Quantification and parameterization of the dynamic spinal evoked potential: Towards advanced neuromonitoring and neurotechnologies

Miranda, Priyanka

Full Metadata

Overview

Year:: 2023
Contributor:: Miranda, Priyanka (creator); Borton, David (Advisor); Lee, Jonghwang (Reader); Truccolo, Wilson (Reader); Brown University. Biology and Medicine: Biomedical Engineering (sponsor)
Genre:: theses
Subject:: Computational neuroscience; Spinal Cord Stimulation; Spinal cord; neuromonitoring
Extent:: ix, 82 p.

Files

Description

Abstract:: Spinal neuromonitoring is a medical technique that involves monitoring the function of the spinal cord and nerve roots during surgery or other procedures to minimize the risk of injury. This thesis explores an alternative method to the current spinal neuromonitoring techniques, with improved ability to identify and quantify changes in the spine intraoperatively. The architecture of this system is designed with the motivation of developing a closed loop system with continual adaptive spinal monitoring. The methodology described herein analyzes spinal cord evoked potentials recorded from the dorsal epidural space, above and below the site of operation. Spinal cord evoked potentials are highly informative and consist of substantial combined information from the ascending and descending tracts. Parameters, such as conduction velocity, that define the flow of information along the spinal cord, are computed from the raw electrophysiological signals. Although these continuously changing values present heterogeneously across subjects, differences from baseline can be calculated to measure intraoperative changes. A clustering model was built to evaluate the efficacy of these carefully curated features, primarily, the conduction velocity, the rostral and caudal first peak (P1) and first negative peak (N1) amplitudes and time lags. This pipeline can be used as a black box to monitor and quantify changes in spinal evoked potentials. To validate the feature extraction pipeline, spinal cord injuries were delivered during terminal surgeries of sheep. The model trained with our feature set shows an average of 22% improvement over the model trained on raw electrophysiological response for both highly discernable states such as pre-injury vs post injury and low discernable states like 30 minutes post injury vs 1 hour post injury. The key takeaways from these results are the feature set extracted that describe spinal ECAPs and their potential use in spinal monitoring. This method of intraoperative analysis of spinal communication can be considered alongside existing methods to improve the specificity and quantify changes in spinal conduction. In addition to measuring intraoperative changes in the spinal evoked potentials, this method can be used to analyze the effect of analgesics on nerve conduction in the spine, to quantify spinal flow changes in different stages of the gait cycle, and provide quantitative neurophysiological information across a wide range of tasks. Applications of this method in the continuous monitoring of the spine in closed loop systems is also a possible future implementation.
Notes:: Thesis (Sc. M.)--Brown University, 2023

Content

Citation

Miranda, Priyanka, "Quantification and parameterization of the dynamic spinal evoked potential: Towards advanced neuromonitoring and neurotechnologies" (2023). Biomedical Engineering Theses and Dissertations. Brown Digital Repository. Brown University Library. https://repository.library.brown.edu/studio/item/bdr:5ubsc9s7/

Relations

Collection:

Biomedical Engineering Theses and Dissertations

Theses and Dissertations for the Biomedical Engineering department.
...