Wearable Health Monitoring

The field studies how sensors such as PPG, ECG, accelerometers, and continuous glucose monitors can track health during everyday life, supporting tasks like sleep staging, blood-pressure estimation, activity recognition, and stress monitoring. Common approaches combine signal denoising and feature extraction with temporal deep learning, multimodal fusion, personalization, and increasingly large self-supervised foundation models. A major practical focus is making these systems reliable on low-power devices and robust to motion, missing data, and differences across people, devices, and body sites. Strong benchmark results remain in tension with the need for participant-disjoint and external validation, fair performance across subgroups, and evidence that predictions are clinically useful in real-world care.