A hybrid continuum-discrete computational model of electrical stimulation of the retinal network

Computational neuroscience can provide powerful insights into the design and optimization of neuroprosthetic devices. This study combines the advantages of two numerical approaches to develop a predictive model of electrical stimulation of the retinal network for visual prostheses. A continuum bidomain formulation was used to model neurons in the retinal inner nuclear layer, which provide synaptic inputs to a population of discrete morphologically-realistic retinal ganglion cells. The ability to model bulk tissue activation for selective layers in the retina, as well as dendritic processing in ganglion cells provides a powerful simulation tool to test and optimize electrical stimulation strategies for future retinal prosthesis devices.