Fig. 3

The ROC curves and calibration curves of development set (A) and validation set (B). In the ROC curve, the y-axis is the sensitivity from 0 to 100%, and the x-axis is the specificity from 100% to 0. The y-axis of the calibration curve is the proportion of positive outcomes observed in the corresponding group, and the x-axis is the average prediction probability of the model, the perfect prediction should be on the 45-degree line. The calibration curve of development set is constructed with the restricted cubic spline (RCS) smoother and the calibration curve of validation set is constructed with the “loess” smoother