Multi-organ Segmentation on BTCV (80/20 split)
91.4Splvox2vec
Evaluation results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| vox2vecPre-training Method=vox2vec [11], Network=3D UNet(FPN) [30]2024.10 | 91.4 | 90.7 | 59.5 | 72.7 | 96.3 | 83.2 | 91.3 | 83.9 | 69.2 | 73.9 | 65.2 | 79.5 | |
| SparKPre-training Method=SparK [36], Network=MedNeXt [31]2024.10 | 90.92 | 87.66 | 62.43 | 74.36 | 95.03 | 84.85 | 86.04 | 80.63 | 68.83 | 76.57 | 61.43 | 79.21 | |
| MAEPre-training Method=MAE [16], Network=UNETR [15]2024.10 | 90.71 | 87.63 | 62.5 | 72.6 | 96.09 | 94.73 | 86.11 | 90.36 | 71 | 75.47 | 63.77 | 79.07 | |
| HySparKPre-training Method=HySparK [33], Network=MedNeXt+ViT [33]2024.10 | 90.67 | 88.32 | 68.18 | 74.2 | 95.03 | 87.46 | 90.17 | 84.5 | 70.04 | 78.36 | 66.75 | 80.67 | |
| SimMIMPre-training Method=SimMIM [36], Network=Swin UNETR [14]2024.10 | 88.33 | 86.82 | 62.43 | 74.36 | 92.35 | 90.7 | 83.03 | 87.43 | 68.04 | 68.43 | 58.65 | 76.44 | |
| SUPPre-training Method=SUP [14], Network=Swin UNETR [14]2024.10 | 84.2 | 86.7 | 58.4 | 70.4 | 94.4 | 76 | 87.7 | 82.1 | 67 | 69.8 | 61 | 75.8 |