Table 4.
Comparison of the clinical model, BI-RADS assessment, combined Rad score, and nomogram.
| Training cohort | Validation cohort | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AUC (95% CI) | Acc (95% CI) | Sen | Spe | PPV | NPV | P | AUC (95% CI) | Acc (95% CI) | Sen | Spe | PPV | NPV | P | |
| M1 | 0.782 (0.687–0.877) | 0.734 (0.623–0.827) | 0.609 | 0.909 | 0.903 | 0.625 | 0.690 (0.531–0.849) | 0.659 (0.494–0.799) | 0.583 | 0.765 | 0.778 | 0.565 | ||
| M2 | 0.954 (0.908–1.000) | 0.912 (0.826–0.964) | 0.978 | 0.818 | 0.883 | 0.964 | 0.945 (0.861–1.000) | 0.927 (0.801–0.985) | 0.958 | 0.882 | 0.920 | 0.938 | ||
| M3 | 0.964 (0.931–0.997) | 0.911 (0.826–0.964) | 0.935 | 0.909 | 0.935 | 0.909 | 0.978 (0.941–1.000) | 0.951 (0.835–0.994) | 0.958 | 0.941 | 0.958 | 0.941 | ||
| M4 | 0.975 (0.948–1.000) | 0.924 (0.842–0.972) | 0.913 | 0.849 | 0.894 | 0.875 | 0.983 (0.955–1.000) | 0.951 (0.835–0.994) | 0.958 | 0.941 | 0.958 | 0.941 | ||
| M1 vs. M2 | <0.001 | <0.001 | ||||||||||||
| M2 vs. M3 | 0.741 | 0.404 | ||||||||||||
| M3 vs. M4 | 0.178 | 0.596 | ||||||||||||
| M1 vs. M4 | 0.001 | <0.001 | ||||||||||||
M1, Clinical model; M2, BI-RADS assessment; M3, Combined Rad score; M4, Nomogram model; AUC, area under the ROC curve; CI, confidence interval; Acc, accuracy; Sen, sensitivity; Spe, specificity; PPV, positive predictive value; NPV, negative predictive value.