Interpretable Framework for Sleep Monitoring: Applying Statistical Control Charts to Physiological Data Streams.

Polysomnography monitors sleep health with non-linear physiological time-series data, consequently making interpretability a challenge. This study explores the feasibility of applying control charts, a statistical process control method, to cardio-respiratory signals derived from polysomnography studies to provide transparent and interpretable analysis of sleep-related physiological variability. Cardio-respiratory signals from a publicly available polysomnography dataset were preprocessed, transformed, and analyzed using univariate control charts. Sleep stage annotations were used as reference information to contextualize physiological variability across wake and non-REM sleep stages. Phase-level control chart rule violations were examined relative to annotated sleep-state transitions and summarized quantitatively. The results indicate that control chart rule violations occur more frequently during wakefulness and at wake-non-REM sleep transitions, while remaining relatively stable during sustained non-REM sleep. These findings indicate structural correspondence between SPC-based variability flags and annotated sleep stage dynamics. This exploratory, feasibility-focused study does not evaluate diagnostic performance or detection accuracy. Instead, it provides evidence that SPC control charts can serve as a transparent and interpretable analytical framework for exploring physiological variability in sleep data and for supporting future research on sleep-state analysis and explainable data-driven methods.