Power-law scaling behavior of A-phase events during sleep: Normal and pathologic conditions

Abstract

ObjectiveA-phases are short-time cortical events during sleep that repeatedly disrupt the basal fluctuations of electrical brain activity. They are the basic units of the cyclic alternating pattern (CAP), which is related to the instability and consolidation of the sleep process. The main purpose of this study is to evaluate the temporal occurrence of A-phases (TOAP) under normal and pathologic sleep conditions. MethodsTo derive a quantitative description of TOAP, we have applied detrended fluctuation analysis (DFA), to unveil scale-free behavior from non-stationary time series. Data from sleep recordings of 15 healthy subjects (H), 37 with nocturnal frontal lobe epilepsy (NFLE), 9 with periodic leg movement (PLM), and 22 with REM Sleep Behavior Disorder (RBD) from the Physionet database were used. TOAP was computed from binary time series constructed from A-phase annotations, where symbols 1 and −1, represent the presence or absence of an A-phase, respectively. These time series were analyzed through DFA and characterized by the scaling exponent. ResultsIn all cases, numerical results show evidence that a statistical fractal structure is embedded in TOAP, where persistent scaling behavior is observed with scaling exponent close to 1. ConclusionThe sleep process maintains a similar structure from the A-phase perspective despite of internal or external factors that could affect the sleep process, suggesting that sleep may be a resilient process. SignificanceThe scaling exponent of the A-phase occurrence provides new information about the sleep process that may have clinical relevance and complementary to standard clinical indices such as the CAP-rate.