Computer-Aided Detection Of Diabetic Retinopathy On Fundus Images Via Statistical Classification

Abstract

Diabetic Retinopathy is commonly known diabetic eye disease which leads to blindness. This causes damages to the blood vessels in the retina and can be a source of losing the vision. Nearly 4.1 million people have the same form of the diabetic retinopathy, one in four of these people suffer from loss of vision. Computer Aided Diagnosis (CAD) systems were developed to detect DR in its early stage using retinal fundus images. The CAD system is developed to obtain the quantitative data for DR stages diagnosis. The main motivation of this work is to detection and classification of Diabetic Retinopathy (DR) from Retinal fundus images using Computed Aided Detection (CAD) model. In this context, the use of automatic image processing techniques such as machine learning, neural networks and deep learning algorithms are proposed and the parameter are compared. Finally the deep learning approach produced better result compared to other techniques. The method could help ophthalmologists identify diabetic retinopathy symptoms early on, allowing for a more effective treatment strategy and improved vision-related quality of life.