Skip to main content
. 2025 Aug 6;16:1636158. doi: 10.3389/fimmu.2025.1636158

Figure 2.

Two graphs labeled A and B comparing diagnostic test performances. Graph A is a Receiver Operating Characteristic (ROC) curve plotting Sensitivity against 1-Specificity. Three lines represent Early-RA versus Control (AUC=0.82), Established-RA versus Control (AUC=0.67), and axSpa versus Control (AUC=0.55). Graph B is a Precision-Recall (PR) curve plotting Precision against Recall. Three lines represent Early-RA versus Control (PR-AUC=0.81), Established-RA versus Control (PR-AUC=0.75), and axSpa versus Control (PR-AUC=0.48). Different colors distinguish the lines.

Predictive values of adipsin concentrations for each RD compared to controls. (A) ROC curves showing the discriminatory capacity of serum adipsin levels between healthy controls and patients with early-RA, established-RA, and axSpA. AUC values are indicated for each comparison. Optimal cut-offs were defined using the Youden Index. (B) Precision-Recall (PR) curves assessing the performance of classification models in distinguishing disease groups from healthy controls. These curves illustrate the balance between precision and recall across different classification thresholds.