Binary Classification on Internal Hold-out n=248 (test)
85SensitivityCluster-Induced Mask Transformer
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Cluster-Induced Mask Transformer2023.07 | 85 | 0.939 | 92.6 | |
| TransUNet-S4Capproach=Segmentation for classification (S4C)2023.07 | 82 | 0.916 | 90.5 | |
| nnUNet-Jointapproach=Joint CNN classification head2023.07 | 81 | 0.924 | 90.5 | |
| nnUNet-S4Capproach=Segmentation for classification (S4C)2023.07 | 80 | 0.907 | 88.5 | |
| Mean of radiologists2023.07 | 73.5 | — | 84.1 |