Name:
Smith MSCI thesis 11_29_22_rev ...
Size:
406.7Kb
Format:
PDF
Description:
Master's thesis
Authors
Smith, Kara MFaculty Advisor
Robert Brown MD DPhilAcademic Program
Master of Science in Clinical InvestigationDocument Type
Master's ThesisPublication Date
2022-11-29
Metadata
Show full item recordAbstract
Background: Cognitive impairment is a common non-motor symptom of Parkinson’s disease (PD), but there are currently inadequate tools available to detect and monitor this complication. Speech undergoes an array of changes in individuals with PD, and speech may become more impaired as cognition declines. Speech markers may serve as useful, easy-to-access proxies of cognitive function. We evaluated differences in speech acoustic features on a reading task and a picture description task, as well as accuracy and pausing on a Stroop sentence task in persons with and without PD. We also assessed whether speech markers from these tasks were associated with mild cognitive impairment in persons with PD (PD-MCI). Methods: We enrolled participants with PD (n=44) and older adult controls (n=8) at the University of Massachusetts Chan Medical School between January 2020 and October 2022. PD participants underwent cognitive testing in order to categorize cognitive status as mild cognitive impairment (PD-MCI) or normal cognition (PD-NC). All participants were audio recorded while completing a protocol of speech and language assessment, and speech data were processed and analyzed to obtain several acoustic features. Results: Standard acoustic measures did not differ significantly between PD-MCI and PD-NC. Performance on reading and picture description tasks worsened over the course of speaking in both groups. Variability in fundamental frequency declined over the course of speaking in the PD-MCI group compared to the PD-NC group during a picture description task. In a Stroop sentence task, accuracy and pausing measures were similar between PD-MCI and PD-NC groups. Conclusions: Novel speech markers may be able to detect PD-MCI, but optimal speech task selection and analytic approach requires further development. Speech markers hold promise in monitoring PD symptoms, since data may be obtained frequently, remotely, and processed using automated algorithms.DOI
10.13028/p1jq-jc59Permanent Link to this Item
http://hdl.handle.net/20.500.14038/51444Rights
Copyright © 2022 SmithDistribution License
https://creativecommons.org/licenses/by/4.0/ae974a485f413a2113503eed53cd6c53
10.13028/p1jq-jc59