Recording properties of an electrode implanted in the peripheral nervous system: A human computational model

Chronically implanted neural interfaces are aiming to create an intimate and long-term contact with neural cells. This would potentially allow the development of neurocontrolled artificial devices. However the precise nature of the compound signal recorded and the interaction between the neural population and the electrode are still poorly understood. Consequently there is limited knowledge available on the optimal strategy for the design of peripheral electrodes in order to achieve high information harvest while insuring long-term reliability of the device. In this paper, we introduce a novel integrated hybrid Finite Elements/Biophysical model for recording, based on anatomical data of the human median nerve. Using this model, we simulated the signal recorded intrafascicularly with implanted Transversal Intraneural Multichannel Electrode (TIME). The preliminary results help in understanding the properties of recorded signals and suggest that a substantial portion of the spikes detected with electrodes implanted in the peripheral nervous system might actually be multi-unit events formed by the superposition of several fibers activity.