Glaucoma Detection using Fuzzy C-means Optic Cup Segmentation and Feature Classification

Abstract

Ophthalmological diseases cause damage to various parts of human retina. Glaucoma damages optic disc which may lead to progressive and irreversible vision loss. Early diagnosis and detection helps in prevention of vision loss and improves the quality of life of patients. The proposed method aims to develop a glaucoma detection system using fundus images. The method focuses on optic cup segmentation using fuzzy c-means (FCM) algorithm. Fusion of segmentation based and global image based features is used for fundus images classification into normal and glaucoma classes using support vector machine (SVM) and ensemble classifiers. Optic cup segmentation and glaucoma classification results are evaluated on publicly available Drishti-GSI database using relevant performance metrics and compared with methods in literature.