System for multimodal fusion of physiological signals for early detection of cardiovascular instability using wearable biosensor networks

Abstract

A system for multimodal physiological signal fusion for the early detection of cardiovascular instability using wearable biosensor networks, wherein the system comprises the following: a portable device structure with a flexible substrate configured to be positioned on a person's body; a plurality of biosensor units distributed over the flexible substrate, each biosensor unit comprising at least one sensor element selected from an electrocardiography sensor element, a photoplethysmography sensor element, a temperature sensor element, and an inertial sensor element; a signal processing circuit that is operationally connected to each sensor element and is configured to perform the amplification, filtering and analog-to-digital conversion of the detected physiological signals; a preprocessing processor integrated into each biosensor unit, configured to perform noise reduction, motion artifact reduction, baseline correction, and segmentation of physiological signals to generate structured feature data; a wireless communication unit configured to transmit the structured feature data from each biosensor unit; a central processing unit consisting of at least one hardware processor and memory, wherein the central processing unit is configured to receive the structured feature data and perform the temporal alignment and synchronization of multimodal data streams; a fusion processor configured to combine the structured feature data into unified representations; a prediction processor configured to analyze the unified representations to identify early signs of cardiovascular instability; an energy management unit configured to regulate the system's energy consumption; and a user interface configured to generate alerts and display information on cardiovascular risk.

Images