Table 5 Performance of clinical, radiomic models and radiomics-clinical nomogram
AUC*(95%CI) | Sensitivity(95%CI) | Specificity(95%CI) | Accuracy(95%CI) | NPV**(95%CI) | PPV***(95%CI) | |
|---|---|---|---|---|---|---|
Train | ||||||
Radiomic | 0.845 (0.781–0.908) | 0.705 (0.585–0.808) | 0.826 (0.739–0.913) | 0.766 (0.686–0.834) | 0.740 (0.642–0.838) | 0.800 (0.698–0.901) |
Clinical model | 0.811 (0.742–0.880) | 0.926 (0.867–0.985) | 0.579 (0.464–0.695) | 0.752 (0.671–0.822) | 0.888 (0.797–0.890) | 0.685 (0.590–0.779) |
Radiomics-clinical nomogram | 0.896 (0.846–0.947) | 0.867 (0.794–0.941) | 0.797 (0.696–0.884) | 0.832 (0.759–0.890) | 0.859 (0.774–0.944) | 0.808 (0.717–0.898) |
Test | ||||||
Radiomic | 0.847 (0.739–0.954) | 0.706 (0.588–0.809) | 0.766 (0.6–0.9) | 0.779 (0.653–0.877) | 0.793 (0.645–0.940) | 0.767 (0.615–0.918) |
Clinical model | 0.808 (0.703–0.913) | 0.862 (0.724–0.965) | 0.567 (0.367–0.733) | 0.712 (0.579–0.822) | 0.809 (0.641–0.977) | 0.657 (0.507–0.809) |
Radiomics-clinical nomogram | 0.887 (0.803–0.971) | 0.793 (0.655–0.931) | 0.733 (0.566–0.867) | 0.763 (0.634–0.864) | 0.786 (0.634–0.938) | 0.742 (0.587–0.895) |