Understanding the Role of M1 in Locomotion: Population Dynamics and Applications to Brain-Machine Interfaces

Xing, David Yuxuan

Full Metadata

Overview

Year:: 2021
Contributor:: Xing, David Yuxuan (creator); Borton, David (Advisor); Truccolo, Wilson (Reader); Hochberg, Leigh (Reader); Sheinberg, David (Reader); Miller, Lee (Reader); Brown University. School of Engineering (sponsor)
Genre:: theses
Subject:: Biomedical engineering; Motor cortex; Neuroengineering; Computational neuroscience; Locomotion; Rhesus monkey
Extent:: 16, 158 p.

Files

Description

Abstract:: One of the hallmarks of our nervous system is its ability to generate a wide variety of movements across a diverse set of contexts and environments. The motor cortex is a crucial component of the central nervous system in controlling and generating these movements. However, while the role of motor cortex in voluntary actions generated by the forelimbs has been an area of significant study over the past few decades, its involvement in the control of the lower limbs, particularly in locomotion, is still poorly understood. It is important to understand how leg movements are generated in motor cortex not only for scientific interest, but also for developing brain-machine interfaces (BMIs) designed to restore lost hind-limb functionality. The goal of this thesis is to discuss the development of novel BMIs for the rehabilitation and restoration of locomotion as well as the recent research characterizing the activity and dynamics of primary motor cortex (M1) during locomotor behaviors. In particular, this thesis describes the development of a brain-spinal interface utilizing M1 activity to drive epidural spinal stimulation in rhesus macaques. Using a linear discriminant analysis classifier, we were able to predict gait events from intracortical M1 recordings with high accuracy and generate temporally controlled spinal stimulation which restored stepping after a thoracic lesion. Additionally, this thesis examines the characterization of population dynamics in M1 during locomotion, and the comparison of neural activity during locomotion to neural activity during voluntary actions such as stepping over an obstacle. Using a Weiner filter decoder, we were able to fully decode hind-limb kinematics from M1 neurons, and that low-dimensional dynamics of only 12 dimensions were required to achieve accurate decoding. Furthermore, the neural activity appears to lie in different subspaces during stationary voluntary movements compared to basic locomotion, suggesting that M1 may be engaged differently during these two behaviors. These findings not only further our understanding of how leg movements are represented in primary motor cortex, but also inform future decoder designs for BMIs aiming to restore not just locomotion, but all aspects of hind-limb movement.
Notes:: Thesis (Ph. D.)--Brown University, 2021

Content

Citation

Xing, David Yuxuan, "Understanding the Role of M1 in Locomotion: Population Dynamics and Applications to Brain-Machine Interfaces" (2021). Engineering Theses and Dissertations. Brown Digital Repository. Brown University Library. https://repository.library.brown.edu/studio/item/bdr:bsvmsbx6/

Relations

Collection:

Engineering Theses and Dissertations

Theses and Dissertations for the Engineering department.
...