Back to search

Boston University

Development and evaluation of computational methods for measuring free-living gait and uncovering neuropathology in Parkinson’s disease

Abstract

dc:description.abstract

Novel advances in engineering and data analytics are revolutionizing both our ability to monitor Parkinson’s disease (PD) patient symptoms and our understanding of neuropathology. Despite promise, key challenges exist before patient monitoring technologies become standard in clinical settings, including 1) industry standardization of sensor-based analytical approaches; 2) validation of endpoint sensitivity to degree of impairment and medication state; and 3) consensus regarding appropriate devices, algorithms, data requirements, and statistical analysis requirements for symptom measurement outside of the clinic. In addition to the need for better patient monitoring, no disease-modifying therapeutics currently exist and thorough understanding of the neuropathology of PD remains elusive. To this end, large network brain simulations that leverage efficient computational frameworks are beginning to provide insight into mechanisms that facilitate pathological oscillations and may serve to identify new therapeutic targets. To address current limitations in patient monitoring and our understanding of neuropathology, in this dissertation I 1) develop and evaluate validity and reliability of an open-source, wearable sensor-based algorithm for measuring gait in PD patients, 2) evaluate and compare sensitivity of at-home measurements relative to in-clinic measurements, 3) evaluate sensitivity of wearable-derived features for measuring degree of gait impairment and treatment response in PD patients, and 4) investigate the effect of synaptic parameters on beta synchrony and entrainment in a large-scale spiking model of the subthalamic nucleus-globus pallidus externa (STN-GPe) network of the basal ganglia. Importantly, I find that sensor-derived features derived from the at-home environment differ from and are more sensitive to small changes compared to in-clinic, traditional assessments. Furthermore, I demonstrate the capacity for a single, lower back sensor-based algorithm to estimate gait features with sufficient sensitivity to detect degree of gait impairment and treatment effect in a mild-to-moderate PD population. Lastly, I demonstrate that weak synaptic connections between STN and GPe allows the STN-GPe network to entrain to a wide range of frequencies outside of the beta range, thus elucidating constraints on conditions required for beta production. Together, my work provides new insights into the feasibility and benefits of sensor-based symptom monitoring and PD-related neuropathology.

Degree

thesis:*
Name thesis:degree_name
Doctor of Philosophy
Level thesis:degree_level
doctoral
Discipline thesis:degree_discipline
Anatomy & Neurobiology
Grantor
Boston University
Year dc:date.issued
2022

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Czech, Matthew
Advisor dc:contributor.advisor
  • Thomas, Kevin
ORCID dc:identifier.orcid

Subjects

dc:subject × 4

Rights

Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/2144/44021

Chain of custody

source
Harvested from
Boston University
Base URL
open.bu.edu/oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Czech, Matthew. Development and evaluation of computational methods for measuring free-living gait and uncovering neuropathology in Parkinson’s disease. doctoral thesis, Boston University, 2022. https://hdl.handle.net/2144/44021