Evaluation of Liver Fibrosis Using Texture Analysis on Combined-Contrast-Enhanced Magnetic Resonance Images at 3.0T

Table 2

Fibrosis prediction model parameters (texture versus Metavir).

	Source image	Texture class	Texture feature
1	Original	Pixel intensity histogram	Mean pixel intensity
2	Original	Gaussian mixture model	STD of the lower intensity pixels
3	Original	Gaussian mixture model	AIC of two-Gaussian fit/AIC of single-Gaussian fit
4	Original	Voronoi polygons	STD of the 1st order inertial moment
5	Gradient	Voronoi polygons	Mean of the 2nd order inertial moment
6	Laplacian	Pixel intensity histogram	Mode/interquartile range

Six most predictive texture features, from strongest to weakest. Keys: STD: standard deviation, AIC: Akaike Information Criterion, inertial moments: mathematical description the shape/area of the Voronoi polygons (see supplementary materials).