A Neuro-behavioral Test and Algorithms for Quantification of Sleepiness and Characterization of Wake-Sleep Transitions

摘要

A neuro-behavioral test has been developed that quantifies sleepiness by means of a convenient and relatively short-duration (15 minute) protocol, and using a class of algorithms that track wake to sleep transitions. Utilizing an ambulatory EEG monitoring combined with an auditory-based vigilance test, this method acquires both behavioral response (i.e., reaction time) and EEG waveforms that are simultaneously analyzed to produce a time-varying EEG-based drowsiness-sleepiness index (DSI). DSI index is computed from the time-varying power spectrum of EEG using an algorithm whose coefficients were determined based on an optimization procedure that maximized the correlation between the time profiles of DSI and those of the behavioral response. Sleepiness is quantified using an algorithm based on the consistency of the neuro-behavioral indices over the test period as well as the degree and profile of their degradation when the subject is incapable of maintaining alertness. This test was administered to 5 volunteers (separate from the subjects whose data were used for the above optimization of DSI) under the conditions of adequate prior night sleep and sleep restriction. Cross correlation analysis showed a close agreement between the time profiles of EEG-based DSI and behavioral response indices. Statistical analysis showed significant difference between the alert and sleepy conditions for the whole group as well as in each individual subject