Non-invasive blood glucose monitoring using breath volatile organic compounds

Sudhir Shresta
Dept. of Engineering Science, Sonoma State University

Stevenson Hall 1300
11:00 AM - 11:50 AM

We are developing a smart breath glycemia reader (BGR), a breathalyzer-type hand-held device, that can predict blood-glucose levels from human breath. The device has a microcontroller, volatile organic compound (VOC) sensors, a rechargeable battery, and a wireless chip. We are currently collecting data from patients with type-2 diabetes. We aim to use the data to train machine learning (ML) models and implement into the device for real-time glycemia predictions. Diabetes is a major health problem in the United States that affects more than 122 million people. It requires continual management of blood glucose (BG) to avoid acute and long-term complications, yet, about half of patients with type-2 diabetes do not adhere to their BG treatment plan. Our device, when fully developed, will give patients an accessible modality to read the glycemic status without having to prick their fingers and allow them to test as many times as they desire and easily track their BG history. This will help address the nonadherence and improve BG management among patients with type-2 diabetes.