Fascicle localisation within peripheral nerves through evoked activity recordings: A comparison between electrical impedance tomography and multi-electrode arrays

Nicole Thompson,Cynthia A Chestek,Enrico Ravagli,Kirill Aristovich,Elissa J Welle,David Holder,Svetlana Mastitskaya

doi:10.1016/j.jneumeth.2021.109140

Abstract

BackgroundThe lack of understanding of fascicular organisation in peripheral nerves limits the potential of vagus nerve stimulation therapy. Two promising methods may be employed to identify the functional anatomy of fascicles within the nerve: fast neural electrical impedance tomography (EIT), and penetrating multi-electrode arrays (MEA). These could provide a means to image the compound action potential within fascicles in the nerve. New methodWe compared the ability to localise fascicle activity between silicon shanks (SS) and carbon fibre (CF) multi-electrode arrays and fast neural EIT, with micro-computed tomography (MicroCT) as an independent reference. Fast neural EIT in peripheral nerves was only recently developed and MEA technology has been used only sparingly in nerves and not for source localisation. Assessment was performed in rat sciatic nerves while evoking neural activity in the tibial and peroneal fascicles. ResultsRecorded compound action potentials were larger with CF compared to SS (∼700 μV vs ∼300 μV); however, background noise was greater (6.3 μV vs 1.7 μV) leading to lower SNR. Maximum spatial discrimination between Centres-of-Mass of fascicular activity was achieved by fast neural EIT (402 ± 30 μm) and CF MEA (414 ± 123 μm), with no statistical difference between MicroCT (625 ± 17 μm) and CF (p > 0.05) and between CF and EIT (p > 0.05). Compared to CF MEAs, SS MEAs had a lower discrimination power (103 ± 51 μm, p < 0.05). Comparison with existing methodsEIT and CF MEAs showed localisation power closest to MicroCT. Silicon MEAs adopted in this study failed to discriminate fascicle location. Re-design of probe geometry may improve results. ConclusionsNerve EIT is an accurate tool for assessment of fascicular position within nerves. Accuracy of EIT and CF MEA is similar to the reference method. We give technical recommendations for performing multi-electrode recordings in nerves.

Highlights

The signal-to-noise ratio (SNR) was highest for SS multi-electrode arrays (MEA), carbon fibre (CF) MEAs, electrical impedance tomography (EIT) (~400, 200 and 55, respectively) (Table 1)
It was possible to record reproducible and significant changes related to neural activity with both EIT and MEA probes as both impedance changes and compound action potentials (CAPs) are significantly higher than background noise
CAP recordings from CF probes had higher peak amplitudes compared to SS probes (CF ~700 μV vs SS ~300 μV, +130%); CF probes showed much higher background noise (CF ~6.3 μV vs SS ~1.7 μV, +270%), leading to an overall lower SNR

Summary

Introduction

Detailed maps of peripheral nerve topography and fascicular organisation are required to allow the desired fascicles to be localised accurately for stimulation or recording with electrodes (Aristovich et al, 2019). Two promising methods may be employed to identify the functional anatomy of fascicles within the nerve: fast neural electrical impedance tomography (EIT), and penetrating multi-electrode arrays (MEA). These could provide a means to image the compound action potential within fascicles in the nerve. New method: We compared the ability to localise fascicle activity between silicon shanks (SS) and carbon fibre (CF) multi-electrode arrays and fast neural EIT, with micro-computed tomography (MicroCT) as an independent reference. We give technical recommendations for performing multi-electrode recordings in nerves

Objectives

Methods

Results

Conclusion