Seeing with subretinal electronic implants in blind volunteers requires direct link of image receiver to eye movement

Purpose To compare subretinal electrode‐generated percepts elicited by stationary stimulation with those elicited by light‐sensitive subretinal arrays during patient initiated natural eye movements. Methods The Tuebingen subretinal implant consists of two arrays: a) 4 x 4 electrode field controlled externally for direct electrical stimulation (DS array), b) a light sensitive “CHIP” (3 x 3 x 0.1mm) with 1500 photodiodes, amplifiers and electrodes. Results Patterns presented via DS electrodes repeatedly to the same retinal position disappear after seconds. With the light sensitive subretinal CHIP, however, objects like grating patterns, tableware or letters can be perceived continuously (repeated at 7 Hz for 1 ms. Clearly objects scanned with natural eye movements via light sensitive retinal CHIP activate a range of adjacent pixels on the CHIP, i.e. eye movements and microsaccades continuously shift the “electrical image” on the retina, thus preventing mechanisms of image fading Conclusion The results clearly support the feasibility of light sensitive subretinal multi‐electrode devices for restoration of useful visual percepts in blind patients. Continuous stimulation of fixed electrodes can provide only quickly fading percepts (similar to the Troxler effect). In contrast, percepts of images constantly moving across the subretinal light‐sensitive chip – due to saccadic and fixational eye movement ‐‐ remain visible without measurable temporal limitation. We conclude that real‐time information of eye movements (an efference copy) is required by the brain to generate stable percepts. This is only available if the light sensitive part of the implant moves exactly with the eye, as presently realized only in the subretinal Tuebingen CHIPCommercial interest