AUTOMATIC DETECTION AND CLASSIFICATION OF DIABETIC RETINOPATHY USING NEURAL NETWORKS FOR OPTHALMIC HEALTHCARE

Abstract

Diabetic retinopathy (DR) is a serious complication of diabetes that can lead to vision loss and blindness. Early detection and proper identification of the disease's severity are crucial to prevent permanent vision damage and improve patient outcomes. This research introduces an automated system that uses deep learning to improve the accuracy and efficiency of detecting diabetic retinopathy from retinal fundus images. The system uses multiple neural network models, like Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), to analyze retinal images and classify them into five categories: Healthy, Mild, Moderate, Severe, and Proliferative Diabetic Retinopathy. The process includes image preparation, feature extraction, sequential processing, and final classification using a Softmax layer. The trained model is used in a Flask-based web application, allowing users to upload retinal images and get real-time diagnostic results. The system outperforms traditional machine learning methods in accuracy, precision, and reliability. This system can be used for large-scale screening and support healthcare professionals, especially in remote areas.