Vision impairment secondary to macular disease occurs in more than 1.75 million individuals in the United States.1 When the foveae are affected, the result is not only a loss in central vision but also a loss of the main reference position for the ocular-motor system.2 As an adaptation to this lack of a reference position, patients commonly and quickly develop a new reference, the preferred retinal locus.3–8 Location of the preferred retinal locus can occur anywhere in the retina. Little is known about how this adaptation is made or how the preferred retinal locus location is chosen. Rees et al.9 found that the location of the preferred retinal locus may correspond to the retinal area that provides the most optimal visual acuity. Fletcher and Schuchard10 and Fine and Rubin11 reported that the distribution of preferred retinal loci locations in relation to the patient's central scotomas was below (48%) or to the left (49%) in most cases. The location of the preferred retinal locus can have a significant effect on visual function. Particularly, preferred retinal loci to the left of a central scotoma have been associated with a significant decline in reading rate for text read from left to right,12,13 as the scotoma in these cases would be to the right of fixation in visual field coordinates.
Eccentric viewing training has long been standard practice in a variety of low-vision rehabilitation services to train an individual to use a different retinal area, that Nilsson and her colleagues14 named the “trained retinal locus.” The rationale for the trained retinal locus is to use a retinal area that has better visual acuity and/or is located in a more optimal position relative to the scotoma, which may in turn optimize visual function. A few studies have indicated that patients can successfully use a trained retinal locus after training, with findings of some improvement in reading rates at the new location.14,15 However, these studies have too few subjects to draw any broad conclusions. In fact, Watson et al.16 found that reading with a trained retinal locus resulted in slower reading rates.
In addition to scotoma location, patients' visual function can be impaired by fixation instability. A bivariate contour ellipse, usually representing 95% of fixation loci, is typically used to represent fixation stability around the preferred retinal locus17 and is a convenient way of summarizing fixation patterns. Other estimates have been used to quantify the parameters of the preferred retinal locus, when and if the patient's fixation pattern is unclear.18 In patients with macular disease, the area of the bivariate contour ellipse can be much larger,10 reflecting poor oculomotor control and fixation instability while using the preferred retinal locus. It has been proposed that eccentric viewing training may be able to improve a patient's fixational stability and visual acuity using a new trained retinal locus.15 In fact, many rehabilitation programs have been designed to this end.19–22 However, there has not yet been a clinical trial to determine if these training programs are efficacious or if they achieve the desired clinical outcome.
The concept and practice of eccentric viewing training pre-date the discovery of the preferred retinal locus and have been consistently practiced within the Veterans Affairs hospital.21,23–26 Eccentric viewing training aims to improve visual skills using the preferred retinal locus or to change the preferred retinal locus to a functionally more optimal retinal location.
This study is an observational study examining whether standard, usual care of eccentric viewing training, as prescribed in the Hines Veterans Affairs Blind Rehabilitation Center, leads to a change in the retinal area used for fixation (i.e., creates a new trained retinal locus that is at a significantly different location from the natural occurring preferred retinal locus) and/or improves fixation stability, with consequently decreased area of the fixation ellipse. Using a convenient sample of patients already undergoing an established eccentric viewing training program allows us to explore the feasibility of conducting a clinical trial in eccentric viewing training. If no effect is observed in the observational study, then it is likely that no effect would be observed in a clinical trial using the Hines Veterans Affairs Blind Rehabilitation Center eccentric viewing training procedures. If an effect is observed, then it is likely that either eccentric viewing training is a successful treatment or test-retest variability is responsible for a pseudo-effect, which will also occur in a control group receiving sham treatment.
Seventy-six patients from the Hines Veterans Affairs Blind Rehabilitation Center were recruited as participants. This study was conducted in accordance with the tenets of the Declaration of Helsinki and was approved by the Hines Veterans Affairs Hospital Institutional Review Board. Written informed consent was obtained. All participants were legally blind with bilateral central scotomas verified upon admission, as a consequence of macular disease. Macular degeneration was the main pathology in 92% of participants, none of whom were receiving ongoing antivascular endothelial growth factor injections or treatments. Participants were enrolled in the study before beginning their low-vision rehabilitation program. Information on the duration of the disease was not known upon admission to the center. Ages and visual acuities of the participants are shown in Table 1.
All participants had a comprehensive low-vision evaluation conducted according to Optometric Clinical Practice Guidelines for Care of the Patient with Visual Impairment published by the American Optometric Association.19 Participants received a comprehensive inpatient blind rehabilitation program administered by a multidisciplinary staff including nurses, physicians, optometrists, psychologists, social workers, and rehabilitation specialists. The program included training in orientation and mobility, daily living skills, manual skills, adaptive computer technology, and visual skills. Program length varied from 4 to 6 weeks, depending on the participants' needs. The low-vision (visual skills) training included instruction to improve use of remaining vision as well as care and use of prescribed, optical, nonoptical, and electronic low-vision devices to meet identified goals. Participants received 5 to 10 hours of training each week.27 Eccentric viewing training was provided using the recommended spectacle prescription for best visual acuity, according to usual care at the Hines Veteran Hospital, Blind Rehabilitation Center. The location of the prescribed trained retinal locus was determined during the low-vision evaluation by the low-vision optometrist, as the eccentric gaze position that produced the best visual acuity. This location was recorded in clock-dial notation and determined by improvement in acuity recorded as the patient looked in all four clock-hour gazes (superior, inferior, left, right) with the Feinbloom acuity chart. A full description of the instructional components of eccentric viewing training and management of scotomas is provided by Stelmack et al.22
The retinal location used for fixation, the preferred retinal locus, was recorded before and after completion of the entire rehabilitation program using a scanning laser ophthalmoscope (Rodenstock, GmbH, Munich, Germany, 1997). In addition, fixation behavior in the participant's better-seeing eye (determined by best-corrected visual acuity) was assessed before and after eccentric viewing training using the scanning laser ophthalmoscope. Participants were asked to “look at” a centrally located red fixation cross (presumably using either their preferred retinal locus pre-eccentric viewing training or trained retinal locus post-eccentric viewing training) while microperimetry was performed. Participants were instructed to fixate on a cross in the center of the scanning laser ophthalmoscope's test screen while the edges of their scotoma were mapped out using a 30-arcmin black (100% contrast) square test stimulus with an exposure duration of 500 milliseconds. While the participant was maintaining fixation on the central target, the scanning laser ophthalmoscope recorded the position of the participant's eye. Thus, the scanning laser ophthalmoscope recorded any changes in eye position that may have occurred from eye movements and/or changes in fixation. The experimenter first manually positioned the stimulus over a retinal area to be tested, at which point the stimulus was then presented (flashed) to the subject. Light stimuli were presented on either side of the scotoma to demarcate the seeing and nonseeing edge of the scotoma. This procedure was repeated until a plot of the entire outer edge of the scotoma was obtained. This assessment, both the recording of eye position while fixating and scotoma edge mapping, was recorded before and after eccentric viewing training. It should be noted that, because fixation stability was obtained while performing microperimetry, it is possible that subjects' fixation stability was reduced as a result of the longer testing time and cognitive load associated with performing microperimetry.
Eccentric Viewing Training
Current standard clinical methods of eccentric viewing training were implemented for all participants.22 These included evaluation of performance on tasks such as word recall, letter recall, and card recall. These were carried out while a certified low-vision therapist manually monitored the participant's fixation. In addition, verbal feedback was given to each participant as reinforcement to use the prescribed eccentric viewing position.
Total training time varied among patients and was tailored to each individual patient based on achievement of optimal acuity. The number of in-office training hours varied per patient, and the eccentric viewing training was one component of the comprehensive low-vision rehabilitation program. The eccentric viewing training occurred with the certified rehabilitation therapist.
Exact eye position coordinates for fixations made during the scanning laser ophthalmoscope scotometry task were established before and after eccentric viewing training for all participants. This was done by digitizing the scanning laser ophthalmoscope printouts that graphically displayed the location of the eye when fixations were made on the centrally located fixation target. The resulting eye position pixel coordinates were then corrected for any distortion inherent with the raster scan of the scanning laser ophthalmoscope optical system. This was done by measuring and correcting for the amount of barrel distortion in our Rodenstock scanning laser ophthalmoscope using the method of Timberlake et al.28
Fixation pixel coordinates were then referenced or normalized to a retinal landmark, typically a vessel crossing that was present in both the pre-eccentric and post-eccentric viewing training scanning laser ophthalmoscope images. This was done so that the fixation pixel eye coordinates from both the pre-eccentric and post-eccentric viewing training scanning laser ophthalmoscope images could be combined onto one graph. The pixel fixation coordinates were then multiplied by a scaling term to convert the pixels to degrees.
To control for any magnification differences and for rotation of the image between the pre-eccentric and post-eccentric viewing training scanning laser ophthalmoscope images, three retinal landmarks were selected that were present in both the pre-eccentric and post-eccentric viewing training scanning laser ophthalmoscope images. Specifically, a magnification correction factor was computed and applied to each post-eccentric viewing training scanning laser ophthalmoscope image by comparing the distance between two landmarks in the pre-eccentric viewing training scanning laser ophthalmoscope image with the distance to the same two landmarks in the post-eccentric viewing training scanning laser ophthalmoscope image. A similar correction factor for rotation was computed and applied to the post-eccentric viewing training scanning laser ophthalmoscope image by comparing the angles formed by the three landmarks in the pre-eccentric and post-eccentric viewing training scanning laser ophthalmoscope images. Thus, in all cases, the pre-eccentric viewing training scanning laser ophthalmoscope image was taken as the “standard” or reference image.
After all scanning laser ophthalmoscope images had been corrected for rotation and magnification differences, the areas over which fixations were made during the fixation task, both pre-eccentric and post-eccentric viewing training, were calculated by fitting a bivariate ellipse. The variability in fixation—length of the major and minor axes—was set at ±2 standard deviations. Thus, the bivariate ellipse represented a 95% confidence ellipse of the fixation area. A bivariate ellipse was fitted because participants' fixations were assumed to follow a bivariate Gaussian distribution.
Three primary outcome measures for preferred retinal locus/trained retinal locus location and fixation stability were obtained from the bivariate ellipse: (i) the x and y coordinates of the centroid of the ellipse, (ii) the orientation angle (relative to the horizontal meridian) of the bivariate ellipse, and the (iii) ellipse area. The ellipse centroid represented the average location of the participant's preferred retinal locus (pre-eccentric viewing training) and trained retinal locus (post-eccentric viewing training). The orientation angle represented the angular direction of the participant's fixation ellipse pre-eccentric and post-eccentric viewing training. The ellipse area was an indicator of fixation stability when the participant was fixating with his/her preferred retinal locus (pre-eccentric viewing training) or trained retinal locus (post-eccentric viewing training). It should be acknowledged that because it was not possible to determine whether the eye visited the same fixation location multiple times while measuring fixation stability, the computed bivariate ellipse areas may be an overestimation of subjects' fixation stability if defined in terms of relative time spent at each point rather than the area covered during a time interval.
The change in preferred retinal locus location, angular direction, and area with eccentric viewing training was computed by taking the difference between the pre-eccentric and post-eccentric viewing training centroid x and y coordinates, the ellipse orientation angle, and the ellipse area, respectively.
Change in Preferred Retinal Locus to Trained Retinal Locus Location
A histogram (Fig. 1) of the distribution of the differences in preferred retinal locus to trained retinal locus location after eccentric viewing training shows a median change of 6.6° in variable directions, with an interquartile range of 3.4 to 10.1°. Fig. 1 also shows that the differences in preferred retinal locus to trained retinal locus location after eccentric viewing training followed a continuous, nonnormal distribution with a central tendency. All participants exhibited some difference in trained retinal locus location after eccentric viewing training as described by a Weibull distribution with α = 1.47 and β = 8.331. The mean of the Weibull function is in close agreement with the empirical median and mean in our data sample, suggesting that the mean is significantly different from zero, as it does not follow an exponential distribution.
In terms of the directionality, as to where the trained retinal locus was established, all directions were equally probable. Fig. 2 illustrates the distribution of the direction of changes in preferred retinal locus to trained retinal locus. The distribution is approximately rectangular, indicating equal probability of all directions.
Using data that were available only in a subset of participants (n = 28), we compared the prescribed direction of the trained retinal locus position with that of the actual trained retinal locus position after training (Fig. 3). This analysis revealed that for 47% of the participants the actual trained retinal locus position (post-training) was within one to two clock dials (30 to 60°) of the prescribed trained retinal locus; 25% of the participants showed a trained retinal locus positioned within one clock dial of the prescribed position (30°).
Changes in Preferred Retinal Locus and Trained Retinal Locus Fixation Area
The change in size of the bivariate ellipse area with eccentric viewing training is shown in Fig. 4. There was a wide range of changes, with some participants showing a smaller ellipse after training and others showing a larger ellipse size after training. No main effect of training was found on fixation (ellipse) area (paired t test, P = .54). Fig. 5A illustrates the log of the bivariate ellipse area before training on the x axis versus log of the bivariate contour ellipse after training. Given that the data are symmetrically distributed along the solid identity line and that the percent change in area remains constant when presented on a log scale, this indicates that there was no effect of eccentric viewing training on fixation stability. The dashed line illustrates the regression line (r = 0.419).
Visual acuity's logarithm of the minimum angle of resolution was found to be a significant predictor of log pre-training bivariate contour ellipse area (R2 = 0.073, P = .02). As illustrated in Fig. 5B, large bivariate contour ellipse areas were associated with worse visual acuity (r = 0.27).
Variability and Measurement Error
To evaluate whether changes in the ellipse size were within measurement error, we compared the change between trials with changes from pre-training to post-training. Average bivariate 95% contour ellipse radii were calculated before and after eccentric viewing training for each subject. The radii of the ellipse for each subject were converted to standard deviation units by dividing each radius by 1.96. We then computed the area of the standard deviation ellipse for each subject and condition. The areas of the standard deviation ellipses were then averaged across subjects to give an average standard deviation ellipse area for pre-eccentric viewing training and post-eccentric viewing training. We then estimated the radius of a circle that was equal in area for pre-eccentric and for post-eccentric viewing training standard deviation ellipses. These two radii where then averaged and multiplied by 2.576, the z score that must be used to obtain the 99% bivariate contour. This radius was used to compute the red circle in Fig. 6, which represents the 99% bivariate contour. A change outside the 99% bivariate contour would represent a change that significantly exceeds test-retest reliability for preferred retinal locus location for each participant. Fig. 6 also shows the amount and direction of change in the location of the center of fixation (from the preferred retinal locus to the trained retinal locus). As illustrated, most participants had a change in preferred retinal locus–trained retinal locus location significantly different from test-retest reliability (P > 1.2 × 10−7, which is significant after a Bonferroni correction for dividing α = .05 among 76 comparisons).
Although eccentric viewing training is routinely assigned as part of many comprehensive visual rehabilitation programs, there are limited data on the efficacy of such training. The preliminary results from this study are promising. We found that the preferred retinal locus location shifted after eccentric viewing training in the subset of participants where these data were available, and it shifted to the general area intended by the instructor in 47% of these participants. Although this is encouraging (or discouraging depending on one's expectations), this finding suggests that more research is needed to improve our understanding of the factors that influence success and failures of eccentric viewing training. Furthermore, the observed shift in preferred retinal locus location found in this study was clinically significant because the amount by which the preferred retinal locus shifted with eccentric viewing training was greater than what would be attributed to within-subject variability in fixation stability (Fig. 6).
We did not find, however, that eccentric viewing training significantly improved fixation stability. This result was not expected because eccentric viewing training is often believed to also improve fixation stability. It is possible that to improve fixation stability at a new preferred retinal locus (i.e., the trained retinal locus) more eccentric viewing training or other types of oculomotor training is required. As a result, more research is required to determine the exact specifications of eccentric viewing training that will result in improved fixation stability at the trained retinal locus.
Further investigations involving a control group are also needed to assess the efficacy and functional outcomes of eccentric viewing training. Because all participants in this study received usual low-vision rehabilitation and care in addition to eccentric viewing training, it is not possible to determine the specific contribution of eccentric viewing training. No functional outcomes, such as reading performance, facial recognition, or mobility tasks, were evaluated before and after eccentric viewing training. Such evaluations should be considered in developing a future clinical trial. Studies on reading performance with age-related macular degeneration have already shown some promise with eye movement and other types of training resulting in improvements of reading speed.29
In addition, much work regarding the reliability of preferred retinal locus measures by fundus imaging methods needs to be done. Although work in animal models indicates that the preferred retinal locus develops instantly and reliably within a small location,30 this may not be the case in human subjects. However, Crossland et al.8 reported that a repeatable preferred retinal locus fixation area developed within the first 6 months of disease onset in 25 patients with macular disease. In addition, an area that remains relatively unexplored is binocularity and preferred retinal loci. Most studies, including this one, have investigated the preferred retinal locus under monocular viewing conditions owing to limitations in equipment; however, future developments (Wiecek et al. IOVS 2015;56 ARVO E-Abstract 549; Timberlake et al. IOVS 2013;54: ARVO E-Abstract 2182) may make analysis under binocular viewing possible. At which time, we would be able to determine the characteristics of a binocular preferred retinal locus and whether this is driven by the dominant eye or by both eyes. Studies exploring reading and gaze changes under monocular and binocular viewing by Kabanarou et al.31 and Kabanarou and Rubin32 indicate that there is some binocular gain.
The results from this study set the stage for a clinical trial by demonstrating the feasibility of detecting a difference between pre-eccentric and post-eccentric viewing training measures, but without a control group, no definitive conclusions can be made regarding the contributions eccentric viewing training had to this difference. A future clinical trial could involve exploration of eccentric viewing training in isolation, apart from other strategies and training involved in low-vision rehabilitation programs. How a patient learns to view objects in the presence of a central scotoma is a complex issue. We have learned a great deal herein, but there is much more to explore before we can determine (1) whether eccentric viewing training helps patients develop skills that might otherwise not acquire, (2) what type of training might be most effective to improve fixation stability and to adopt the prescribed trained retinal locus, (3) and which patients might demonstrate the most benefit from an eccentric viewing training program.
1. Friedman DS, O'Colmain BJ, Tomany SC, et al. Prevalence of Age-related Macular Degeneration in the United States. Arch Ophthalmol 2004;122:564–72.
2. White JM, Bedell HE. The Oculomotor Reference in Humans with Bilateral Macular Disease. Invest Ophthalmol Vis Sci 1990;31:1149–61.
3. Schuchard RA. Preferred Retinal Loci and Macular Scotoma Characteristics in Patients with Age-related Macular Degeneration. Can J Ophthalmol 2005;40:303–12.
4. Crossland MD, Dunbar HM, Rubin GS. Fixation Stability Measurement Using the MP1 Microperimeter. Retina 2009;29:651–6.
5. Chung ST. The Glenn A. Fry Award Lecture 2012: Plasticity of the Visual System Following Central Vision Loss. Optom Vis Sci 2013;90:520–9.
6. Chung ST, Legge GE. Functional and Cortical Adaptations to Central Vision Loss. Vis Neurosci 2005;22:187–201.
7. Heinen SJ, Skavenski AA. Adaptation of Saccades and Fixation to Bilateral Foveal Lesions in Adult Monkey. Vision Res 1992;32:365–73.
8. Crossland MD, Culham LE, Kabanarou SA, et al. Preferred Retinal Locus Development in Patients with Macular Disease. Ophthalmology 2005;112:1579–85.
9. Rees AL, Kabanarou SA, Culham LE, et al. Can Retinal Eccentricity Predict Visual Acuity and Contrast Sensitivity at the PRL in AMD Patients? Int Congress Ser 2005;12182:694–8.
10. Fletcher DC, Schuchard RA. Preferred Retinal Loci Relationship to Macular Scotomas in a Low-vision Population. Ophthalmology 1997;104:632–8.
11. Fine EM, Rubin GS. Reading with Simulated Scotomas: Attending to the Right Is Better than Attending to the Left. Vision Res 1999;39:1039–48.
12. Timberlake GT, Mainster MA, Peli E, et al. Reading with a Macular Scotoma I. Retinal Location of Scotoma and Fixation Area. Invest Ophthalmol Vis Sci 1986;27:1137–47.
13. Legge GE, Rubin GS, Pelli DG, et al. Psychophysics of Reading II—Low Vision. Vision Res 1985;25:253–65.
14. Nilsson UL, Frennesson C, Nilsson SE. Patients with AMD and a Large Absolute Central Scotoma Can Be Trained Successfully to Use Eccentric Viewing, as Demonstrated in a Scanning Laser Ophthalmoscope. Vision Res 2003;43:1777–87.
15. Tarita-Nistor L, González EG, Mandelcorn MS, et al. Fixation Stability, Fixation Location, and Visual Acuity After Successful Macular Hole Surgery. Invest Ophthalmol Vis Sci 2009;50:84–9.
16. Watson GR, Schuchard RA, De l'Aune WR, et al. Effects of Preferred Retinal Locus Placement on Text Navigation and Development of Advantageous Trained Retinal Locus. J Rehabil Res Dev 2006;43:761–70.
17. Timberlake GT, Sharma MK, Grose SA, et al. Retinal Location of the Preferred Retinal Locus Relative to the Fovea in Scanning Laser Ophthalmoscope Images. Optom Vis Sci 2005;82:177–85.
18. Crossland MD, Sims M, Galbraith RF, et al. Evaluation of a New Quantitative Technique to Assess the Number and Extent of Preferred Retinal Loci in Macular Disease. Vision Res 2004;44:1537–46.
19. Freeman KF, Cole RG, Faye EE, et al. Optometric Clinical Practice Guideline: Care of the Patient with Visual Impairment (Low Vision Rehabilitation). St. Louis, MO: American Optometric Association; 2007: Available at: https://www.aoa.org/documents/optometrists/CPG-14.pdf
. Accessed October 3, 2018.
20. Mogk LG, Mogk M. Macular Degeneration: The Complete Guide to Saving and Maximizing Your Sight. New York: Ballantine; 1999.
21. Quillman RD. Low Vision Training Manual. Kalamazoo, MI: Western Michigan University; 1980.
22. Stelmack JA, Rinne S, Mancil RM, et al. Outcomes from a Structured Curriculum Used in the Veterans Affairs Low Vision Intervention Trial. J Vision Impair Blind 2008;102:636–48.
23. Stelmack JA, Massof RW, Stelmack TR. Is There a Standard of Care for Eccentric Viewing Training? J Rehabil Res Dev 2004;41:729–38.
24. Holcomb JG, Goodrich GL. Eccentric Viewing Training. J Am Optom Assoc 1976;47:1438–43.
25. Goodrich GL, Quillman RD. Training Eccentric Viewing. J Vision Impair Blind 1977;71:377–81.
26. Goodrich GL, Mehr EB. Eccentric Viewing Training and Low Vision Aids: Current Practice and Implications of Peripheral Retinal Research. Am J Optom Physiol Opt 1986;63:119–26.
27. Stelmack JA. Low Vision Program: Central Blind Rehabilitation Center. J Vision Impair Blind 1991;5:19–24.
28. Timberlake GT, Sharma MK, Gobert DV, et al. Distortion and Size Calibration of the Scanning Laser Ophthalmoscope (SLO) Laser-beam Raster. Optom Vis Sci 2003;80:772–7.
29. Seiple W, Szlyk JP. Eye-movement Training for Reading in Patients with Age-related Macular Degeneration. Invest Ophthalmol Vis Sci 2005;46:2886–96.
30. Skavenski A, Acocella C, Heinen S, et al. Patterns of Eye Movement Adaptation to Foveal Lesions in Adult Primates. In: Fuchs AF, ed. Contemporary Ocular Motor and Vestibular Reseach: A Tribute to David A. Robinson, International Meeting Eibsee 1993. New York: Thieme Medical Publishers; 1994:134–41.
31. Kabanarou SA, Crossland MD, Bellmann C, et al. Gaze Changes with Binocular Versus Monocular Viewing in Age-related Macular Degeneration. Ophthalmology 2006;113:2251–8.
© 2019 American Academy of Optometry
32. Kabanarou SA, Rubin GS. Reading with Central Scotomas: Is There a Binocular Gain? Optom Vis Sci 2006;83:789–96.