Technical University of Denmark (DTU) · Department of Health Technology, Digital Health
How can routinely collected physiological signals — a night of sleep, a week of wrist accelerometry — become robust, clinically meaningful measures of health and disease? This work develops the signal-processing and machine-learning methods to make that possible.
Extended wearable iRBD screening beyond simple actigraphy, combining sleep, motor, and cardiorespiratory signatures from wrist accelerometry for more accurate multicenter detection of isolated REM sleep behavior disorder.
Advanced self-supervised and foundation-model approaches for wrist accelerometry, including large-scale prediction of future disease risk from one week of wrist movement in 97,696 UK Biobank participants.
Co-developed a multimodal sleep foundation model trained at very large scale for broad disease and mortality prediction.
Physiology-aware masked-autoencoder pretraining for wrist accelerometry that learns pulse- and respiration-related motion, enabling accurate sleep staging and sleep-apnea evaluation directly from a wrist-worn device.
Introduced deep-learning age estimation from polysomnography and showed that excess sleep age is associated with mortality, supporting sleep-derived age deviation as an interpretable marker of sleep-related aging.
01/05
539
Citations · h-index 13
46
Publications & preprints
23
Completed student & PhD projects
4
Major research collaborations
Citation metrics via Google Scholar, updated 2026-08-23.
Extract interpretable markers of sleep health, aging, fragmentation, and neurological disease from polysomnography and related physiological signals.
Develop and evaluate self-supervised and foundation-model approaches that learn reusable representations across physiological modalities, cohorts, and downstream health tasks.
Use wrist accelerometry and other low-cost sensors to derive scalable physiological markers for sleep, activity, neurodegenerative disease screening, and future health risk.
No open calls posted right now — strong candidates are still welcome to reach out.
See opportunitiesInterested in a research collaboration, data access, or an applied project?
Get in touch