Analysis of evoked EEG patterns from controlled cognitive states /
| Main Author: | |
|---|---|
| Other Authors: | , , |
| Format: | Thesis Book |
| Language: | English |
| Published: |
1988.
|
| Subjects: | |
| Online Access: | ProQuest, Abstract Link to OAKTrust copy |
| Abstract: | The effects of cognitive visual stimuli on EEG patterns were studied to establish relationships between the electroencephalogram (EEG) and cognitive states evoked by visual stimuli as a feasibility study on the implementation of brain wave robotic systems. Human subjects were asked to perceive provocative visual stimuli while EEG data were being measured and recorded on analog magnetic tape for later digitization and data analysis using the ASYST scientific software system. Two types of spectral parameters (relative and absolute power difference) were calculated from the power spectra of control and stimulus EEG activity for the delta, theta, alpha, and beta frequency bands (1-50 Hz) to characterize spectral differences in the EEG. The rectangular and Parzen data windows were used to measure extreme effects of data tapering on parameters. "EEG templates" were developed to describe parameter variability using 95% mean confidence intervals, and the general linear analysis of variance (GLM ANOVA) model was used to establish significant differences in parameter means between stimuli. The experiment was designed mainly for the analysis of within-subject variability. Results showed that (1) all subjects exhibited reproducible EEG signatures for several stimuli, (2) subjects varied in their abilities to produce signatures, (3) the absolute power difference parameters showed less variability than the relative power difference parameters, and the rectangular windowed parameters showed less variability than the Parzen windowed parameters, (4) the alpha and betal frequency bands for the rectangular windowed absolute power difference parameters showed the least amount of variability, (5) the electrode locations that showed the least amount of variability were T3, T6, C3, P3, P4, O1, and O2, and (6) stimuli differed significantly in their EEG signatures mainly in the alpha and beta bands. Most importantly, results demonstrated that the parameters of choice (smallest amount of variability) for characterizing the EEG in real-time were those calculated using the absolute power difference data processing technique and the rectangular data window. |
|---|---|
| Item Description: | Typescript (photocopy). Vita. "Major subject: Bioengineering." |
| Physical Description: | xii, 303 leaves : illustrations ; 29 cm |
| Bibliography: | Includes bibliographical references. |