Sleep disorders affect a substantial proportion of the global population and are associated with cardiovascular, metabolic, and neurodegenerative comorbidities. In clinical practice, gold-standard diagnosis relies on overnight polysomnography (PSG), a multivariate and multimodal recording of physiological signals, typically reduced to discrete 30-seconds sleep stages, thereby discarding the continuous oscillatory dynamics underlying brain sleep microstructure. In particular, phase-amplitude coupling (PAC) between slow oscillations (SOs, i.e., characteristic of deep non-REM sleep) and sleep spindles (i.e., transient faster thalamocortical bursts) may represent a stable, individual electroencephalographic (EEG) fingerprint, which is systematically distorted in pathological conditions and may provide biologically grounded basis for automated classification of sleep-wake disorders.