A model of electrical impedance tomography on peripheral nerves for a neural-prosthetic control interface

摘要

Objective: A model is presented to evaluate the viability of using electricalimpedance tomography (EIT) with a nerve cuff to record neural activity inperipheral nerves. Approach: Established modelling approaches in neural-EIT areexpanded on to be used, for the first time, on myelinated fibres which areabundant in mammalian peripheral nerves and transmit motor commands. Mainresults: Fibre impedance models indicate activity in unmyelinated fibres can bescreened out using operating frequencies above 100 Hz. At 1 kHz and 10 mmelectrode spacing, impedance magnitude of inactive intra-fascicle tissue andthe fraction changes during neural activity are estimated to be 1,142{\Omega}.cm and -8.8x10-4, respectively, with a transverse current, and 328{\Omega}.cm & -0.30, respectively with a longitudinal current. We show that anovel EIT drive and measurement electrode pattern which utilises longitudinalcurrent and longitudinal differential boundary voltage measurements coulddistinguish activity in different fascicles of a three-fascicle mammalian nerveusing pseudo-experimental data synthesised to replicate real operatingconditions. Significance: The results of this study provide an estimate of thetransient change in impedance of intra-fascicle tissue during neural activityin mammalian nerve, and present a viable EIT electrode pattern, both of whichare critical steps towards implementing EIT in a nerve cuff for neuralprosthetics interfaces.