Glaucoma is a disease that many people around the world are affected. The disease leads to changes on the patient's retina, which may lead to blindness patients. Early detection of glaucoma could significantly reduce disease effects. People who are susceptible to glaucoma need to be screened regularly. Thus, the need for automatic screening system for detecting glaucoma disease is obvious. Among the goals of this report are the following: Automatic segmentation of retinal fundus images including localization and boundary extraction of optic disc and cup areas, computing cup to disc ratio (CDR), segmenting RNFL layer in the OCT images to obtain RNFL layer thickness and Rim area for feature extraction to screen glaucoma disease based on OCT and Fundus images.

We aim to evaluate the validity of OCT for glaucoma screening using by two factor: 1) Cup to disk ratio by fundus images 2) Find ganglion cell layer thickness in Rim region by using OCT images. We used active contour method for OCT image segmentation. In proposed method, the thickness of RNFL layer was determined by active contour. Then, Rim area diameter was calculated. Retinal image processing included segmentation of the optic disc and cup area to obtain CDR. Proposed system combines features of both OCT and fundus images and is thus able identify glaucoma disease more precisely.

The proposed algorithm was evaluated on 46 OCT and 46 Fundus images. Accuracy of the OCT images segmentation algorithm is equal to (4.18 ± 10) pixels which is equal to (1.69 ± 4.32) um. The data of 23 patients eyes (consist of OCT and Fundus images for left and right eyes for each patient) were analyzed. Combining the best performing optic nerve head parameters (CDR) and nerve fiber layer parameters (The average of RNFL diameter in Rim area) resulted in a sensitivity of 93% , specificity of 87%.

When adequate quality scans may be obtained, the Stratus has moderate sensitivity and high specificity for definitive glaucoma. Specificity is increased when parameters from both the optic nerve head and retinal nerve fiber layer scans are combined.