Bioelectric nerve fiber to electrode coupling for unit identification and tracking

Shaoyu Qiao, Purdue University

Abstract

Neuroprosthetics is a technology which aims to help those with neurological disorders or injuries to restore or improve function by accessing and using the residual capacity of the nervous system. Critical in this are neuroprosthetic interfaces which monitor or influence the electrical activity in neural tissues. Selective advanced neuroprosthetic electrodes are currently capable of detecting and measuring the electrical activity of small sets of nerve fibers surrounding the electrode superimposed upon the interference waveforms of the activities of more distant nerve fibers. However, current techniques for monitoring the health of neuroprosthetic devices are limited to post-hoc analyses of excised tissues and devices, and the statistical averaging of many instances to generate a picture of what happened. Monitoring the active nerve fiber to electrode coupling could provide a longitudinal, non-destructive, and high-resolution means to track the interaction between nerve, tissue and electrode. To test this hypothesis, we first developed a three dimensional bioelectrical volume conductor finite element method (FEM) model of a thin-film recording electrode residing in a nerve fascicle to explore the potential distribution in the nerve fascicle and further derive the electrode-fiber coupling function in the time-frequency domain. It was found that the spectral distribution of the single fiber action potential (SFAP) was multimodal in nature, and could be predicted by taking the single fiber action currents (SFACs) filtered by the electrode-fiber coupling function. This function manifested itself as a Bessel-like low-pass filter, dependent upon the fiber's location relative to the electrode and conduction velocity, determining the shape of the SFAP waveform which is a unique footprint for identifying and tracking the nerve unit. Taking advantage of variations in the power spectral density (PSD) of the SFAPs, an approach that enables the differential quantification of electrode-fiber distance and unit conduction velocity through spectral analysis of the single unit waveform was developed and demonstrated in an in vitro two-fiber earthworm peripheral nerve model. The utility of the approach was further explored using an in vivo rabbit peripheral nerve model. The spatial and temporal information about nerve fibers generating the SFAP detected by neuroprosthetic interfaces from multi-unit records originating from thin-film intrafascicular electrodes implanted in the mammalian peripheral nerve were extracted and characterized. This work is foundational and the implications of the results extend beyond in situ fiber identification and tracking to in situ diagnostics of the nerve fibers and implanted electrodes, micro-scale in situ monitoring of the interaction between nerve fibers and neural interface, and assessment of the biocompatibility of the neural interface and the health of the reporting nerve fibers.

Degree

Ph.D.

Advisors

Yoshida, Purdue University.

Subject Area

Neurosciences|Biomedical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS