Informatics Publishing Limited

T. R. Sivapriya ¹, A. R. Nadira Banu Kamal ², V. Thavavel ³

Affiliations
1 Department of Computer Science, Lady Doak College, Madurai, Tamil Nadu, IN
2 Department of MCA and Department of Computer Science, TBAK College for Women, Kilakarai, Ramnad District, Tamil Nadu, IN
3 Department of MCA, Karunya University, Coimbatore, Tamilnadu, IN

Abstract

Feature selection is of great importance in medical image classification especially neuroimaging classification for determining the most relevant features that will aid in accurate diagnosis of neuropsychological diseases. This paper presents a comparison of feature selection algorithms based on Support Vector Machine (SVM). To achieve robust performance and optimal selection of parameters involved in feature selection, and classification, prior knowledge is embedded to generate multiple versions of training and testing sets for parameter optimization. The integrated feature extraction and selection method is applied to a Structural Magnetic Resonance image based Alzheimer's dementia (AD) study with four different sets of non-demented and demented subjects. Cross-validation results of our study clearly indicate that the algorithm SVM-RFE trained with prior knowledge achieves 98% accuracy with Radial Basis Function (RBF) kernel and can improve performance of the classifier. This novel method of inculcating prior knowledge in SVM-RFE method which is tested in 4 different sets of datasets reveals that RBF kernel is found to outperform other kernels with a mean sensitivity of 97%, and thereby aids in quick and efficient classification of dementia.

Keywords

Support Vector Machine, Classification, Dementia, SVM-RFE.

Full Text

     

T. R. Sivapriya ¹, V. Saravanan ²

Affiliations
1 Lady Doak College, Madurai, IN
2 Dr. N.G.P Institute of Technology, Coimbatore, IN

Abstract

This paper presents a texture based classification method using Support vector machine and decision tree for diagnosis of dementia. Support Vector Machine has been proved to be an effective classifier in several applications. In this work, a comparison of linear and non-linear kernels of SVM with BPN is investigated. Rules are extracted from a trained SVM which is compared with rules extracted from BPN and C5.0. OASIS dataset is utilized for training and testing of the classifiers. Wavelet based textural features from the brain MRI images are given as input feature vectors for classification. From the analysis it is found that SVM outperforms other classifiers. Rules extracted from the trained SVM improve the comprehensibility of the classifier.

Keywords

SVM, BPN, Decision Tree, MRI.

Full Text