Table 2 Performance features of five-marker stratification model
| Â | Â | Prediction | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | Balanced accuracy (%) | Overall accuracy (%) | Overall AUC | ||
|---|---|---|---|---|---|---|---|---|---|---|---|
Non-BC | LMR-NMIBC | HR-NMIBC + MIBC | |||||||||
Cohort 1 (model development) | |||||||||||
Reference pathology | Non-BC | 63 | 4 | 9 | 87.8 | 82.9 | 88.6 | 81.8 | 85.4 | 78.0 | 0.881 |
LMR-NMIBC | 5 | 12 | 9 | 46.2 | 93.9 | 54.5 | 91.7 | 70.0 | |||
HR-NMIBC + MIBC | 9 | 6 | 74 | 83.1 | 82.4 | 80.4 | 84.8 | 82.7 | |||
Cohort 2 (validation) | |||||||||||
Reference pathology | Non-BC | 34 | 3 | 2 | 91.1 | 87.2 | 91.1 | 87.2 | 89.1 | 82.1 | 0.889 |
LMR-NMIBC | 3 | 6 | 5 | 42.9 | 93.8 | 54.5 | 90.5 | 68.3 | |||
HR-NMIBC + MIBC | 2 | 2 | 38 | 90.5 | 86.8 | 84.4 | 92.0 | 88.6 | |||