GLAUCOMA DETECTION USING DEEP LEARNING

This paper presents a web-based intelligent system for automated Glaucoma Detection using image processingand deep learning techniques. The system is designed to assist ophthalmologists and healthcare professionals byproviding a rapid and accurate analysis of ocular images to detect signs of glaucoma, a leading cause of irreversibleblindness worldwide. The model is trained on a publicly available retinal fundus image dataset, such as the RIMONE or DRISHTI-GS datasets, which include labelled samples of healthy and glaucomatous eyes. Deep learningalgorithms, particularly Convolutional Neural Networks (CNNs), are employed to extract features and classifythe images with high precision. The backend is built using Django, and the front end provides a user-friendlyinterface where users can upload retinal images for prediction. The trained model achieves an accuracy ofapproximately 95%, with high sensitivity and specificity, making it a viable tool for clinical support. By enablingearly detection of glaucoma, the system has the potential to improve patient outcomes and reduce the burden onhealthcare systems, particularly in remote or resource-limited regions.Keywords: Glaucoma Detection, Retinal Fundus Images, Deep Learning, Convolutional Neural Networks(CNN), Django Web Application, Computer-Aided Diagnosis (CAD), Artificial Intelligence in Healthcare