Author ORCID Identifier

https://orcid.org/0000-0001-5486-4507

Year of Publication

2021

Degree Name

Master of Science in Biomedical Engineering

Document Type

Master's Thesis

College

Engineering

Department/School/Program

Biomedical Engineering

First Advisor

Dr. Sridhar Sunderam

Abstract

Electroencephalography (EEG) is a widely used technique for monitoring and analyzing brain activity in experimental, diagnostic, and therapeutic applications. Since EEG is sensitive to noise and artefact sources, referential signals at different locations can be combined in different ways to improve signal quality and better localize cortical activity. Four signal derivations were compared against referential EEG in terms of their ability to measure the alpha rhythm modulation (or reactivity) and spatial coherence associated with an eye closure task: a common average reference (CAR), a local average reference (LAR), a large Laplacian (LL), and a focal Laplacian (FL) estimated using a specialized electrode. Results showed significant differences in the alpha reactivity averaged across all electrodes between EEG derivations: the CAR showed significantly greater reactivity than all other derivations while the LL showed significantly lower reactivity compared to all other derivations. No significant differences in alpha reactivity were found between the referential EEG, LAR, and FL when averaged across all locations. LL and FL displayed a trend of increasing alpha reactivity from frontal to occipital regions while the CAR and LAR showed no such trend. The referential EEG showed a linear decrease in spatial coherence as distance increased while the FL showed an exponential decrease. Further, the referential EEG showed no change in spatial coherence related to eye closure while all other derivations showed a significant increase. The focal Laplacian improves detection of alpha reactivity and signal localization without the need for multiple electrodes.

Digital Object Identifier (DOI)

https://doi.org/10.13023/etd.2021.224

Funding Information

This study was supported by National Science Foundation Grant No. 1539068 from 2018-2021.

Available for download on Friday, June 17, 2022

Share

COinS