Transformer-Based Sleep Stage Classification Enhanced by Clinical Information
This addresses the problem of labor-intensive and variable manual sleep staging for clinicians by enhancing accuracy with contextual cues, though it is incremental as it builds on existing deep learning methods.
The paper tackled automated sleep stage classification from polysomnography by incorporating clinical metadata and expert event annotations into a Transformer-based model, achieving macro-F1 0.8031 and micro-F1 0.9051, which improved over a baseline of macro-F1 0.7745 and micro-F1 0.8774.
Manual sleep staging from polysomnography (PSG) is labor-intensive and prone to inter-scorer variability. While recent deep learning models have advanced automated staging, most rely solely on raw PSG signals and neglect contextual cues used by human experts. We propose a two-stage architecture that combines a Transformer-based per-epoch encoder with a 1D CNN aggregator, and systematically investigates the effect of incorporating explicit context: subject-level clinical metadata (age, sex, BMI) and per-epoch expert event annotations (apneas, desaturations, arousals, periodic breathing). Using the Sleep Heart Health Study (SHHS) cohort (n=8,357), we demonstrate that contextual fusion substantially improves staging accuracy. Compared to a PSG-only baseline (macro-F1 0.7745, micro-F1 0.8774), our final model achieves macro-F1 0.8031 and micro-F1 0.9051, with event annotations contributing the largest gains. Notably, feature fusion outperforms multi-task alternatives that predict the same auxiliary labels. These results highlight that augmenting learned representations with clinically meaningful features enhances both performance and interpretability, without modifying the PSG montage or requiring additional sensors. Our findings support a practical and scalable path toward context-aware, expert-aligned sleep staging systems.