Implications of low prosthetic contrast sensitivity for delivery of visual information

Abstract

Purpose : To measure contrast sensitivity of the retinal response to subretinal photovoltaic stimulation, and assess its impact on delivery of visual information.

Methods : We measure ex-vivo the full-field contrast sensitivity of healthy rat retina stimulated with white light, and of degenerate rat retina stimulated with a subretinal prosthesis above the flicker fusion frequency (20Hz). We model the effects of fixational eye movements on retinal response by decomposing simulated images on the retina into a static and a dynamic component. We use the measured contrast sensitivity function to predict retinal ganglion cell activity elicited by the dynamic component of the image.

Results : Prosthetic vision exhibits significantly reduced contrast sensitivity and no OFF responses in degenerate retina, with 65% contrast changes required to elicit responses, as compared to 3-7% changes required with visible light. The maximum number of action potentials elicited is, at most, half of its natural counterpart for the ON pathway. Our model predicts that for the majority of natural images the contrast sensitivity of prosthetic vision is insufficient for triggering RGC activity by fixational eye movements. For these images, increasing the contrast alone is insufficient. Therefore, image processing should locally enhance both contrast and sparsity, or stimulation frequencies should be reduced to below the flicker fusion to elicit retinal response.

Conclusions : The low contrast sensitivity and lack of OFF responses hamper delivery of visual information via a subretinal prosthesis. Fixational eye movements, which normally prevent static images from fading, are likely to be insufficient for eliciting responses to natural scenes with subretinal stimulation above flicker fusion frequency. Image pre-processing might alleviate this limitation. Alternatively, subretinal implants may need to rely on lower stimulation frequencies.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.