Development of Retinal Ganglion Cell Model to Assess Optimal Stimulation Parameters

Abstract

Purpose:: The Boston Retinal Implant Project is developing a sub-retinal prosthesis designed to excite surviving retinal cells in the retina of AMD and RP patients. Two important design criteria for implanted electrodes are the magnitude and field of excitation they create.As an aid to defining the appropriate values for these criteria, a computer simulation is being developed in order to observe realistic simulated effects of different design approaches.

Methods:: We built a simulation of retinal ganglion cell excitation using Microsoft Visual Studio 2005. This gave two advantages over existing available neuron simulations written in a script language: higher performance and detailed tailoring to our problem. The basic model of Hodgkin and Huxley, as implemented on the MIT Project Athena network was used as a starting point. The model was expanded by explicit simulation of metabolic currents and their effect on resting potential and explicit simulation of various other ionic currents.The general approach is to simulate ionic concentrations and impressed electrode charge distributions as the basic determining variables, from which we compute the resulting force fields. Drift and diffusion currents are then computed, along with the resultant charge distribution changes. Various integration schemes are evaluated.

Results:: At this time the model has been completed to the point of showing the steady state ionic currents that produce the resting potential. A metabolic current of 400,000 ions/sec-uM2 has been shown to be required to support a resting potential of -72 mV. We have also shown a graphical solution to the LaPlacian that depicts the static potential distribution. Planned further work will produce estimates of the effect of external field strength required to produce action potentials. Preferred electrode geometries will also be determined.

Conclusions:: The computer model has been shown to have potential as an efficient and flexible tool for the prediction of important design criteria for the design of retinal implants. The computer model will be continuously expanded and tested for accuracy of its conclusions.

Keywords: computational modeling • retinal connections, networks, circuitry • ganglion cells