Comparison between visual scoring of cyclic alternating pattern (CAP) and computerized assessment of slow EEG oscillations in the transition from light to deep non-REM sleep.

Ten healthy volunteers (six men and four women) aged between 20 and 30 years underwent a nocturnal polygraphic recording on paper and on tape. Spectral analysis was accomplished and EEG sleep scored according to standard criteria and to the guidelines for the identification of cyclic alternating pattern (CAP). The initial 25 min of sleep, starting from the first clear-cut k-complex and ending within stage 4, was subdivided into five consecutive blocks of 5 min each. Using a zero-crossing technique, we evaluated the number of total power oscillations in each block and we tested the hypothesis of significant modifications of the number of total power oscillations and of their periodicity in the successive sessions. In addition, we measured the gap between the maximum and minimum values, respectively, of two successive half-waves making up each oscillation. The hypothesis of time-related trends of the values of the gaps was tested by means of linear regression techniques. Within the 25-min time span, the number of periodic oscillations and the number of CAP cycles showed significant increases. The amplitude gaps underwent a decreasing trend. The present data suggest that slow rhythmic oscillations expressed by CAP can be detected by means of spectral analysis. Their dynamics suggests a close relationship with the EEG synchronization processes.