Quantitative evaluation method of liver fibrosis using multi-Rayleigh model with three echo envelope components

Quantitative diagnosis method of liver fibrosis using ultrasound echo signal is highly required. The distribution of echo envelope from a homogeneous medium with high scatterer density, such as normal liver, can be approximated by Rayleigh distribution. As the fibrotic liver disease progresses, the distribution begins to deviate from Rayleigh distribution. In our study, we proposed a quantitative method for liver fibrosis staging based on new echo envelope model. We examined a multi-Rayleigh model with three Rayleigh distributions. This model is expressed by the combination of three Rayleigh distributions, corresponding to the echo signal from hypoechoic, normal, and fibrous tissue respectively. Clinical data classified into F0 (normal liver), F1 (early-hepatitis), F2 (moderate hepatitis), and F3 (late-hepatitis) in accordance with the new-Inuyama classification based on the result of biopsy were analyzed by the proposed model. We showed that the multi-Rayleigh model can approximate amplitude distribution of the fibrotic liver accurately using KL divergence, that is an index of the difference between two probability distributions. Additionally, it was found that fibrosis indices of the multi-Rayleigh model obviously increase as the liver fibrosis progresses. These results suggest that the multi-Rayleigh model with three Rayleigh distributions can evaluate the degree of progression of liver fibrosis.