Automated Classification of Glaucoma Stages using Higher Ordercumulant Features

Bhandary, Sulatha V (2013) Automated Classification of Glaucoma Stages using Higher Ordercumulant Features. Biomedical Signal Processing and Control, 30. pp. 1-10.

[img] PDF
Automated classificationDr.Sulatha & Kevein articlePDF.pdf - Published Version
Restricted to Registered users only

Download (2MB) | Request a copy

Abstract

tGlaucoma is a group of disease often causing visual impairment without any prior symptoms. It is usuallycaused due to high intra ocular pressure (IOP) which can result in blindness by damaging the optic nerve.Hence, diagnosing the glaucoma in the early stage can prevent the vision loss. This paper proposes anovel automated glaucoma diagnosis system using higher order spectra (HOS) cumulants extracted fromRadon transform (RT) applied on digital fundus images. In this work, the images are classified into threeclasses: normal, mild glaucoma and moderate/severe glaucoma. The 3rd order HOS cumulant featuresare subjected to linear discriminant analysis (LDA) to reduce the number of features and then theseclinically significant linear discriminant (LD) features are fed to the support vector machine (SVM) andNaïve Bayesian (NB) classifiers for automated diagnosis. This work is validated using 272 fundus imageswith 100 normal, 72 mild glaucoma and 100 moderate/severe glaucoma images using ten-fold crossvalidation method. The proposed system can detect the early glaucoma stage with an average accuracyof 84.72%, and the three classes with an average accuracy of 92.65%, sensitivity of 100% and specificity of92% using NB classifier. This automated system can be used during the mass screening of glaucoma.

Item Type: Article
Uncontrolled Keywords: Fundus image;Glaucoma;Radon transform;Higher order cumulant;Naïve Bayesiana
Subjects: Medicine > KMC Manipal > Ophthalmology
Depositing User: KMC Manipal
Date Deposited: 04 Jul 2015 04:15
Last Modified: 04 Jul 2015 04:15
URI: http://eprints.manipal.edu/id/eprint/143236

Actions (login required)

View Item View Item