Multimodal wearable biosensing meets multidomain AI: A pathway to decentralized healthcare
Published in Advanced Science, 2026
This review presents a unified framework integrating multimodal wearable biosensing with multidomain AI to enable decentralized healthcare. Wearable systems provide continuous monitoring of physiological and biochemical signals, while AI enables fusion of heterogeneous data sources, including biosignals, electronic health records, and medical knowledge.
The paper highlights the evolution from single-modality sensing to multimodal platforms combining biochemical, electrical, mechanical, and imaging signals. It further outlines a three-layer AI pipeline: patient profiling through multimodal fusion, clinical assessment using population-scale data, and therapeutic reasoning supported by knowledge graphs.
Together, these advances enable more accurate diagnosis, early risk prediction, and personalized decision-making. The work also discusses key challenges in data integration, generalization, and privacy, and outlines future directions toward closed-loop, AI-driven healthcare systems.
Recommended citation: Liu, Chenshu, et al. "Multimodal wearable biosensing meets multidomain AI: a pathway to decentralized healthcare." Advanced Science (2026): e22900.
Download Paper