Abstract
PURPOSE. This study compared the performance of subjects with simulated low-resolution prosthetic vision on basic visual tasks to that of blind patients implanted with 4x4 epiretinal prostheses.
METHODS. Nine participants with at least 20/30 monocular visual acuity performed seven independent forced-choice basic-visual tasks previously tested on three blind subjects implanted with 4x4 epiretinal prostheses. These tasks were performed under 4x4, 8x8, and 6x10 pixel resolutions and 256 levels of grayscale. After training, each subject completed 40 repetitions of each task:resolution combination in a random order determined by a customized C++ computer program. Performance was recorded as stimulus presented and answer provided and the time required to provide each answer.
RESULTS. Of the 40 repetitions conducted at each setting, the number correctly identified divided by 40 provided the % correct. Participants performed similarly to implant patients at the 4x4 setting and quite better (>93% accuracy and up to twice as fast for all tasks tested) at the 8x8 and 6x10 settings. Variance in performance at the 4x4 setting was large and appears to be a function of the age of the subject tested. No significant difference in performance was observed between 8x8 and 6x10 settings.
CONCLUSIONS. These results imply that sighted subjects wearing a simulator with our parameters perform visually-based tasks at a level similar to that of blind patients implanted with 4x4 epiretinal prostheses. Age-dependent differences in performance cannot be differentiated between intrinsic ability to perform visually-based tasks and time needed to adapt to new visual conditions.