Abstract
Abstract: :
Purpose: Anatomical studies indicate that, in early age-related macular degeneration (AMD), individual or small bundles of retinal photoreceptors undergo degeneration. Since most vision test stimuli contain redundant information, they are not sensitive to moderate amounts of photoreceptor loss. The Degraded Texture Discrimination Test (DTDT) uses degraded texture stimuli that contain barely enough information to support a given level of discrimination performance for subjects with intact retinas. When patients with photoreceptor loss are tested, the amount of available information is further reduced on the retina, thus their discrimination performance should be inferior to normal subjects. The potential of DTDT in early detection of AMD was studied. Methods: To evaluate DTDT, three groups of subjects, young normal, old normal and early AMD patients, were tested. Early AMD is defined as soft drusen or hard/soft drusen with RPE changes and with visual acuity of at least 20/60. The stimulus was a pair of black-and-white textures positioned side-by-side. One member was a random texture. The other was either a texture with obvious structure or a random texture. The texture pair was degraded by setting randomly selected pixels to black. There were 9 levels of degradation, ranging from 10% to 60% pixel deletion. Each texture pair was presented for 250 msec and the observer decided whether the pair was the same or different. Three types of texture renderings, black/white checks, white dots on black background and blurred black/white checks, were tested. Discrimination performance was measured by the d' at each degradation level. Prior to the test, a training session was given to the subject to familiarize him/her with discriminating degraded textures. Results: Preliminary results show that young and old normal subjects perform similarly in DTDT. Their discrimination was nearly perfect when degradation was less than 35%, but decayed with further texture degradation. Early AMD patients showed inferior discrimination compared to the other two groups with check and dot textures. With the dot textures, their d' scores were similar to the normals at 10% degradation but generally lower at greater levels of degradation. With check textures, d' scores for AMD subjects were lower than those of normals across all levels of degradation. Differences between AMD subjects and normals were minimized with the blurred checks, possibly due to the spread of useful information across space. Conclusions: DTDT, especially with checks and dots, shows promise for detecting AMD in its earliest stages.
Keywords: age-related macular degeneration • clinical (human) or epidemiologic studies: sys • texture