Diabetic Retinopathy Detection using Android Application

Abstract

Diabetes is a chronic health condition that arises due to inability to maintain a healthy glucose level in the blood. Over a period of time, due to this condition body organ of any individual may get damaged as diabetes affects primary organs like heart, blood vessels, eyes, brain, etc. The main cause of Diabetic Retinopathy is diabetes mellitus, which causes vision problems due to excess swelling of blood vessels of the retina, which further causes leakage of fluids and blood, into retina membrane. Almost 60% to 80% of diabetes patients who are suffering from chronic diabetes suffer from diabetic retinopathy. It is a leading factor for blindness in people from age 21 to 60 years. Diabetic Retinopathy can be treated in the early stage by observing abnormal growth of tissues called lesions which start from Micro-aneurysms in the non-proliferative stage of Diabetic Retinopathy. Many researchers throughout the world have proposed numerous Machine Learning models for early detection of Diabetic Retinopathy from developing into later stages, that is, to prevent blindness. In this paper, android application is developed to detect severity of Diabetic Retinopathy using deep learning techniques.