2-D compact variational mode decomposition- based automatic classification of glaucoma stages from fundus images

Glaucoma is one of the leading causes of vision loss worldwide. It leads to reduced quality of life for individuals and substantial economic loss for society. This problem can be reduced by the early and reliable diagnosis of glaucoma. The traditional instrument-based methods are nonautomated and laborious. Recently, many computer-based approaches have been proposed for glaucoma detection. However, none of the existing approaches can be efficiently used for the classification of glaucoma stages. In this study, we proposed a novel method to classify the glaucoma stages (healthy, early-stage, and advanced-stage) using a 2-D compact variational mode decomposition (2-D-C-VMD) algorithm. In this work, the preprocessed input images are first decomposed into several variational modes (VMs) employing 2-D-C-VMD. Next, various features, namely, Kapur entropy (KE), Renyi entropy (RE), Shannon entropy (SE), Yager entropy (YE), energy (En), and fractal dimension (FD) features, which are extracted from the first VM. Then, linear discriminant analysis (LDA) has been used for dimensionality reduction. Finally, a trained multiclass least-squares-support vector machine (MC-LS-SVM) classifier has been utilized for classification purpose. The proposed approach has been tested on two different public glaucoma database. Our method achieved the highest classification accuracy of 98.11% with tenfold cross-validation. The experimental results show that the proposed approach performed far better as compared to state-of-the-art approaches.