Date of Publication :17th May 2024

Abstract:Diabetic retinopathy (DR) is a prevalent complication of diabetes and a leading cause of blindness worldwide. This research presents a novel approach for automated DR detection through a convolutional neural network (CNN)-based model deployed on a web platform. Utilizing a dataset of retinal images, the CNN model with ReLU and Softmax activation functions achieved an training accuracy of 84.84%. The methodology involves preprocessing steps and training the model with 50 epochs, a batch size of 32, and a learning rate of 1e-5. Results demonstrate promising performance metrics, including precision, recall, and F1 score. The study contributes to the advancement of DR screening methods, offering a convenient and accessible tool for early detection and intervention. Further enhancements and validation studies could extend the applicability and reliability of this automated system in clinical settings, potentially mitigating the burden of DR-related vision impairment.

Reference :