Probability image of tissue characteristics for liver fibrosis using multi-Rayleigh model with removal of nonspeckle signals

We have been developing a quantitative diagnostic method for liver fibrosis using an ultrasound image. In our previous study, we proposed a multi- Rayleigh model to express a probability density function of the echo amplitude from liver fibrosis and proposed a probability imaging method of tissue characteristics on the basis of the multi-Rayleigh model. In an evaluation using the multi-Rayleigh model, we found that a modeling error of the multi-Rayleigh model was increased by the effect of nonspeckle signals. In this paper, we proposed a method of removing nonspeckle signals using the modeling error of the multi-Rayleigh model and evaluated the probability image of tissue characteristics after removing the nonspeckle signals. By removing nonspeckle signals, the modeling error of the multi-Rayleigh model was decreased. A correct probability image of tissue characteristics was obtained by removing nonspeckle signals. We concluded that the removal of nonspeckle signals is important for evaluating liver fibrosis quantitatively.