Project Summary / Abstract The most promising treatment option for photoreceptor degeneration, which is the leading cause of blindness in the United States, are retinal prostheses that are surgically implanted on the anterior retina in order to gain electrical access to the retinal ganglion cells (RGCs), bypassing the damaged photoreceptor cell layer. Despite there being a few FDA approved epiretinal prostheses on the market, the high density of RGCs at the surface of the retina makes it difficult to deliver current from electrodes with enough precision to recapitulate the natural patterns of ganglion cell activity, and thus useful visual perception for blind patients. This is due to coarse electrical stimulation that evokes unwanted activity in cells and axons at the epiretinal surface, especially in the central retina where cells are the densest. The goal of this research is to use information from the natural RGC activity recorded on the multi-electrode array in order to guide precise, spatially targeted electrical stimulation. Careful application of the properties of electric fields propagating in tissue to calibrate stimulation currents will allow epiretinal implants to produce meaningful visual perception in blind patients. The relationship between the recorded signature of a given RGC’s activity on the array—its Electrical Image (EI)—and its sensitivity to single or multi-electrode stimulation—its Electrical Receptive Field (ERF)—will be determined. First, experiments will be conducted in peripheral primate retina to collect data on RGC activation characteristics in response to delivering current from each of the ~500 electrodes on the array. Next, we will extend this model to experimental data collected on the smaller, densely-packed RGCs in the central Raphe region of the retina. Despite dense electrode spacing, arrays are limited in their ability to deliver precisely targeted stimulation by the distance between electrodes. This can be addressed by weakly stimulating with multiple neighboring electrodes at the time same, pushing the strength of stimulation current at the intersection of the multiple generated electric fields over the threshold required for target RGC activation. We will stimulate with combinations of two to seven neighboring hexagonally arranged electrodes to collect RGC ERFs in the peripheral and central retina. An cascading linear-nonlinear model, popular for modeling neuronal spiking, will be fit with EIs as input to predict ERFs across the array. A thorough understanding of the EI-ERF relation in the retina will enable the closed-loop, precise, spatially localized stimulation necessary for designing a high-fidelity epiretinal device, and uncover general principles applicable elsewhere in the central nervous system.