Date of Award

12-2021

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Computer Engineering

Committee Chair/Advisor

Dr. Yongkai Wu

Committee Member

Dr. Eric Patterson

Committee Member

Dr. Richard Brooks

Committee Member

Dr. Nathan Rowland

Abstract

Electroencephalography (EEG) is a non-invasive technique used in both clinical and research settings to record neuronal signaling in the brain. The location of an EEG signal as well as the frequencies at which its neuronal constituents fire correlate with behavioral tasks, including discrete states of motor activity. Due to the number of channels and fine temporal resolution of EEG, a dense, high-dimensional dataset is collected. Transcranial direct current stimulation (tDCS) is a treatment that has been suggested to improve motor functions of Parkinson’s disease and chronic stroke patients when stimulation occurs during a motor task. tDCS is commonly administered without taking biofeedback such as brain state into account. Additionally, the administration of tDCS by a technician during motor tasks is a tiresome process. Machine learning and deep learning algorithms are often used to perform classification tasks on high-dimensional data, and have been successfully used to classify movement states based on EEG features. In this thesis, a program capable of performing live classification of motor state using machine learning and EEG as biofeedback is proposed. This program would allow for the development of a device that optimally administers tDCS dosage during motor tasks. This is achieved by surveying the literature for motor classification techniques based on EEG signals, recreating the methods in the surveyed literature, measuring their accuracy, and creating an application to perform online capturing and analysis of EEG recordings using the classifier with the highest accuracy to demonstrate the feasibility of real-time classification. The highest accuracy of motor classification is achieved by training a random forest on binned spectral decomposition from a normalized signal. While live classification was successfully performed, accuracy was limited by external changes to the recording environment, skewing the input to the trained model.

Author ORCID Identifier

https://orcid.org/0000-0001-7511-6057

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.