All-optical sensing of dermal activity enables an AI-powered wearable ring for continuous stress monitoring
Electrodermal activity (EDA), a primary stress-sensing signal in today's wearables, requires skin-electrode contact that drifts and degrades over time, limiting long-term use. Here, we present an artificial intelligence (AI)-powered ring for all-optical sensing of dermal activity (ODA), using light rather than electrodes to track sweat gland activation, thereby enhancing durability and comfort. An optimized dual-wavelength design (960 and 1250 nanometers) captures perspiration dynamics while concurrently recording photoplethysmography. The ODA ring integrates temperature and motion sensors for multimodal monitoring with on-device inference. In correlation studies, ODA and EDA signals showed 86% agreement. In classification tasks, the ODA ring achieved 92.3% binary accuracy versus 77.95% for EDA. In multiclass tasks, ODA outperforms EDA by 100 to 74.58% (short term) and 89.72 to 43.20% (long term). On-device one-dimensional convolutional neural network inference enabled 87.53% autonomous accuracy. This electrode-free modality establishes a scalable, all-optical platform for rings, watches, and patches for accurate, robust, and long-term stress monitoring.
Authors
- Md. Farhad Hassan (ORCID: https://orcid.org/0000-0003-3413-7286)
- Emily Rawson (ORCID: https://orcid.org/0009-0006-7299-4085)
- Yasser Khan (ORCID: https://orcid.org/0000-0003-2290-0854)
- Burçin Becerik-Gerber (ORCID: https://orcid.org/0000-0001-8648-0989)
- A. K. M. Rakib (ORCID: https://orcid.org/0000-0001-5349-4524)
- Adam Charles Frank (ORCID: https://orcid.org/0000-0001-8203-8480)
- Gale Lucas (ORCID: https://orcid.org/0000-0003-3089-9283)
- Samiha Tasnim
- Mohammed Arfan (ORCID: https://orcid.org/0009-0002-1955-2668)
- Mohammad Shafiqul Islam (ORCID: https://orcid.org/0000-0002-0998-3643)
- Leikhang Xiong
- Antonin Brun (ORCID: https://orcid.org/0009-0005-5736-0474)
- Shounak Das (ORCID: https://orcid.org/0009-0001-3104-6744)
- Dolendra Vikas Addepalli
- Xiao Yu
Institutions
- University of Southern California (US)
- University of Maryland, College Park (US)
Publication Details
- Journal
- Science Advances
- Published
- 2026-11-13
- DOI
- https://doi.org/10.1126/sciadv.aeh0459
- Primary Topic
- Non-Invasive Vital Sign Monitoring
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Science Foundation