Table 4.
Accuracy and predictive value of three models
Accuracy | 95% CI | Sensitivity | Specificity | PPV | NPV | |
---|---|---|---|---|---|---|
Training cohort | ||||||
Clinical model | 0.696 | 0.582–0.794 | 0.730 | 0.629 | 0.791 | 0.548 |
Radiomic model | 0.746 | 0.636–0.837 | 0.942 | 0.370 | 0.742 | 0.769 |
Combined model | 0.810 | 0.706–0.889 | 0.826 | 0.777 | 0.877 | 0.700 |
Validation cohort | ||||||
Clinical model | 0.636 | 0.451–0.795 | 0.590 | 0.727 | 0.812 | 0.470 |
Radiomic model | 0.636 | 0.451–0.795 | 0.727 | 0.454 | 0.727 | 0.454 |
Combined model | 0.696 | 0.512–0.844 | 0.727 | 0.636 | 0.800 | 0.538 |
CI, confifidence interval; NPV, negative-predictive value; PPV, positive-predictive value