Perceptually localized phosphenes, resulting from electrode-based microstimulation of the visual pathway, are reported as unnatural in appearance by observers. There are no systematic methods to artificially encode naturally occurring visual stimuli into the brain, or for stimulating optimized spatiotemporal cortical patterns that evoke naturalistic visual responses, including those related to eye movements, with normal contrast sensitivity and visual acuity.

Major gaps in knowledge include (1) the lack of spatiotemporal kernels to transform video images into a correlated activation map within the ascending visual system, specifically between the lateral geniculate nucleus (LGN) to the primary visual cortex (V1); (2) the failure to account for corollary signals from oculomotor areas or top-down inputs from extrastriate cortex to V1; and (3) a technique to determine the visual functional anatomical map in a blind patient, a crucial step for prosthetically stimulating an individual’s cortical circuits to achieve full visual restoration.

This project aims to (1) develop the computational theory and methods necessary to stimulate LGN inputs to V1 in a targeted pattern that accounts for both bottom-up and top-down signals; (2) optimize the system to activate LGN inputs into V1 optogenetically, and to determine the spatiotemporal transform for visual perception and eye-movement-related top-down signals, and (3) develop techniques to map visual cortex without visual activation, using solely optogenetic stimulation.