Table 2.
Predictive model performances for the three endpoints shown as mean AUCs (with 95% CIs) from 10×5-fold cross-validation and retrospective-to-prospective cohort transfer. The table includes the number of cases and the target distribution per endpoint.
| n | Conventional, imaging and radiomics parameters: | Conventional and imaging parameters only: | Conventional parameters only: | |
|---|---|---|---|---|
| 10x5-fold cross validation | 10x5-fold cross validation | 10x5-fold cross validation | ||
| Train retrospective – test prospective | Train retrospective – test prospective | Train retrospective – test prospective | ||
| Extracapsular extension | 246 (66 yes, 180 no) | 0.90 (0.86, 0.94) 0.90 |
0.90 (0.86, 0.94) 0.91 |
0.71 (0.63, 0.77) 0.77 |
| Positive surgical margins | 248 (45 yes, 203 no) | 0.64 (0.57, 0.72) 0.63 |
0.60 (0.52, 0.68) 0.67 |
0.50 (0.41, 0.58) 0.49 |
| Nerve-sparing approach decision | 249 (133 yes, 116 no) | 0.80 (0.75, 0.85) 0.81 |
0.79 (0.73, 0.83) 0.79 |
0.81 (0.76, 0.86) 0.82 |