Non-invasive blood glucose monitoring device and system

Abstract

A system and method for non-invasive blood glucose level determination and proactive management are disclosed. A sensing device comprises a capacitance-to-digital converter (CDC) sensor, featuring an active shield electrode and single-ended mode operation, to measure capacitance variations indicative of net body charge modulation by systemic physiological activity correlated with blood glucose. The device also includes at least one skin conductance sensor. A processor determines blood glucose levels from combined CDC and skin conductance data, optionally using machine learning. Further, a proactive artificial intelligence (AI) system, employing supervised and reinforcement learning models, analyzes real-time and historical data to predict blood glucose trajectories and the likelihood of future dysregulation. Crucially, the AI dynamically generates a personalized schedule of recommended actions (e.g., relating to diet, exercise, or medication timing) for managing the subject's blood glucose, providing alerts and the action plan to enhance diabetes self-management.

Images