Extraction of Retinal Features in Fundus Images for Glaucoma Diagnosis

Abstract

Glaucoma is a major cause for blindness, identified early by the structural changes of optic nerve head in the retina. Retinal image analysis is emerging as an important screening tool for prior detection of eye diseases. Glaucoma is a chronic eye disease in which damage to the optic nerve leads to progressive, irreversible vision loss and it is considered as second leading cause for blindness worldwide. In this research work, segmentation of optic disc and optic cup and extraction of retinal features from the fundus images is proposed. The optic cup and optic disc are segmented by active contour, level set, K-means clustering based segmentation techniques. The proposed method has been evaluated for 110 fundus images of DRIONS-DB database. The features are extracted from the processed segmented image and used to identify the severity level of glaucoma. The cup to disc ratio (CDR) is calculated from the segmented fundus image and the severity level of disease is identified if the CDR value exceeds 0.6. Then, the texture based feature extraction is done and these features are used for detection of glaucoma. The features such as CDR (Cup to Disc Ratio) and GLCM (Gray level Co-occurrence Matrix) are used to diagnose the glaucoma.   Keywords: Glaucoma, optic disc, optic cup, K-means clustering, cup to disc ratio (CDR) Cite this Article Keerthanashree T, Ponni Bala M. Extraction of Retinal Features in Fundus Images for Glaucoma Diagnosis. Current Trends in Information Technology. 2016; 6(1):21–28p.