Classification of mammogram images using shearlet transform and kernel principal component analysis

• Main Authors: Ibrahim, A.M., Baharudin, B.
• Format: Conference or Workshop Item
• Published: Institute of Electrical and Electronics Engineers Inc. 2016
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85010432805&doi=10.1109%2fICCOINS.2016.7783238&partnerID=40&md5=389f7b764431248aa738a8255f73e92a; http://eprints.utp.edu.my/30483/

In this paper, we have automatically classified the breast tumor in mammogram images to benign and malignant classes using shearlet transform. First the region of interest (ROI) of the mammogram image is subjected to shearlet transform and various texture features are extracted from different levels and orientations. The dimensionality of extracted features are reduced by kernel principal component analysis (KPCA) method and ranked based on T-value. Ten ranked features are fed to k-nearest neighbor (KNN) classifier using minimum features. Our results show that shearlet transform coupled with KPCA is superior to shearlet transform.We have reported an accuracy of 89.8 , sensitivity of 92.7 and specificity of 93.8 using KNN classifier for shearlet-KPCA method. © 2016 IEEE.