Automated Diagnosis of Diabetic Retinopathy using image processing for non-invasive biomedical application

In today’s world, Diabetes is a very common disease which affects a lot of people’s health. As the number of diabetic patients is increasing significantly in India, there is a rise in numerous associated diseases that have disturbed the society. Diabetic Retinopathy (DR) is considered to be one of such silent diseases which occur as a result of either Type 1 or Type 2 diabetes. Late diagnosis of this disease may lead to permanent eye blindness. Thus, for early diagnosis of Diabetic Retinopathy, a software-based algorithm is designed here. This technique can be promising for the pre-detection of DR without any involvement of an expert doctor and hence will save both time and money. Here, MATLAB based image processing is used which exploits the knowledge of Computer Science and Biomedical Engineering to identify whitish lesions, cotton wool spots and hard exudates associated with DR. Based on the value of pixel counts, the image of the patient’s eye under examination is classified as a Diabetic Retinopathic eye or a Non-Diabetic Retinopathic eye.