A model of electrical impedance tomography implemented in nerve-cuff for neural-prosthetics control

In neural interfaces for peripheral nerve a trade-off exists between the level of invasiveness and the selectivity of neural recordings. In this study, we implement electrical impedance tomography (EIT) in a nerve cuff with the aim to investigate the achievable level of selectivity.

Established modelling approaches in neural-EIT are expanded on to be used, for the first time, on myelinated fibres which are abundant in mammalian peripheral nerves and transmit motor commands. The model is then used to evaluate the viability of using EIT with a nerve cuff to record neural activity in peripheral nerves.

Fibre impedance models indicate activity in unmyelinated fibres can be screened out from activity in myelinated fibres using operating frequencies above 100 Hz. At 1 kHz the transverse impedance magnitude, which is perpendicular to the fibre length axis, of inactive intra-fascicle tissue and the fraction change during neural activity are estimated to be 1142 Ω cm and  -8.8  ×  10^-4, respectively. At 1 kHz and 10 mm spacing between the impedance measurement electrode pair, the longitudinal impedance magnitude, which is parallel to the fibre length axis, and the fraction change during neural activity are estimated to be 328 Ω cm and  -0.30, respectively. We show that a novel EIT drive and measurement electrode pattern which utilises longitudinal current and longitudinal differential boundary voltage measurements could distinguish activity in different fascicles, as well as simultaneous activity in multiple fascicles, of a three-fascicle mammalian nerve using simulated data.

The results of this study provide an estimate of the transient change in impedance of intra-fascicle tissue during neural activity in mammalian nerve, and present a viable EIT electrode pattern, both of which are critical steps towards implementing EIT in a nerve cuff for a recording neural interface.