Pattern Recognition of EEG to Determine Level of Alertness
Technical progress rept. no. 4, 15 Nov 1969-14 May 1970
MCDONNELL DOUGLAS ASTRONAUTICS CO NEWPORT BEACH CA ASTROPOWER LAB
Pagination or Media Count:
The report documents the work accomplished during the fourth reporting period in applying the principles of pattern recognition technology to the analysis of EEG. Several additional sleep scoring decision systems were investigated. Most classifications are performed satisfactorily. However, no reliable single channel frequency relationships have been discovered for separation of Stage 1 and REM. Also the transitional nature of Stage 3 patterns leads to confusing these patterns with Stage 2 or Stage 4. Cluster analysis was applied to frequency indicators derived from an overnight sleep record. Selection of clusters independent of human scoring demonstrated well defined groupings for awake, Stage 1 and REM combined, Stage 2 and Stage 4. Stage 3 patterns were split between the Stage 2 and Stage 4 clusters. Sleep scoring decision systems based upon frequency analysis of a single channel of EEG have been designed for each of two Tektite I aquanauts.
- Anatomy and Physiology