Table 3.
Performance of the predicative model
Cohort | Accuracy (95% CI) | Sensitivity | Specificity | Pos. pred. valuea | Neg. pred. valueb | |
---|---|---|---|---|---|---|
Radiomics signature | Training | 0.63 (0.58–0.67) | 0.30 | 0.95 | 0.84 | 0.61 |
Testing | 0.74 (0.68–0.80) | 0.42 | 0.90 | 0.61 | 0.80 | |
Clinical model | Training | 0.68 (0.64–0.73) | 0.32 | 0.94 | 0.78 | 0.67 |
Testing | 0.68 (0.61–0.75) | 0.33 | 0.90 | 0.66 | 0.69 | |
Radiomics nomogram | Training | 0.65 (0.60–0.70) | 0.60 | 0.78 | 0.93 | 0.68 |
Testing | 0.76 (0.70–0.82) | 0.78 | 0.61 | 0.89 | 0.79 |
aPositive predictive value
bNegative predictive value