People with vascular diseases of the retina characteristically have reduced contrast enhancement abilities -- especially under low luminances. The marked decrement in visual function and in task performance, even for people with relatively intact vision at high light levels, is not fully appreciated by professional practitioners, rehabilitation personnel, and rehabilitation researchers.
In order to convey the full meaning and impact of the marked decrement in visual function under low luminance conditions, we have been able to secure unusual patients with asymmetric defects; i.e., with one eye relatively normal, and the fellow eye with early ARM with only minimal acuity decrement under high light levels. By means of comparing the right eye's scene to the left eye's scene under a variety of high and low luminance conditions, these subjects were able to vary the TV display parameters of contrast, color, brightness and resolution, so that we were able to obtain photographs of exactly what the subjects saw with their better eye as well as the poorer eye. These dramatically demonstrate the task performances that suffer least and most.
The initial photographs in this series have been improved using strictly photographic methods and darkroom manipulations. This demonstration was presented and published in the Pisart Tenth Anniversary Award Symposium on Low Vision, September 1990, and should result in an improved understanding of the real-world functional problems faced by those with early macular disease.
The RERC was asked by several private firms to supply consulting expertise on various aspects of state-of-the-art heads-up displays, and especially on modifications needed for applications in low vision. Our staff, including Christopher Tyler, Ph.D., and John Brabyn, Ph.D., have advised such companies as Spread Spectrum Technologies, Inc. on the relevant features needed for possible low vision applications of their head-mounted display, which utilizes a half-silvered mirror and LCDs to project an image in front of the wearer. Our input acquainted them with possible uses of their system for those with impaired vision, and specific criteria for critical and non-critical parameters which would need to be modified for the low vision user.
Similarly, in collaboration with Drs. Colenbrander and Scheinholtz of our affiliated Low Vision Clinic, we addressed similar issues relating to a new design from the Human Interface Technology Laboratory, at the Washington Technology Center, University of Washington, Seattle. Their system used a color LCD display in a relatively cosmetic headband package which allows placement of the electronically displayed image in either the upper or lower visual field, leaving the remaining field available for mobility and orientation. A prototype of this system was loaned to Dr. Scheinholtz for evaluation and pre-production feedback.
In collaboration with Drs. Colenbrander and Scheinholtz, we have tested various combinations of this system with different camera inputs and compared the output with more conventional systems such as CCTV reading systems. Dr. Scheinholtz redesigned the optical system for use by persons with low vision, and submitted the design to the manufacturer.
To provide a closer electronic analogy to an optical magnifier, we also developed a novel self-contained electronic magnifier prototype based on commercially available cameras and displays. A liquid crystal display was interfaced to a small camera and packaged in a single box with the camera underneath and the display on top, for initial experimentation with this concept.
A commonly suggested solution for making best use of undamaged portions of the retina in patients with macular disease was originally put forward by NASA engineers in 1986. This concept, visual "warping" or "remapping," involves moving the image on the retina out of the "blind spot" and displaying it in adjacent parts of the visual field.
For a practical, real-world task such as reading, the remapping concept effectively reduces to a one-dimensional problem involving the creation of a moving gap in the text corresponding to the location of the patient's scotoma. As the patient scans his eyes along a line of text, each letter encountered just to the right of the scotoma will "jump" across the gap to a position just to the left of the scotoma so that all the text information is "remapped" to an area outside the affected retinal area.
Methods previously proposed to implement this concept have involved complex image processing computers. We developed a much simpler approach, utilizing a novel bi-prism arrangement in which text is viewed through two prisms placed base to base. The separation, elevation and power of the prisms can be used to control the size of the apparent "scotoma." In use, the patient simply fixates (using marks on the bi-prism surface as a guide) so that his scotoma falls on the area between the bases of the prisms, and scans the prism over the text being read (or alternatively, the bi-prism can be held stationary while the text is scanned underneath).
A prototype bi-prism of this type was built and successfully demonstrated. For clinical testing, we engaged the services of a low vision Fellow (Simon Warner, M.D.) in collaboration with August Colenbrander, M.D., of our affiliated low vision clinic. Dr. Warner constructed two prototypes of the bi-prism "remapper" -- one designed to be attached to standard optical magnifiers, and one for attachment to a CCTV reading aid.
Initial pilot testing in the low vision clinic yielded mixed results. As we hypothesized, adjustment to such remapping is not an entirely natural function, and patients appear to require time to become accustomed to it. In addition, the model of macular disease assumed by the remapping concept -- i.e., that scotomas are "black holes" -- is seldom true in practice, and this fact is bound to reduce the range of applicability of the concept. The bi-prism approach is now freely available to any researchers wishing to test the remapping concept.
A new version of our spectacle-clip illuminator was developed in collaboration with the California Pacific Medical Center Low Vision Clinic. This device solves the problem faced by low vision persons who read with a 20- to 40-diopter spectacle lens, necessitating a 1- to 2-inch working distance. At this distance, the head and reading material largely occlude whatever ambient light is present, making reading difficult. In the original design the light source and reflector were attached to the bridge section of the user's spectacles using a clip adapted from commercially available sunglasses. However, this design meant that separate versions would have to be made for use with the right and left lenses. The new version allows attachment to either left or right spectacle lens.
The new device has recently been undergoing evaluation in a comparative testing trial with other illuminator devices by the VA RR&Center in Atlanta, and commercial manufacturers have expressed interest.
We also developed a modified form of quartz-halogen dichroic head-mounted low vision illuminator, incorporating a lightweight battery pack and head-mounted switch. This will also be offered to commercial manufacturers.
An effective rehabilitation aid to restore binocularity when slight distortion (metamorphopsia) is present in a particular part of the field of one eye is a "dab of nail polish" on a spectacle lens. This blots out only that small section of the field with distortions in one macular (for distance or reading), allowing fusion in the remaining, more forgiving, peripheral retina. While this method has previously been used informally and experimentally, we have developed a practical series of translucent stickers for this purpose. We determined optimal sizes and configurations of these simple but effective aids, and made the technique available to other practitioners by presenting it in a paper (entitled "Functional Characteristics of Patients With Early Maculopathy") at the Pisart Tenth Anniversary Award Symposium, sponsored by the Lighthouse, Inc., New York.
Under Research and Demonstration funding from NIDRR, innovations in low vision telescope design were pursued. One of the devices is a new ergonomic grip which can be adapted to a wide variety of telescopes. This is an attachment for the most popular low vision telescopes enabling a steadier, more comfortable grip and the ability to hold and focus the telescope with one hand. The steadier grip also allows higher powered scopes to be practical for elderly users. Response from patients and clinicians has been extremely positive, and Sam Walters, Inc. of Los Angeles, manufacturers of low vision aids, has taken on production of the device.
In collaboration with Alan Scott, M.D., Smith-Kettlewell Executive Director, we developed an entirely new concept for a high power (10X) illuminated spectacle-mounted binocular magnifier. This new device employs a pair of 10X magnifiers with appropriate base-in prism calculated to allow comfortable use at close working distances with the user's eyes comfortably focused at distance and unconverged. Illumination is provided by a focused bulb powered by an "N"-sized cell positioned between the two lenses. The complete device (lenses, bulb, and battery) is packaged in a compact unit which mounts easily on any pair of glasses. Total weight is 1 oz.
In use, the device functions as an "upside down" bifocal. Since it covers only the top quarter of the spectacle lenses, it does not impede normal viewing for walking, reading, etc. When the user desires to examine an object closely in detail, he can switch on the illuminator (by rotating the light bulb) and tilt his head down so that the object of regard appears in the magnifier. This mode of operation is similar to that used in bioptic telescopes for driving. We hypothesize the design may be useful in broader applications (fly fishing, hobbies, etc.) in addition to its utility for low vision. After fabricating several prototypes with different designs, we performed initial informal testing of the resulting experimental unit. These tests demonstrated a need for greater light output; we plan to revisit this project when improved battery and bulb technology becomes available.
We have begun a major study of vision impairment in the elderly covering 1,000 persons over 55. The overall goals of the study are:
We have obtained additional funding from the National Eye Institute, and in-kind support from the Buck Center for Research on Aging in Novato, California, to carry out this study, which addresses the NIDRR goal of improving our understanding of visual impairments and their functional impact on individual's lives, in order to improve rehabilitation strategies. The need for this study arises from the fact that relatively little is known about vision function in the elderly, especially those over 75. Many patients presenting at eye clinics complain of difficulties performing daily living activities such as driving at night or at dusk, reading menus in restaurants, walking outdoors at night, adapting to changes in light level, etc. The commonly used clinical tests (such as the well-known Snellen chart) measure vision only under optimal viewing conditions of high contrast and high light levels, and often do not show any loss in vision function in spite of the patient's real-life problems with vision.
To address these problems, we wish to evaluate a number of new, practical, low-cost vision function tests designed to predict real-world visual performance problems in non-ideal viewing conditions. We will determine whether these new and unique tests, designed to give clinically practical measures of vision function in conditions such as poor lighting, contrast and glare, will predict the actual visual performance problems faced by the older person in such tasks as reading, driving, walking, and other daily living activities. We will specifically address the following questions:
1. Does performance on the battery of vision function tests predict:
- subjective visual complaints as reported via questionnaire responses?
- real-life task performance (driving, reading, low-light walking speed)?
2. Does actual task performance relate to subjectively reported performance via questionnaire? In other words, do people have an accurate sense of their sensory limitations on tasks of daily living?
In the course of the study, we will collect a large, comprehensive and valuable set of data about vision function in the elderly. In spite of the growing importance of this population, there is still currently a scarcity of such data, and a virtual lack of information about vision in people over age 75. The data collected in this study will therefore be a valuable resource.
To date, over 500 people ranging in age from 58 to 95 years have been tested. Testing is scheduled to be completed by June 1996.
The SKILL Card is a simple-to-administer, inexpensive vision test which we have devised in order to provide better clinical assessment of vision under conditions of poor illumination and contrast. These viewing conditions represent a major problem for the elderly and those with various retinal and optic nerve diseases. For example, driving at night or at dusk, reading menus in dimly lit restaurants, and many other tasks become difficult for these patients, even though most of the widely-used clinical vision tests may indicate that their vision is normal. Few objects in our visual world have the high contrast and optimal lighting conditions of these clinical tests. The new SKILL (Smith-Kettlewell Institute Low Luminance) Card can detect, quantify, and monitor the types of vision difficulties mentioned above by simulating poor viewing conditions under normal office lighting. This provides the eye care professional with a convenient test which better correlates with the patient's reported real world vision problems.
The SKILL Card consists of two letter charts, mounted back-to-back, designed to be held at reading distance. Side one of the card is a standard black on white letter chart for assessing visual acuity. On the flip side is a low contrast, low luminance chart consisting of black letters printed on a dark grey background, allowing low luminance to be obtained using normal office lighting. In practice, side one is tested first. The card is then flipped over and side two is tested. The difference between the number of letters that the patient reads on the two sides is the SKILL score.
The resulting test takes only a couple of minutes, and meets our goal of rapid and simple utilization by the busy eye care practitioner.
Current goals for this project during this reporting period were:
Production
We have determined the best method for production and packaging of the SKILL Card, and 1000 cards have been produced under private funding. Distribution to interested clinicians has begun. The cards are mounted on foam core for rigid support. Attached to each is a cord for measuring the appropriate 40 cm test distance. Instructions for use and norms are included with each card.
Age Norms
Data have been collected from approximately 80 observers who showed no abnormalities on any vision measure and had no known ocular disease. These data were used to establish age norms and confidence limits for individuals from 15 to 100 years of age. The SKILL score shows a linear increase with age.
Effects of Eye Disease on SKILL Card Score
Among much other testing, we have used the SKILL Card to measure vision in 15 recovering optic neuritis patients, 6 patients with age-related maculopathy and 10 of their family members. The SKILL Card has proven to be extremely sensitive. All of the optic neuritis patients had abnormal SKILL scores (see below). The SKILL scores of the ARM patients were also quite abnormal.
At ARVO 1992 we presented a pilot study of diverse visual functions in optic neuritis patients. Patients were seen an average of 4 months after the acute attack, at which time all had acuity of 20/40 or better (mean 20/25). We found that the patients' SKILL scores were independent of any other measure, suggesting that it measured a unique aspect of vision function. Indeed all patients had abnormal SKILL scores for their affected eyes -- some dramatically so -- with SKILL scores as much as 5 standard deviations from normal; unaffected fellow eyes were commonly abnormal as well. These findings speak to the sensitivity of the SKILL Card in detecting subtle anomalies of vision in recovering optic neuritis.
"Recovered" optic neuritis patients whose measured acuity and contrast sensitivity have returned to normal (or near-normal) often report that their vision is not normal, especially with reduced light levels, or when viewing low contrast objects. They report problems in everyday living situations which are not reflected in the "standard" vision test results. To further assess the sensitivity of the SKILL Card, therefore, a second group of optic neuritis patients was tested further along in their recovery (generally > above ). All were found to have abnormal SKILL scores even with return of central vision (Snellen acuity) and color vision (desaturated D-15) to normal values.
Other Applications and Dissemination of the SKILL Card
The SKILL Card is being incorporated in our testing of vision function, vision performance (driving, reading and walking under dim illumination) and subjective visual experience (via questionnaire) in a group of individuals over the age of 55. This study is being conducted at and in conjunction with the Buck Center for Research on Aging in Marin County, California (see section "Impact of Vision Impairments on Everyday Tasks," above).
In an effort to find a better method for assessing vision as it relates to driving performance, the California Department of Motor Vehicles has administered the SKILL Card, as well as three other tests, to over 1700 people at three of its centers (see "Studies of Driving and Visual Impairment," below). Other researchers interested in vision and driving have now adopted the SKILL Card as an integral part of their vision test battery. These include such well-known driving researchers as Jane Stutts, Ph.D., and JoAnne Wood, Ph.D.
Due to our encouraging preliminary results and the collaboration of Barrett Katz, M.D., of our staff, we were privileged to have the SKILL Card included as a test in the national NIH-sponsored Optic Neuritis Treatment Trial (ONTT) follow-up study. The ONTT has been monitoring visual recovery in over 400 people with optic neuritis. This study, with its many participating physicians, will provide a larger testing ground for the SKILL Card for monitoring the residual visual loss and individuals' subjectively reported visual disability in optic neuritis. It appears that the SKILL Card may afford clinicians an additional easily administered test of optic nerve function which addresses real-world visual impairments.
The Berkeley Disability Glare Tester is being used in our recently commenced study of visual impairment in the elderly. Disability glare is a significantly handicapping phenomenon in the large proportion of elderly individuals with cataracts or media opacities. The impact on daily tasks of this type of visual impairment is being investigated in our study, in collaboration with the Buck Center for Research on Aging in Marin County, California.
We have developed an innovative add-on module for the tester to add a flicker test to the battery of mainly experimental tests in the study. Tests of flicker response, assessing the dynamics of retinal function, have up until now been expensive, special-purpose machines. However, flicker is an important component of a test battery for use in aging studies, as it has been shown to be an extremely sensitive index of retinal integrity. For example, a loss of flicker sensitivity specific to the mid-frequency range is an early visual change in age-related maculopathy -- the leading cause of blindness. Our new approach allows this type of test to be provided as a simple add-on module to an existing instrument, increasing its clinical practicality.
The modification consists of an 8 x 8 array of red light emitting diodes (LEDs) mounted on one of the interchangeable plexiglass diffusers which constitute the front panel of the disability glare tester. The LEDs cover an area of 1.2" x 1.2". The array is placed behind the plexiglass diffuser to give the appearance of a single source of the above-mentioned dimensions.
In addition, the frequency of the modulation can be set between approximately 5 and 80 Hertz. The drive circuit is enclosed in a separate cabinet. The module is extremely flexible, providing for obtaining flicker threshold sensitivity measurements as a function of both frequency and per cent modulation. The test procedure is simple, and the necessary controls consist of two calibrated knobs attached to ten-turn potentiometers for accuracy. The simplicity and efficiency of the device enable it to fit easily into the battery administered by trained volunteers in our large study of vision in 2,000 people over age 55.
It has proven difficult to establish strong associations between driving performance and standard vision test results, despite the obvious dependence of driving on vision. The topic of driving with impaired vision using visual aids is also a controversial one. For example, the most-often cited study found that total visual field, as well as two measures of "night vision" (glare recovery and thresholds at low light level), were not predictive of accident rate and that two others (static and dynamic visual acuity) were only weakly associated with accident rates. Nonetheless, visual acuity is still the primary screening measure employed by the California Department of Motor Vehicles (DMV). The failure of vision tests to predict driving performance in the visually impaired is attributable to (1) the failure of the tests to adequately assess the visual impairment and define the visual limitations of the individual, and (2) the failure of the vision measures to tap the visual task demands of driving.
Under supplementary funding from the California Department of Motor Vehicles, we conducted a study to relate vision functions which are known to decrease with aging to decrements in driving performance, using a battery of functional (practical, low cost) vision tests designed specifically to predict visual performance. We compared vision test scores with the driving records of 97 drivers over 55 years of age. Half of the group had no accidents in the last three years, while the other half had three or more accidents. The vision test battery included visual acuity under high-contrast and low-contrast conditions as well as low-contrast/low-luminance conditions, peripheral visual fields with and without attentional demand, contrast detection for larger targets, sensitivity to disability glare, recovery from exposure to glare, and color vision.
Several visual skills which are known to decrease with age were found to be reduced in the accident-prone subjects. The skills with the strongest statistical relationship to accident involvement included low-contrast/low-luminance visual acuity, disability glare, and the extent of the standard visual field directly down and to the right or to the left. The most highly accident-prone subjects also showed reductions in contrast sensitivity, and their visual skills showed larger vision decrements than the moderately accident-prone drivers.
On the attentional visual field test there tended to be a sharp drop from those with no accidents to those having had accidents. It was not a steady graduation as there was in the regular visual field; this may be a test with more discriminating power.
There was a significant reduction in contrast sensitivity in those people with the highest accident-proneness index. Significant differences were also found in all analyses of the disability glare test.
These results are remarkable in that previous studies of vision and driving -- using conventional tests of vision -- have generally failed to show any strong correlation between accident rates and vision. The new, more subtle and realistic tests utilized in our study, however, enabled us to pick up losses in real-world vision function not noticeable on standard tests. In addition, the number of subjects in our study was very small compared with previous studies on driving, making it even more remarkable that statistically significant differences were found.
It may be that only people with the most severe vision decrements contribute significantly to accidents. So as an adjunct analysis, we compared the people with the ten worst scores to the people with the ten best scores on each test. We found that people with the worst scores on the attentional visual field, standard visual field, and bright glare tests were several times more likely to be in the accident group than in the non-accident group.
People who wore bifocals were also more likely to be in the accident group and at fault than people with only single-vision correction lenses. If this observation is found to hold up in larger samples, and under more study, perhaps people could be advised to use single-vision lenses while driving, or further research could indicate what the problems are using bifocals. In our study, as in previous work, standard acuity tests did not correlate at all with any of the driving measures.
The fact that the new RERC-developed vision tests (designed to simulate real-world visual tasks and detect practical handicaps) were predictive of driving performance was encouraging. It appears that the new tests, in contrast to the traditional Snellen-type acuity tests and other currently used methods, relate more closely to real-world task handicaps.
In collaboration with (and with augmented funding from) the DMV, we have been evaluating the ability of a variety of vision measures designed to overcome those two limitations, to allow us to specify which types of visual impairments are of practical significance in causing accidents. Each of over 1200 individuals, ranging in age from 24 to 91, was tested using all the measures. Three of the measures were letter charts: the Pelli-Robson Contrast Sensitivity test, the SKILL Card, and the Berkeley Glare Tester. Another was a novel Attentional Visual Field Test developed by our RERC; this test uses a modification to a standard perimeter to provide a measure of visual field constriction in the presence of a cognitive task.
To relate these measures to actual driving performance, driving habit surveys were completed by each participant at the time of testing. Driving performance data (on accidents and moving violations) will be provided by the DMV. The data gathered here will also provide a unique visual profile of the California driving population, a type of sample that may differ from others measured previously with these tests. Upon receipt of the DMV accident and violations data for the study participants, we will perform the analyses necessary for publication of the results. The goal will be to specify more precisely which aspects of vision function and impairment are of practical significance in affecting driving performance and accidents.