A New Approach of Cup to Disk Ratio Based Glaucoma Detection Using Fundus Images

Article type: Research Article

Authors: Khan, Muhammad Waseem^a | Sharif, Muhammad^a | Yasmin, Mussarat^{a; *} | Fernandes, Steven Lawrence^b

Affiliations: [a] Department of Computer Science, COMSATS Institute of Information Technology, Wah Cantt, Pakistan | [b] Department of Electronics and Communication Engineering, Sahyadri College of Engineering & Management, Mangalore, Karnataka, India

Correspondence: [*] Corresponding author. Email: mussaratyasmin@comsats.edu.pk

Abstract: This paper introduces a procedure for glaucoma exposure by computing cup to disk ratio for low contrast digital fundus images. The designated approach is based on pre-processing, processing and post processing steps. Pre-processing consists of enhancing contrast of optic cup using brightness preserving dynamic fuzzy histogram equalization (BPDFHE) and optic disk regions, the processing step involves the segmentation of optic cup and optic disk and post processing will finally measure cup to disk ratio of given fundus image to detect glaucoma in the image. The method was successful in 381 images out of 403, achieving 98.20% sensitivity, 96.70% specificity and 97.67% global accuracy in pre diagnosis.

Keywords: Glaucoma detection, fundus image, cup to disk ratio, optic cup, optic disk, BPDFHE, watershed

DOI: 10.3233/jid-2016-0004

Journal: Journal of Integrated Design and Process Science, vol. 20, no. 1, pp. 77-94, 2016