Diabetic Retinopathy Detection using Machine Learning

Diabetic retinopathy (DR) is a retinal disorder characterized by damage to the blood vessels in the eye, often leading to visual impairment or even blindness if not diagnosed and treated early. Automated detection methods are crucial for early intervention, as manual diagnosis is time-consuming and prone to human error due to the complexity of retinal structures. Various automated techniques have been introduced to detect DR from retinal images; however, many of these approaches struggle to identify subtle features, particularly in the early stages of the disease. In this study, we introduce an advanced detection system based on deep learning. Our method utilizes a convolutional neural network (CNN) architecture that integrates two powerful pre-trained models—ResNet50 and InceptionV3. These models are used to extract features from retinal fundus images, and their outputs are combined and passed through the CNN for final classification. To evaluate the effectiveness of the proposed approach, we conducted experiments using a publicly available dataset of fundus images. The performance of our model was measured using several key metrics, including accuracy, sensitivity, specificity, precision, and F1 score. The experimental results demonstrated that our framework outperformed existing methods, achieving an impressive average accuracy of 98.65%.