Quantitative Evaluation of Liver Fibrosis Using Multi-Rayleigh Model with Hypoechoic Component

Abstract

To realize a quantitative diagnosis method of liver fibrosis, we have been developing a modeling method for the probability density function of the echo amplitude. In our previous model, the approximation accuracy is insufficient in regions with hypoechoic tissue such as a nodule or a blood vessel. In this study, we examined a multi-Rayleigh model with three Rayleigh distributions, corresponding to the distribution of the echo amplitude from hypoechoic, normal, and fibrous tissue. We showed quantitatively that the proposed model can model the amplitude distribution of liver fibrosis echo data with hypoechoic tissue adequately using Kullback–Leibler (KL) divergence, which is an index of the difference between two probability distributions. We also found that fibrous indices can be estimated stably using the proposed model even if hypoechoic tissue is included in the region of interest. We conclude that the multi-Rayleigh model with three components can be used to evaluate the progress of liver fibrosis quantitatively.