Date of Award

Summer 2014

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Electrical/Computer Engineering

Committee Director

Dean J. Krusienski

Committee Member

Jiang Li

Committee Member

Christian Zemlin

Committee Member

Frederic D. McKenzie

Abstract

A Brain-Computer Interface (BCI) is a system that allows people with severe neuromuscular disorders to communicate and control devices using their brain signals. BCIs based on scalp-recorded electroencephalography (s-EEG) have recently been demonstrated to provide a practical, long-term communication channel to severely disabled users. These BCIs use time-domain s-EEG features based on the P300 event-related potential to convey the user's intent. The performance of s-EEG-based BCIs has generally stagnated in recent years, and high day-to-day performance variability exists for some disabled users. Recently intracranial EEG (i-EEG), which is recorded from the cortical surface or the hippocampus, has been successfully used to control BCIs in experimental settings. Because these recordings are closer to the sources of the neural activity, i-EEG provides superior signal-to-noise ratio, spatial resolution, and broader bandwidth compared to s-EEG. However, because i-EEG requires surgery and the long-term efficacy for BCIs must still be explored, this approach is still not an option for patients. In order to improve s-EEG BCI performance, it is important understand the underlying brain phenomena and exploit the relationships between the s-EEG and generally superior i-BEG signals. Because the human head acts as a volume conductor consisting of the brain, cerebrospinal fluid, skull, and scalp tissue, linear mathematical models can be used to relate s-EEG and i-EEG. This dissertation presents unique s-EEG and i-EEG data that were recorded from the same subjects and used to develop novel empirical models to estimate s-EEG from i-EEG. These new empirical models can be used to better understand the sources and propagation of the relevant neural activity, as well as to validate existing theoretical volume conduction models. It is envisioned that this knowledge will help to advance algorithms for improving s-EEG BCI performance.

DOI

10.25777/hmpg-x793

ISBN

9781321301953

Share

COinS