SVM-based characterisation of liver cirrhosis by singular value decomposition of GLCM matrix

Abstract

Early diagnosis of liver cirrhosis is essential as cirrhosis is an irreversible disease most often seen as precursor to development of hepatocellular carcinoma. Early diagnosis helps radiologist in better disease management by adequate scheduling of treatment options. In the present work, features derived from GLCM mean matrix, GLCM range matrix and singular value decomposition of GLCM matrix have been used along with SVM classifier for designing an efficient computer-aided diagnostic system to characterise normal and cirrhotic liver. The study has been carried out on 120 regions of interest ROIs extracted from 31 clinically acquired B-mode liver ultrasound images. It is observed that the first four singular values obtained by singular value decomposition of GLCM matrix result in highest accuracy and sensitivity of 98.33% and 100%, respectively. The promising results obtained by the proposed computer-aided diagnostic system indicate its usefulness to assist radiologists in diagnosis of liver cirrhosis.