Machine learning powers many new diagnostic technologies, but AI doesn’t have a total lock on medical advances. Engineers at Pohang University of Science and Technology (POSTECH) developed a contact lens wearable that can diagnose diabetes and treat diabetic retinopathy. The POSTECH engineers collaborated with researchers from Stanford University’s Department of Engineering and Stanford Medicine Ophthalmology.

The multifunctional aspect of POSTECH technology relates to its potential for use in both diagnosis and treatment. The technology starts with a smart contact lens. The lens has an integrated micro-LED and a photodetector. The light and the detector assist diabetes diagnosis by measuring glucose in the conjunctival blood vessels using near-infrared (NIR) light.

The POSTECH engineers then treated diabetic retinopathy disease by using the smart lenses to shine light in their eyes. After a month of this exposure to NIR light, the researchers observed a significant reduction in angiogenesis in the retina, which is the production of new blood vessels. This indicates that similar lenses might be used to treat retinopathy in humans.

It is important to note that while the researchers used a smart contact lens to measure glucose levels, they were not relying on levels in the rabbits’ tears as the source for the data. Sampling tears was the approach that was explored and abandoned by Google Verity and Alcon after they determined that the correlation between glucose levels in tears and in blood is too inconsistent to be reliable. Instead, this study used near-infrared light to measure blood glucose levels in the eye’s blood vessels.