Fig. 5

Decision curves of training set (A) and test set (B). The X-axis represents the model power, ranging from 0–100%, the Y-axis measures the net benefit, and purple, green, blue, and orange represent the decision curves of different models, respectively. None and ALL are two reference lines; the closer the other model curves are to the two reference lines, the less value is applied; the high ordinate in the case of the same abscissa represents the better model efficacy