Quantitative Tissue Characterization Program, Computing and Color-Mapping of Ultrasound Parameters for Computerized Diagnosis of Liver Pathologies

In this paper a new quantitative tissue characterization (QTC) program was developed for computing a standardized sets of parameters and quantitative color-mapping of these parameters for non-invasive accurate diagnosis of liver pathologies. In this program, histogram, co-occurrence matrix, runlength matrix, gradient texture feature coding, backscattering, attenuation and speckle parameters were numerically calculated and color mapped in the ultrasound image. The standardization of these parameters using AIUM tissue phantom simulating liver texture, were done for accurate differentiation and estimation of the quantitative parameters. A QTC program was linked to the Neural Net and Fuzzy Logic modules so as to diagnose an unknown liver pathologies. This program is used for quantification and calculation of physical properties of tissues, color mapping of such properties and diagnosis of these tissues. Also it can be extended for real time color mapping of ultrasound parameters with motion.