Entropy-based Axon-to-Axon Mutual Interaction Characterization via Iterative Learning Identification

Abstract

This paper presents a novel strategy to characterize the mutual interaction between the neural axons. Based on the mutual coupling factor and assumptions about the recording micro-electrode array, the estimated potential is affected by the coupling factor matrix. In other words, the axon-to-axon interaction can be described by the extended membrane potential model with some unknown coupling factors. To identify these factors, entropy analysis is applied with an iterative learning approach using the simulated data. The advantage of this approach is to avoid solving the nonlinear dynamic membrane potential equation. Simulation results indicate the effectiveness and correctness of our interaction model.