A Novel Framework for Glaucoma Classification and Detection

Abstract

Glaucoma is a progressive optic neuropathy and one of the leading causes of irreversible blindness worldwide. Early detection and accurate classification are crucial for effective treatment and preventing vision loss. This study proposes a robust and automated approach for glaucoma detection and classification using advanced machine learning techniques applied to retinal fundus images. The methodology involves preprocessing, feature extraction, and classification stages to distinguish between normal, early-stage, and advanced glaucoma cases. Various classifiers, including support vector machines, random forests, and deep convolutional neural networks, were evaluated for their performance. Experimental results demonstrate high accuracy, sensitivity, and specificity, confirming the effectiveness of the proposed system. This automated framework can assist clinicians in timely diagnosis and improve patient management, ultimately reducing the risk of glaucoma-related blindness.