Paper
Addressing Wearable Sleep Tracking Inequity: A New Dataset and Novel Methods for a Population with Sleep Disorders
Proceedings of the Conference on Health, Inference, and Learning (CHIL 2024), PMLR 248:380-396
TL;DR
Introduces DREAMT — 100 Duke Sleep Disorders Center patients wearing an Empatica E4 during overnight PSG — and a sleep/wake method for a sleep-disordered population. Per-30-s-epoch features from BVP (Oura-style PPG preprocessing), ACC, EDA and skin temperature feed LightGBM; GPBoost mixed-effects models add obesity or apnea severity as random effects, and an LSTM post-processing step adds temporal context. Best (GPBoost + apnea severity + LSTM), participant-level 5-fold CV on 80 participants: wake-positive F1 0.823 ± 0.019, accuracy 0.857, AUROC 0.926, AUPRC 0.915, kappa 0.694 ± 0.025 (baseline LightGBM kappa 0.605).
- Form factor: wrist
Results on ChatPPG leaderboards · 4
All leaderboards| Leaderboard | Method | Result | Rank |
|---|---|---|---|
| DREAMT accuracySleep/wake detection (actigraphy-style) | Addressing Wearable Sleep Tracking Inequity | 0.857accuracy, higher is better | #1 of 1 |
| DREAMT AUROCSleep/wake detection (actigraphy-style) | Addressing Wearable Sleep Tracking Inequity | 0.926AUROC, higher is better | #1 of 1 |
| DREAMT Class F1 for one stated class , not macroSleep/wake detection (actigraphy-style) | Addressing Wearable Sleep Tracking Inequity | 0.823Class F1 for one stated class , not macro, higher is better | #1 of 1 |
| DREAMT Cohen's kappaSleep/wake detection (actigraphy-style) | Addressing Wearable Sleep Tracking Inequity | 0.694Cohen's kappa, higher is better | #1 of 1 |
4 rows are credited to this paper. Ranks apply within each leaderboard only, as reported by the source.
Tasks & signals
Tasks
Signals
Devices
Keywords
Datasets used
How it works
Cite this paper
Wang, W. K., Yang, J., Hershkovich, L., Jeong, H., Chen, B., Singh, K., Roghanizad, A. R., Shandhi, M. M. H., Spector, A. R., & Dunn, J. (2024). Addressing Wearable Sleep Tracking Inequity: A New Dataset and Novel Methods for a Population with Sleep Disorders. Proceedings of the Conference on Health, Inference, and Learning , PMLR, 248, 380–396. https://proceedings.mlr.press/v248/wang24a.html
Wang, Will Ke, Jiamu Yang, Leeor Hershkovich, Hayoung Jeong, Bill Chen, Karnika Singh, Ali R Roghanizad, Md Mobashir Hasan Shandhi, Andrew R Spector, and Jessilyn Dunn. 2024. "Addressing Wearable Sleep Tracking Inequity: A New Dataset and Novel Methods for a Population with Sleep Disorders." Proceedings of the Conference on Health, Inference, and Learning , PMLR 248: 380–396. https://proceedings.mlr.press/v248/wang24a.html.
Wang, W. K., Yang, J., Hershkovich, L., Jeong, H., Chen, B., Singh, K., Roghanizad, A. R., Shandhi, M. M. H., Spector, A. R. and Dunn, J. (2024) 'Addressing Wearable Sleep Tracking Inequity: A New Dataset and Novel Methods for a Population with Sleep Disorders', Proceedings of the Conference on Health, Inference, and Learning , PMLR, 248, pp. 380–396. Available at: https://proceedings.mlr.press/v248/wang24a.html
Wang, Will Ke, et al. "Addressing Wearable Sleep Tracking Inequity: A New Dataset and Novel Methods for a Population with Sleep Disorders." Proceedings of the Conference on Health, Inference, and Learning , PMLR, vol. 248, 2024, pp. 380–396. https://proceedings.mlr.press/v248/wang24a.html.
W. K. Wang et al., "Addressing Wearable Sleep Tracking Inequity: A New Dataset and Novel Methods for a Population with Sleep Disorders," Proceedings of the Conference on Health, Inference, and Learning , PMLR, vol. 248, pp. 380–396, 2024, [Online]. Available: https://proceedings.mlr.press/v248/wang24a.html.
@inproceedings{wang2024addressing,
title = {{Addressing Wearable Sleep Tracking Inequity: A New Dataset and Novel Methods for a Population with Sleep Disorders}},
author = {Wang, Will Ke and Yang, Jiamu and Hershkovich, Leeor and Jeong, Hayoung and Chen, Bill and Singh, Karnika and Roghanizad, Ali R. and Shandhi, Md Mobashir Hasan and Spector, Andrew R. and Dunn, Jessilyn},
year = {2024},
booktitle = {Proceedings of the Conference on Health, Inference, and Learning , PMLR},
volume = {248},
pages = {380--396},
url = {https://proceedings.mlr.press/v248/wang24a.html}
}Added to the ChatPPG research hub
