Honors College Thesis
 

Decoding EEG Signals to Detect and Classify Hand Movement

Public Deposited

Downloadable Content

Download PDF
https://ir.library.oregonstate.edu/concern/honors_college_theses/44558n22z

Descriptions

Attribute NameValues
Creator
Abstract
  • Electroencephalogram (EEG) signals have recently gained attention to classify associated hand movement. This is particularly useful for amputees with prostatic hands or people who are unable to move their hands voluntarily. Among the decoding algorithms used to classify hand movements are support vector machines and multilayer perceptron networks. In this thesis we apply power spectral density and principal component analysis on EEG data to classify EEG readings into one of three groups: right-hand movement, left-hand movement, and no hand movement. We employ a one vs. all approach using three distinct support vector machines and a single multiclass multilayer perceptron network to classify the data. Utilizing collected EEG and hand position data across four healthy subjects, we achieve an average accuracy of 56.2% using a single multi-class multilayer perceptron network across all of the four subjects. For the support vector machine approach, across all of the four subjects we arrive at average accuracies of 68.9%, 72.0%, and 76.1% for no movement, left movement, and right movement against all other classes, respectfully. Combining these models, we achieve an overall accuracy of 55.6% for correctly classify the processed EEG signals.
  • Key Words: electroencephalogram, EEG, signal processing, machine learning, power spectral density, principal component analysis
Resource Type
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Non-Academic Affiliation
Rights Statement
Publisher
Peer Reviewed
Language

Relationships

Parents:

This work has no parents.

In Collection:

Items

Thumbnail Title Date Uploaded Visibility Actions
file details McGrathPatrick2021.pdf 2021-06-04 Public Download