Proposal of compound amplitude envelope statistical analysis model considering low scatterer concentration

Abstract

Many methods for the analysis of amplitude envelope statistics have been proposed in recent decades to enable echo signal characterization and realize quantitative diagnosis based on these statistics. In the statistical analysis of ultrasound signals, the spatial resolution of the signal is an important factor. An analysis method that offers higher sensitivity than the current analysis model is required to allow the effective use of recently developed high-resolution ultrasonic diagnostic equipment. In this study, we propose a multi-amplitude envelope statistical model called the double Nakagami probability density function (PDF) model that assumes the physical structure of fatty liver disease under high-frequency ultrasound excitation. Application of the proposed model to actual rat livers demonstrated that it was possible to evaluate fatty liver disease at an early stage with a lower scatterer density than that of a normal liver. However, it is difficult to detect the disease at this stage using existing technology.

Ultrasound Assessment of Hepatic Steatosis by Using the Double Nakagami Distribution: A Feasibility Study

“…The details for the initial model and hypothesis can be found in the previous study [26]. In brief, the Nakagami distribution model assumes that only one type of scatterer exists in the scattered media.…”

“…This means that normal and fat-infiltrated hepatocytes (lipid droplets) are two primary scattering sources with different acoustic properties for fatty liver. The DND model combines two Nakagami distributions by using a weighted factor α to describe two scattering sources, as follows [26]:…”

“…Compared with the conventional Nakagami parameter, the DND parameter specifically reflects the statistical distribution of ultrasound signals backscattered from fat-infiltrated hepatocytes in the liver parenchyma [26]. The effects of hepatic steatosis on the Nakagami parameter have been discussed by modeling the liver tissue as a scattering medium [23][24][25]; the same concept could be used to explain how the DND parameter varies with the degree of hepatic steatosis.…”

“…It has been shown that the DND model can fit the envelope statistics relatively accurately in cases of small and large numbers of lipid droplets; this advantage over the single Nakagami distribution is attributed to the increase in the number of degrees of freedom in the data analysis [26]. However, the estimated parameter of the DND model is typically higher than that obtained using the conventional Nakagami distribution [26].…”

“…To resolve this problem, the double Nakagami distribution (DND) has recently been proposed as a new model for describing the backscattered statistics of hepatic steatosis under high-frequency ultrasound excitation. The DND model combines two Nakagami distributions to allow the estimation of the Nakagami parameter corresponding to lipid droplets, and the feasibility of DND-based parametric imaging for detecting the fatty liver was explored through ex vivo animal experiments [26]. While proposing the DND model as a new approach to detect hepatic steatosis using high-frequency ultrasound, the diagnostic performance of DND imaging in clinical evaluations of the hepatic steatosis grade remains unknown.…”