Abstract

Glaucoma is an eye disease which may slowly lead to loss of vision. Hence early detection of this disease may help in proper treatment and prevention of vision loss. Retinal Nerve Fiber Layer (RNFL) thining in one of the important indicators of Glaucoma. In this work, a simple RNFL thickness measurement technique from OCT images using color space conversion and morphological processing has been proposed. Seven features like vertical Cup to Disc Ratio (CDR), Horizontal CDR, Average RNFL thickness and average thickness of the four quadrants of RNFL has been used for the detection of Glaucoma. The method has been tested using Back Propagation Artificial Neural Network (ANN), by taking different combination of these features and it has been found that the higher accuracy is achieved when the thicknesses of the RNFL quadrants are considered.