Ultrasound velocity in fixed human liver: empirical anova and regression modelling on histologically assessed...

Anova and "dummy variables" least squares regression methods are used to investigate the dependence of ultrasound propagation velocity on the presence or absence of fat and on the rank ordered level of cirrhosis in fixed human liver. Cirrhosis is found to affect velocity in a nonlinear fashion, while fat depresses the velocities toward lower values. Quantifications of these effects and statistical analyses of their magnitudes are presented. The results represent a contribution toward the empirical understanding of the effects of tissue abnormalities on the velocity of ultrasound propagation in fixed human tissues.