Shear-wave elasticity imaging of a liver fibrosis mouse model using high-frequency ultrasound.

The objective of this study was to develop a high-frequency imaging platform for evaluating liver fibrosis in mice based on shear-wave elasticity imaging (SWEI). Although SWEI has been used to diagnose hepatic fibrosis clinically, it is performed at relatively low frequencies (<20 MHz). For preclinical ultrasound imaging in small animals, a high-frequency (>30 MHz) single-element transducer with mechanical scanning is often used. In this study we developed a new SWEI system based on a 40-MHz single-element transducer for imaging and a separate 20-MHz excitation transducer for producing the radiation force and the associated shear waves. Liver fibrosis was induced in ten C57BL/6 (B6) mice using carbon tetrachloride; the other ten mice served as the control group. Synchronizing the excitation beam (i.e., the beam from the excitation transducer) and the detection beam sequence (i.e., the beam from the imaging transducer) allows this mechanical-scanning setup to analyze the shear-wave dispersion relation. The liver viscoelastic properties were determined in vivo by measuring the shear-wave dispersion curve followed by fitting to the Voigt model. The mice were then killed and the fibrosis stage was evaluated (from F0 to F4) based on the METAVIR score. The measured mean values of liver elasticity and viscosity, respectively, ranged from 1.06 to 1.89 kPa and from 1.29 to 1.75 Pa∙s for normal F0 and fibrosis stages of F3 and F4. The Spearman coefficients for the correlations between the measured elasticity and viscosity at various fibrosis stages as assessed by the METAVIR score were 0.73 (p < 0.001) and 0.634 (p = 0.0013), respectively. We also found that the collagen content in the liver was linearly correlated with the measured elasticity (r(2) = 0.54, p < 0.001) and less strongly with the viscosity (r2 = 0.26, p = 0.022). Finally, the diagnosis performance of high-frequency SWEI was evaluated using multivariate receiver operating characteristic curve (ROC) analysis. The areas under the multivariate ROC curve for diagnosing fibrosis stages of F ≥ 3, F = 4, F0 vs. F3, F0 vs. F4, and F3 vs. F4 were 0.9, 0.98, 0.83, 1.0, and 0.96, respectively. Compared with traditional ROC analysis, an improved diagnosis performance was found for diagnosing fibrosis stages of F ≥ 3 and F0 vs. F3. These results demonstrate that the developed high-frequency SWEI platform can yield quantitative viscoelastic properties for diagnosing various fibrosis stages in mice. It is a promising tool for studying the progression of liver fibrosis in preclinical animal models both noninvasively and quantitatively.