Fig. 2

Histogram of the intra-class correlation coefficient (ICC). For the 20 random selected patients from the overall dataset, we extracted the radiomics features from the test and re-test scans. The ICC was used to determine the stability of the features. Features with an ICC <0.75 were excluded from the analysis. After robustness test, 420 of the initial 647 CT image features in the arterial phase, 350 in the portal venous phase, and 455 in the parenchymal phase were retained