GLAUCOMA IMAGE CLASSIFICATION USING DISCRETE ORTHOGONAL STOCKWELL TRANSFORM

Abstract

Glaucoma is an eye condition which is caused by the improved blood pressure in the optic nerve. It causes a functional failure of the visual field and irreversible. A Computer Aided Diagnosis (CAD) can help the doctors to find glaucoma at the earliest. In this paper, a CAD system for glaucoma diagnosis using Discrete Orthogonal Stockwell Transform (DOST) is presented. DOST distribute its coefficients based on sample spacing paradigm where low frequencies have a lower sampling rate, and high frequencies have higher sampling rate. All DOST coefficients are considered for the diagnosis of glaucoma using Random Forest (RF) classifier. Results show that the glaucoma diagnosis system has 96% sensitivity, 92% specificity, and 94% accuracy using 100 fundus images of normal and glaucoma cases.

Keywords

• fundus image, glaucoma, discrete orthogonal stockwell transform, random forest classifier