Diagnosis of liver tumour from CT images using contourlet transform. (1st January 2011)

• Record Type: Journal Article
• Authors: Kumar, S.S.; Moni, R.S.
• Is Part Of: International journal of biomedical engineering and technology. Volume 7:Number 3(2011)
• Page Start: 276
• Page End: 290
• Publication Date: 2011-01-01
• Subjects: CAD -- contourlet transform -- liver tumour -- segmentation -- feature extraction; Biomedical engineering -- Periodicals; 610.28
• DOI: 10.1504/IJBET.2011.043300
• Languages: English
• ISSNs: 1752-6418

Abstract

In this paper, a novel feature extraction scheme based on multiresolution contourlet transform for the automatic diagnosis of benign and malignant liver tumours from Computed Tomography (CT) images is presented. The liver is segmented automatically from CT images using region growing and the suspected tumour region is extracted from the segmented liver using Fuzzy C Means (FCM) clustering. The textural information obtained from the segmented tumour by contourlet transform is used to detect and classify the liver tumours using feed-forward neural network classifier. The tumours diagnosed by this method are haemangioma (benign) and hepatocellular carcinoma. A comparison with a similar algorithm based on wavelet texture descriptors shows that using contourlet-based texture features significantly improves the classification rate of liver tumour from CT scans.