Skip to main content
Fig. 1 | European Radiology Experimental

Fig. 1

From: Multiphase CT-based prediction of Child-Pugh classification: a machine learning approach

Fig. 1

Multiphase computed tomography of three patients. ac Transversal reconstruction of three patients in portal-venous phase. df Pre-processing with semiautomatic liver and vein delineation. a, d Patient with Child-Pugh class A: no changes in liver size or liver parenchyma were observed; all models rated the liver as Child-Pugh class A. b, e Patient with Child-Pugh class B: slight changes in liver configuration as well as heterogeneity of liver parenchyma were observed; only the convolutional network and the expert radiologists’ prediction rated the liver correctly as Child-Pugh class B, whereas the linear regression and the random forest rated it as a Child-Pugh class C. c, f Patient with Child-Pugh class C: overall appearance of the liver exhibits characteristic changes (liver configuration, size, and parenchyma texture); all models rated the liver as Child-Pugh class C.

Back to article page