Studymoderate evidence

BCGNet: an AI model trained on 600 K hours of sleep data for a novel under-pillow contactless monitoring device.

NPJ digital medicine·July 3, 2026·PMID 42399358

Why it matters

Longitudinal home sleep monitoring is limited by the cost and discomfort of PSG. BCGNet offers a scalable, contactless alternative that accurately estimates sleep stages, apnea severity, and sleep continuity, potentially enabling population screening and personalized sleep medicine without burdening the patient.

Infographic summary of BCGNet: an AI model trained on 600 K hours of sleep data for a novel under-pillow contactless monitoring device.
Visual abstract · MedLit

The papers that matter, first.

MedLit ranks the medical literature into a specialty feed with concise summaries, evidence-aware triage, and visual takeaways.

§

MedLit · Evidence, Distilled