Embedded AI Health Monitor

AI/ML
Biosensors
Embedded Systems
Embedded AI Health Monitor

Project Overview

This project focuses on creating a next-generation implantable health monitor. The device combines a flexible, multi-sensor array with a custom-designed, ultra-low-power AI chip. It continuously monitors a panel of biomarkers (e.g., glucose, lactate, inflammatory cytokines) directly from interstitial fluid.


Key Innovations

  • **Ultra-Low-Power AI Chip:** Designed a bespoke ASIC for on-device data processing and anomaly detection, allowing for over 5 years of battery life.
  • **Flexible Sensor Array:** Created a flexible, biocompatible substrate that conforms to tissue, minimizing foreign body response.
  • **Predictive Analytics:** The onboard AI model is trained to not only report current levels but also to predict adverse health events hours before they would be clinically apparent.

  • Outcomes

  • The device consumes less than 1 microwatt during continuous operation.
  • In pre-clinical trials, the system predicted hypoglycemic events in diabetic models with 95% accuracy up to 2 hours in advance.
  • The technology has been licensed to a major medical device company for further development.