Enhanced Detection of Diabetic Retinopathy from Fundus Images Using Novel Computing Techniques

Abstract

Diabetic retinopathy is a serious eye condition that can cause blindness and visual loss. Diabetic retinopathy is a complication in blood sugar levels that affects the fundus of the eye. We have devised an approach that is based on the classification of probable diseases such as hemorrhage and exudates, followed by feature extraction from the pixel level result and a machine learning method to predict the severity of diabetic retinopathy. Anatomical structures in retinal pictures for instance blood vessels, exudates, and micro aneurysms are segmented, and images are identified as standard or DR images using characteristics extracted from these structures and the Gray Level Co-occurrence Matrix in our research (GLCM). The Support Vector Machine classifier's issue area is these extracted candidates. The Support Vector Machine classifier sorts the images to identify whether the candidate extraction conclusions are microaneurysms or not. The algorithms have been simulated, and the results have been presented. The classifier employed is the Support Vector Machine (SVM), which has a 96 percent accuracy rate.