Computer-aided glaucoma diagnosis system using Support Vector Machine (SVM)

Abstract

The Glaucoma is one of the highest primary reasons of vision loss and vision damage in adults and youngsters. It is very important to find the disease early and can take precaution from vision loss particularly through regular screening. In this same, many of diagnosis methods are utilized ranging famous methods centered on an expert to famous diagnostic techniques, most of times fully diagnosis computerized. The computerized diagnostic systems are developed based on the early finding and glaucoma clinical signs classifications are highly enhanced the diagnosis of the glaucoma disease. Many of the researchers have developed methods permitting the glaucoma clinical signs automatic classification of disease.  However, most of the techniques does not perform vise enough and not optimized. The problem of data stability in method construction and performance test is not taken into account. In this research paper, a predictive technique on the Support Vector Machine (SVM) is produced to optimize the diagnosis techniques of disease, signs using patient retinal and glaucoma sample images. The methodology of this research was Optic Disc (OD) Segmentation, Optic Disc Ratio (ODR), texture feature extraction and classification for glaucoma finding. One comparative study of the proposed algorithm performance for segmentation has been done with the parameters is Jaccard Index, Coefficient of Disc, Sensitivity, and Specificity. For classification performance has been done with these parameters of proposed algorithm namely recall, precision, accuracy,                f-measure, kappa and execution time.