Fig. 3

The radiomics models and their receiver operating (ROC) curves. The radiomics signature (a) and radiomics-clinical model (c) described in the form of nomograms to estimate the risk of a focal liver lesion to be HCC. Locate each variable on the corresponding axis, draw a line straight upward to the Points axis to determine the number of points, add the points from all the variables to get a total point, and draw a line straight down from the “Total Points” axis to the “Risk of hepatocellular carcinoma” axis to determine the HCC probability. b ROC curves of the radiomics signature in the training (red line) and validation cohorts (blue line). No difference (p = 0.521) (DeLong test) was detected between the area under the curve (AUCs) of the radiomics signature in the training cohort (0.861, 95%CI: 0.789–0.932) and in the validation model (0.810, 95%CI: 0.690–0.931). d ROC curves of the radiomics signature (red line) and radiomics-clinical model (blue line) in the validation cohort. No difference (p = 0.213) (DeLong test) was detected between the AUCs of the radiomics signature (0.810, 95%CI: 0.690–0.931) and the radiomics-clinical model (0.866, 95%CI: 0.782–0.951)