Skip to main content
. 2026 Feb 23;13:1729151. doi: 10.3389/fnut.2026.1729151

Figure 3.

Panel A and B show calibration plots of predicted versus actual probability of PHT_ANC for training and validation cohorts, with apparent, bias-corrected, and ideal lines. Panel C and D display decision curve analyses for the training and validation cohorts, comparing standardized net benefit across high-risk thresholds for strategies of all, none, and the nomogram.

Calibration curves of the PHT_ANC risk nomogram in the training cohort (A) and validation cohort (B). DCA curves of the PHT_ANC risk nomogram in the training cohort (C) and validation cohort (D).