A model for compound action potentials and currents in a nerve bundle II: A sensitivity analysis of model parameters for the forward and inverse calculations

Abstract

We present a detailed analysis of the sensitivity of simulated Compound Action Current (CAC) and Compound Action Potential (CAP) recordings to specific model parameters, including the Single Fiber Action Currents (SFACs) and Single Fiber Action Potentials (SFAPs) that represent the contributions of each axon in the nerve bundle. In the preceding paper, we described a general method for simulating CACs and CAPs. This method uses a volume conduction model that incorporates the effects of the nerve bundle and other anisotropic properties of the region of the bundle that surrounds an individual nerve axon. In this paper, we present a complete analysis of the effects of incorrectly assigned model parameters on the simulated CAC and CAP. We also investigate the effects of incorrectly assigned parameters, recording noise, and data smoothing on the Conduction Velocity Distributions (CVDs) predicted from the CAC and CAP. We find that the simulated CAC is less sensitive to most of the parameters than is the CAP.