A-phase Occurrence During Sleep in the Deep Brain Recordings: Multiscale-Entropy and Multiscale-DFA Analysis.

This study evaluates the temporal occurrence of A-phases (TOAP) across superficial and deep electroencephalogram (EEG) recordings to investigate their self-similarity and regulation dynamics across different brain depths during sleep. Sleep recordings from 10 epileptic patients were analyzed to characterize the TOAP in scalp, neocortex (NC), and hippocampus (HPC) signal recordings. TOAP was represented as a binary series, with '1' indicating the presence and '0' the absence of an A-phase. Detrended fluctuation analysis (DFA) was applied to evaluate scale-free properties, and entropy metrics were used to quantify the occurrence probability of symbolic patterns in the binary signal. Mono and multiscale approaches captured TOAP dynamics across different time scales. DFA results revealed consistent correlation properties in TOAP across scalp and deep brain recordings. Monoscale analysis showed persistent long-range correlations, and multiscale analysis indicated a decrease in the scaling exponent from approximately 1.5 toward 0.5 at short scales, and the opposite trend at longer scales. Entropy analysis showed that the TOAP pattern distribution varies with brain region and scale, with higher diversity at the scalp and lower in the NC and HPC. Shannon entropy was positively correlated with cyclic alternating pattern rate. TOAP exhibits consistent scaling across brain regions, suggesting that A-phases follow a unified temporal structure throughout the brain, potentially reflecting a global mechanism of EEG modulation during sleep. In addition, entropy variations suggest that anatomical location and the analysis scale modulate A-phase occurrence. There is a potential relationship between symbolic entropy of TOAP patterns and sleep instability.