Athena Seminar Series: Engineering Observability for Continuous Health Intelligence

Presenter: Sijie Ji

Human health changes continuously through movement, speech, sleep, behavior, and environment, yet most medical AI still relies on episodic clinical measurements and retrospective records, missing how disease risk and functional decline emerge in everyday life. In this talk, I will show how this agenda spans a continuum from what people do, to what happens inside the body, to how they feel, and ultimately to how intelligent systems can support them. Through wireless sensing, mobile and wearable computing, biochemical sensing, and edge AI, my work develops low-burden systems that recover latent behavioral, physiological, and affective states from weak real-world signals, including WiFi-based fall detection, fine-grained hand motion reconstruction, continuous lipid profiling, pain biomarkers, and proactive mental health support. The broader goal is to build cyber-physical intelligence that helps shift healthcare from late clinical reaction to everyday prevention and intervention.