A washable, textile-based sensor array combined with deep learning can identify six distinct sleep states, including sub-healthy and high-risk patterns, with up to 98.6% accuracy, according to a new study.
Sleep disorders affect health, productivity and quality of life worldwide, yet conventional methods like polysomnography are invasive and impractical for daily monitoring.
Current wearable devices often struggle with motion artifacts or require multiple sensors, limiting their capacity for long-term, comfortable use.
This study aimed to address this unmet need by developing a user-friendly ‘diagnostic e-textile’ that detects subtle extrinsic laryngeal muscle vibrations, while ignoring regular sleep movements such as tossing and turning, to classify various sleep disorders while remaining comfortable, robust and scalable for home use.
These sleep disorders range from normal nasal breathing, mouth breathing, snoring, bruxism, central sleep apnoea to obstructive sleep apnoea.
Detecting sleep disorders
The research team integrated a strain sensor array onto the collar of a standard garment using specialised printing processes and chemical treatments. The sensors captured minuscule vibrations without requiring tight skin contact or precise positioning, before feeding these signals into a deep learning neural network called SleepNet.
The SleepNet analysis model was ‘trained’ on five participants and then applied to a dataset from two new participants. Simulated conditions for bruxism and sleep apnoea were established under clinical guidance.
Of the tested conditions, the garment-algorithm combination achieved 98.6% overall accuracy. The sensor array performed reliably against motion artifacts, without the need for precise positioning. SleepNet classified sleep abnormality episodes in real time and performed robustly, the authors concluded.
Reliable monitoring for preventative care
By providing a comfortable, single-modality garment with high fidelity in classifying multiple sleep disorders, this system could improve early detection and management of sleep issues. The authors consider their innovation as ‘ultimately improving understanding and management of sleep disorders’.
Commenting on the results, Professor Luigi Occhipinti, principal investigator of the study, said: ‘Sleep is so important to health, and reliable sleep monitoring can be key in preventative care. Since this garment can be used at home, rather than in a hospital or clinic, it can alert users to changes in their sleep that they can then discuss with their doctor.’
Small sample sizes and partly simulated data for complex apnoea states underscore the need for broader, real-world evaluations, the authors said. Ongoing work should extend testing to larger, more diverse populations, they added.
The upcoming Clinical Excellence in Respiratory Care event includes a session from Sophie West, consultant respiratory and sleep physician at Newcastle upon Tyne NHS Foundation Trust, who will discuss the future of sleep medicine with a focus on diagnostic wearables and artificial intelligence.
Register now to attend on 7 May 2025, or watch on-demand shortly after the event via our Clinical Excellence Catch-up zone.
Reference
Tang C et al. A deep learning–enabled smart garment for accurate and versatile monitoring of sleep conditions in daily life. Proceedings of the National Academy of Sciences, 122(7): e2420498122 (2025).
