Skip to main content
. 2024 Dec 19;15:302. doi: 10.1186/s13244-024-01852-z

Table 3.

Predictive performance of different variables

Variables AUC (95% CI) p-value Cutoff Sensitivity Specificity Comparison with DKI
f1 0.839 (0.738–0.913) < 0.001 0.233 96.77% 54.35% Z = 2.360, p = 0.018
f2 0.519 (0.402–0.634) 0.784 / / / /
f3 0.733 (0.620–0.828) < 0.001 0.305 77.42% 65.22% Z = 2.825, p = 0.005
Dapp (× 10−3mm2/s) 0.860 (0.762–0.929) < 0.001 1.200 77.42% 78.26% Z = 1.657, p = 0.098
Kapp 0.874 (0.779–0.939) < 0.001 0.616 83.87% 82.61% Z = 1.499, p = 0.134
DWI/ADC (× 10−3mm2/s) 0.771 (0.661–0.859) < 0.001 1.400 93.55% 52.17% Z = 3.453, p < 0.001
RSI 0.842 (0.741–0.915) < 0.001 / 96.77% 56.52% Z = 2.113, p = 0.035
DKI 0.908 (0.820–0.962) < 0.001 / 87.10% 86.96% /

DWI = ADC, RSI = f1 + f3, DKI = Dapp + Kapp

CI confidence interval