Real-time physiological stream processing for health monitoring services

We introduce an algorithmic framework that uses nonparametric Bayesian models to process real-time physiological data in the context of developing and testing personalized wellness monitors and tailored intervention strategies. A wearable device aggregates signals from various sensors while periodically transmitting the collected data to a backend server. The server performs the computationally challenging model inference tasks offline and builds custom user profiles based on inferred hidden Markov states. We discuss how these user profiles can be used to detect possible physiological changes in a simple application based on a two-week study hosted at Jaslok Hospital, where relaxation therapy is given to eight healthy male subjects as intervention against stress from the workday. A heuristic is introduced to enable real-time state identification using the modest processing capabilities of the wearable device.