Macrostructural EEG characterization based on nonparametric change point segmentation: application to sleep analysis

In the present investigation a new methodology for macrostructural EEG char acterization based on automatic segmentation has been applied to sleep anal ysis. A nonparametric statistical approach for EEG segmentation was chosen, because it minimizes the need for a priori information about a signal. The method provides the detection of change-points i.e. boundaries between qua si-stationary EEG segments based on the EEG characteristics within four fun damental frequency bands (delta, theta, alpha and beta). Polysomnographic d ata of 18 healthy subjects were analyzed. Our findings show that nonparamet ric change-point segmentation in combination with cluster analysis enables us to obtain a clear picture of the hierarchical macrostructural organizati on of sleep, which is impossible to deduce from the unsegmented EEG data. A nalysis of correlations between classically defined sleep stages and piecew ise stationary power step functions reveals that three basic patterns can b e distinguished: SWS (stage III/stage IV). stage II and stage I/REM. In acc ordance with correlation analyses, cluster detection shows that the cyclic sleep patterns during the course of the night become clearly observable by implementation of only three classes. Since the described methodology is ba sed on a minimum of a priori assumptions. it may be useful fur the developm ent of a new sleep classification standard, which goes beyond the establish ed Rechtschaffen and Kales scheme. (C) 2001 Elsevier Science B.V. All right s reserved.