Publication

Abstract : Diabetic Retinopathy (DR) is a solemn eye ailment that can root vision loss if not diagnosed. It is a complication of diabetes and is caused when high blood sugar destroys the retina. While DR cannot be reversed, early detection and treatment can prevent or slow down vision loss. Traditional physical prognosis of DR by ophthalmologists can be inefficient, prone to errors, and expensive. Computer-aided finding systems, especially those using deep learning mechanisms, have produced great promise by enhancing the accuracy of DR recognition. Convolutional Neural Networks (CNNs) are a particular kind of deep learning mechanism that has been highly successful in analyzing medical images. In this study, we achieved a 95% accuracy in detecting DR using a CNN-based approach.