The winner of this year’s Microsoft Imagine Cup is Bryan Chiang, who has developed an app which uses a simple picture of your eye to accurately estimate your blood sugar levels, valuable for diabetics who are tracking their diet and insulin needs.

Called EasyGlucose, the app provides a non-invasive method of blood glucose monitoring for diabetic patients using high-resolution images of their eye taken with your smartphone camera augmented with a low-cost $10 anterior segment iris imaging adapter.

The app then uses a deep learning computer vision framework developed with Azure Virtual Machines to analyzes iris morphological variation in the eye image to predict the patient’s blood glucose level.

Despite being extremely non-invasive and even non-touch, the results are very clinically accurate and outperform other non-invasive techniques by 30%.

In day to day use the app works off-line, but the app would, of course, record results on the cloud so users can analyse trends to help them optimise their treatment.

Bryan, who is an undergraduate at UCLA studying Computer Science and won $100,000 in the competition, is validating the technology with Stanford University before applying for FDA approval. When it does come to market he hopes to charge a $20/ month subscription fee for what could be a world-changing application for diabetics.