Exploring Texture-Based Parameters for Noninvasive Detection of Diffuse Liver Diseases and Liver Cancer from Ultrasound Images Eng. Delia MITREA 1 , Prof. PhD. Eng. Sergiu NEDEVSCHI 1 , Eng. Calin CENAN 1 , Med. Monica LUPSOR 2 , Prof. PhD Med. Radu BADEA 2 1 Computer-Science Department, Technical University of Cluj-Napoca str. Baritiu no. 26-28, ROMANIA, 2 Department of Ultrasonography, 3 rd Medical Clinic, University of Medicine and Pharmacy, Cluj-Napoca, ROMANIA Abstract : Non-invasive, image based detection of diseases is one of the most important issues in the nowadays research of biomedical images, because it prevents from some serious problems, that could be generated by the invasive techniques and could be dangerous for the patients. Texture is a fundamental visual property of the tissue providing a lot of information concerning its pathological state. Thus, we developed specific methods for texture analysis and recognition, for automatic and semi-automatic detection of some liver diseases from ultrasound images, in order to assist the medical personal in establishing a diagnostic in non-invasive way. We also performed some studies concerning the relevance of these parameters in the case of various liver diseases. Key-Words: - Textural parameters, Diffuse liver diseases, Cancer, Data Mining, Automated diagnosis 1. Introduction Liver cancer is considered one of the most frequent causes of mortality in the entire world. The hepato- cellular carcinoma (HCC) is one of the primary malign liver tumors, with hepatic origin, such as the hepatoblastoma, colangiocarcinoma and hepatic sar- coma [1]. One of the non-invasive methods for HCC diagnosis is the ultrasonography. This method can be optimized through computerized methods of echo- graphic image processing. Beside the features analy- zed by the medical specialists, that can be noticed with the eyes, the specific methods for image analysis can provide subtle information, like those referring to the textural features of the hepatic pa- renchyma – first and second order statistics of the grey levels, multi-scale features, the fractal dimen- sion, features referring to the structure and distribution of microstructures, to the bi-dimensional shape of the contour, respectively to the tri-dimensi- onal aspect of the tumor. Other important category is that of diffuse liver diseases, such as steatosis – which imply the accumulation of fat in hepatocytis (liver cells); hepatitis – implying the inflammation and degradation of the hepatic tissue, resulting in fibrosis or necrosis, and, finally, the cirrhosis, implying the presence of the fibrosis and necrosis, but also the nodules formation, usually preceding cancer. Properties like the grey levels mean variance, coarseness and second order statistics like entropy, homogeneity, contrast, correlation, based on the Grey level co-occurrence matrix (GLCM), but also the fractal index can provide the information which is necessary for diffuse liver diseases characterization and distinction. We apply texture-based features in order to describe and to compare the main pathological stages of the liver – steatosis (fat accumulation), fibrosis (liver tissue degradation) [10], [11] and liver tissue reorganization (nodules and tumors formation). We will study all these cases from the point of view of the texture-based features (first and second order statistics, respectively the fractal index) . 2. Visual appearance of liver tissue in ultrasound images in various patholo- gical stages Ultrasound images are formed as the result of ultrasounds propagation from the transducer in the human body and from the human body back to the transducer. They are obtained through echography Proc. of the 8th WSEAS Int. Conf. on Mathematical Methods and Computational Techniques in Electrical Engineering, Bucharest, October 16-17, 2006 259