62.
21. Wright VW, Watson GR. Learn to use your vision for read-
ing workbook. Lilburn (GA): Bear Consultants; 1995.
22. Leat SJ, Fryer A, Rumney NJ. Outcome of low vision aid
provision: The effectiveness of a low vision clinic. Optom
Vis Sci. 1994;71(3):199–206. [PMID: 8196946]
23. Stelmack JA, Stelmack TR, Fraim M, Warrington J. Clini-
cal uses of the Pepper Visual Skills for Reading Test in low
vision rehabilitation. Am J Optom Physiol Opt. 1987;
64(11):829–31. [PMID: 3425678]
24. Goodrich GL, Mehr EB, Darling NC. Parameters in the use
of CCTV’s and optical aids. Am J Optom Physiol Opt.
1980;57(12):881–92. [PMID: 6164297]
25. Nilsson U. Visual rehabilitation with and without educa-
tional training in the use of optical aids and residual vision.
A Prospective study of patients with age-related macular
degeneration. Clin Vis Sci. 1990;6:3–10.
26. Fridal G, Jansen L, Klindt M. Courses in reading develop-
ment for partially sighted students. J Vis Impair Blind. 1981;
75(1):4–7.
27. Watson GR, Wright V, De l’Aune WR. The efficacy of
comprehension training and reading practice for print read-
ers with macular loss. J Vis Impair Blind. 1992;86(1):37–43.
28. Baldasare J, Watson GR, Whittaker SG, Miller-Schafer H.
The development and evaluation of a reading test for low
vision individuals with macular loss. J Vis Impair Blind.
1986;80(6):785–89.
29. Stelmack JA, Reda D, Ahlers S, Bainbridge L, McCray J.
Reading performance of geriatric patients post exudative
maculopathy. J Am Optom Assoc. 1991;62(1):53–57.
[PMID: 1813493]
30. Watson GR, Baldasare J, Whittaker SG. The validity and
clinical uses of the Pepper Visual Skills for Reading Test.
J Vis Impair Blind. 1990;84(3):119–23.
31. Legge GE, Ross JA, Luebker A, LaMay JM. Psychophysics
of reading. VIII. The Minnesota Low-Vision Reading Test.
Optom Vis Sci. 1989;66(12):843–53. [PMID: 2626251]
32. Nilsson UL, Frennesson C, Nilsson SE. Location and stability
of a newly established eccentric retinal locus suitable for
reading, achieved through training of patients with a dense
central scotoma. Optom Vis Sci. l998;75(12):873–78.
[PMID: 9875992]
33. Nilsson UL, Frennesson C, Nilsson SE. Patients with AMD
and a large absolute central scotoma can be trained suc-
cessfully to use eccentric viewing, as demonstrated in a
scanning laser ophthalmoscope. Vision Res. 2003;43(16):
1777–87. [PMID: 12818347]
34. Cummings RW, Rubin GS. Reading speed and saccadic eye
movements with an artificial paracentral scotoma [abstract].
Invest Ophthalmol Vis Sci. 1992;33(Suppl):1418.
35. Fine EM. Reading with central scotoma: What can we
learn from simulation studies? Vis Impair Res. 1999;1(3):
165–73.
36. Sunness JS, Schuchard RA, Shen N, Rubin GS, Dagnelie
G, Haselwood DM. Landmark-driven fundus perimetry
using the scanning laser ophthalmoscope. Invest Ophthal-
mol Vis Sci. 1995;36(9):1863–74. [PMID: 7635660]
37. McMahon TT, Hansen M, Stelmack JA, Oliver P, Viana
MA. Saccadic eye movements as a measure of the effect of
low vision rehabilitation on reading rate. Optom Vis Sci.
1993;70(6):506–10. [PMID: 8336914]
Submitted for publication July 7, 2005. Accepted in
revised form February 27, 2006.