Table 6. The predictive performance of the nomogram, clinical predictors, and Rad-score.
Cohort | Nomogram | Clinical | Rad-score | P | ||
---|---|---|---|---|---|---|
Nomogram vs. clinical | Nomogram vs. Rad-score | Clinical vs. Rad-score | ||||
Training cohort (n=108) | 0.016 | 0.023 | 0.708 | |||
AUC (95% CI) | 0.830 (0.746, 0. 896) | 0.766 (0.675, 0.842) | 0.744 (0.651, 0.823) | |||
ACC | 0.778 | 0.704 | 0.704 | |||
SEN | 0.768 | 0.652 | 0.638 | |||
SPE | 0.795 | 0.795 | 0.821 | |||
Validation cohort (n=45) | 0.068 | 0.017 | 0.546 | |||
AUC (95% CI) | 0.879 (0.748, 0.957) | 0.781 (0.633, 0.891) | 0.718 (0.564, 0.842) | |||
ACC | 0.822 | 0.644 | 0.689 | |||
SEN | 0.759 | 0.483 | 0.655 | |||
SPE | 0.938 | 0.75 | 0.938 |
ACC, accuracy; AUC, area under the receiver operating characteristic curve; CI, confidence interval; Rad-score, radiomics score; SEN, sensitivity; SPE, specificity.