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

Nonrapid eye movement (NREM) sleep is characterized by the interaction of multiple oscillations essential for memory consolidation, alongside a dynamic arrhythmic 1/<i>f</i> scale-free background that may also contribute to its functions. Recent spectral parametrization methods, such as fitting oscillation and one-and-over-F and irregular resampling auto-spectral analysis, enable the dissociation of rhythmic and arrhythmic components in the spectral domain; however, they do not resolve these processes in the time domain. Instantaneous measures of frequency, amplitude, and phase-amplitude coupling (PAC) are thus still confounded by fluctuations in arrhythmic activity. This limitation represents a pitfall for studies of NREM sleep relying on instantaneous estimates to investigate oscillatory coupling. To address this limitation, we introduce "Rhythms and Background" (RnB), a novel wavelet-based methodology designed to dynamically denoise time series data of arrhythmic interference. This enables the extraction of purely rhythmic time series, suitable for enhanced time-domain analyses of sleep rhythms. We first validate RnB through simulations, demonstrating it accurately estimates the spectral profiles of individual and multiple oscillations across a range of arrhythmic conditions. We then apply RnB to publicly available intracranial electroencephalogram sleep recordings, showing that it provides an improved spectral and time-domain representation of hallmark NREM rhythms. Finally, we demonstrate that RnB significantly enhances the assessment of PAC between cardinal NREM oscillations, outperforming traditional methods that conflate rhythmic and arrhythmic components. This methodological advance offers a substantial improvement in the analysis of sleep oscillations, providing greater precision in the study of rhythmic activity critical to NREM sleep functions.