FIGURE 1.

Development and validation of the artificial intelligence‐derived prognostic index (AIDPI). (A) C‐indices of multiple models derived from various machine‐learning algorithm combinations in three cohorts. (B–D and H–J) Time‐dependent ROC curve analysis of the AIDPI across multiple cohorts. (E–G and K–M) Kaplan–Meier survival analyses of the AIDPI across multiple cohorts. (N) Results of univariate Cox regression analyses of the AIDPI and previously published prognostic signatures. (O) Comparison of the C‐indices of the AIDPI and a signature established by Xu et al., ns: p > .05; *p < .05; ****p < .0001 by Z test.