Rhythms and Background (RnB): The Spectroscopy of Sleep Recordings

Non-rapid eye movement (NREM) sleep is characterized by multiple coupled rhythms involved in memory consolidation, alongside an arrhythmic 1/f scale-free background also hypothesized to contribute to NREM sleep functions. Spectral parametrization methods have been recently introduced to describe the properties of rhythmic and arrhythmic processes in the spectral domain. However, these approaches fall short disentangling rhythms from arrhythmicity regarding the time series. This is crucial as arrhythmicity can impact the morphology of individual oscillations, therefore affecting their frequency, amplitude, waveform, and their cross-frequency coupling estimated transiently. We introduce ’Rhythms & Background’ (RnB), an innovative wavelet-based methodology that compensates for arrhythmicity within the time domain. We validate RnB against spectral methods and demonstrate its enhanced capability to characterize the hallmark features of NREM sleep rhythms. Furthermore, RnB reveal the rhythmic time series, highlighting key NREM sleep rhythms with improved sensitivity to their phase-amplitude coupling.