Image Processing for Identifying Different Stages of Diabetic Retinopathy

Diabetic Retinopathy (DR) is the damage caused to the retina of the eye due the complications of diabetes. There are a large number of people, who are suffering from Diabetic Retinopathy that leads to blurring of vision or even blindness, if not treated at an early stage. Hence it is important to detect DR at an earlier stage and provide treatment otherwise it may lead to blurring of vision or total blindness. In medical practice so far, the stages of DR are detected by merely looking at the fundus images. This paper presents a more mathematical and statistical approach to identify the different stages of Diabetic Retinopathy. In this paper development of an algrithm for the identification of different stages of DR is discussed.This is carried out by extracting blood vessels from a fundus image and then calculating the perimeter and area of the blood vessels. Retinal blood vessel segmentation is employed by edge detection techniques after pre-processing the images. This is followed by perimeter and area analysis that helps in identifying the severity of Diabetic Retinopathy. The proposed algorithm is simulated on LabVIEW platform. The algorithm is tested on the images of MESSIDOR Data Set and has yielded an accuracy of 85 %..