Approximately 35% of people living with diabetes have developed a form of diabetic retinopathy. High sugar levels in the blood, associated with diabetes, can lead towards the blocking of the tiny blood vessels in the retina. It is the leading cause of blindness in adults, worldwide. Therefore it is essential to diagnose the disease as early as possible. However, most people experience no discomfort at first, but urgent treatment is needed at this stage. Hence the necessary (annual) examination.

But these examinations by ophthalmologists are time consuming, and early signs might be missed. Therefore, scientists at VITO are designing a computer model that uses artificial intelligence to examine retina scans and detect diabetic retinopathy.

To develop and train the computer model, a large set of annotated images is necessary. There is a database with 10.000 raw images of retina scans, and the project relies on volunteers to annotate these images. Citizen scientists detect, localise an characterise (by size and contours) abnormalities in the scans. Eye for Diabetes uses a Zooniverse application for this task.