Title: Texture-based medical image classification of computed tomography images using MRCSF

Authors: R.S. Sabeenian, V. Palanisamy

Addresses: Department of Electronics and Communication Engineering, Sona College of Technology, Sona Nagar, Salem, Tamil Nadu 636 005, India. ' Info Institute of Engineering, Sathy Road, Kovil Palayam, Coimbatore, Tamil Nadu 641 107, India

Abstract: Texture classification is a trendy and a catchy technology in the field of texture analysis. It finds its applications in various fields like medical image classification, computer vision, remote sensing, machine vision, agricultural field and many more. Our main focus is on medical images which is a kind of natural texture. The basic idea of the paper is to find out the abnormalities of a patient|s abdomen and more particularly in organs such as kidney, liver and spleen. It is found out by analysis and classifying the CT scan images of a patient|s abdomen. The CT scan images are collected from various sources such as internet, hospitals, etc. The classification is done by extracting the statistical and spectral characteristics of images and by fixing the threshold. The classification results show whether the part of the patient is normal or abnormal. It is a new technology for automation of medical image classification.

Keywords: texture classification; computed tomography; wavelet transforms; wavelets; Markov random field; MRF; grey level cooccurrence matrix; GLCM; multiresolution Markov random field; MRMRF; multi resolution combined statistical and spatial frequency; MRCSF; image classification; medical images; abdomen abnormalities; kidney; liver; spleen; CT scan images; medical engineering; informatics.

DOI: 10.1504/IJMEI.2009.026813

International Journal of Medical Engineering and Informatics, 2009 Vol.1 No.4, pp.459 - 472

Published online: 26 Jun 2009 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article