Diabetic Retinopathy Images Classification via Multiple Instance Learning

Diabetic retinopathy (DR) is a complication of diabetes that affects eyes. It is among the primary cause of blindness and low vision and originates from the damage of the blood vessels of the light-sensitive tissue of the retina. The International Diabetes Federation (IDF) [1] estimates that, by 2035, there will be 600 million of person with diabetes, and by 2045 the number will be 700 million. At the present, IDF reports that about 463 million people (1 in 11 adults) worldwide have diabetes and 1.6 million deaths are directly attributed to diabetes each year. Considering the number of patients affected by diabetes worldwide it is straightforward that an affective screening of potential number of patients affected by DR is of paramount importance. While the primary method for evaluating diabetic retinopathy involves direct and indirect ophthalmoscopy, Artificial Intelligent (AI) has been on the rise in the eye care sector. Diabetic Retinopathy can be revealed by analysing fundus photograph data sets of patients and therefore is a disease to which AI tools can provide effective support. In this work we present some preliminary numerical results obtained from classification of eye fundus of healthy people against those of people with severe diabetic retinopathy, by means of Multiple Instance Learning (MIL) algorithm.