Liver Segmentation on LiTS
96.52Dice ScoreVoCo
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| VoCoNetwork=Swin-UNETR, Pre-training=Medical SSL2024.02 | 96.52 | — | — | — | |
| VoCoNetwork=3D UNet, Pre-training=Medical SSL2024.02 | 96.03 | — | — | — | |
| MNet + (SG + VM)#Params (M)=7.42, Epochs=150, CV Folds=3 (80/20)2026.06 | 95.9 | — | — | — | |
| MNet + VMamba#Params (M)=7.42, Epochs=150, CV Folds=3 (80/20)2026.06 | 95.8 | — | — | — | |
| Rubik+++Network=Swin-UNETR, Pre-training=Medical SSL2024.02 | 95.72 | — | — | — | |
| SwinMMNetwork=Swin-UNETR, Pre-training=Medical SSL2024.02 | 95.52 | — | — | — | |
| Rubik++Network=3D UNet, Pre-training=Medical SSL2024.02 | 95.46 | — | — | — | |
| Swin-UNETRNetwork=Swin-UNETR, Pre-training=Medical SSL2024.02 | 95.33 | — | — | — | |
| JigsawNetwork=Swin-UNETR, Pre-training=General SSL, re-implemented=true2024.02 | 95.24 | — | — | — | |
| RubikNetwork=3D UNet, Pre-training=Medical SSL2024.02 | 94.93 | — | — | — | |
| PCRLv2Network=3D UNet, Pre-training=Medical SSL2024.02 | 94.5 | — | — | — | |
| ROTNetwork=3D UNet, Pre-training=Medical SSL2024.02 | 94.49 | — | — | — | |
| MNet (Original)#Params (M)=8.77, Epochs=150, CV Folds=3 (80/20)2026.06 | 94.4 | — | — | — | |
| MNet + Spatial Gate#Params (M)=8.77, Epochs=150, CV Folds=3 (80/20)2026.06 | 94.4 | — | — | — | |
| JigsawNetwork=3D UNet, Pre-training=General SSL2024.02 | 94.38 | — | — | — | |
| MNet (Original [3])#Params (M)=8.77, Epochs=5002026.06 | 94.3 | — | — | — | |
| MNet (Ours)#Params (M)=8.77, Epochs=150, CV Folds=3 (80/20)2026.06 | 94.3 | — | — | — | |
| KiU-Net 3DType=Voxel2020.10 | 94.23 | — | — | — | |
| PositionLabelNetwork=Swin-UNETR, Pre-training=General SSL, re-implemented=true2024.02 | 94.13 | — | — | — | |
| nnU-Net (Original [3])2026.06 | 94.1 | — | — | — | |
| MAE3DNetwork=UNETR, Pre-training=General SSL2024.02 | 94.02 | — | — | — | |
| PCRLv1Network=3D UNet, Pre-training=Medical SSL2024.02 | 93.87 | — | — | — | |
| MoCo v3+Network=UNETR, Pre-training=General SSL2024.02 | 93.86 | — | — | — | |
| Li et al.Type=Voxel2020.10 | 93.8 | 0 | — | — | |
| U-Net 3DType=Voxel2020.10 | 93.46 | 0 | — | — | |
| Swin-UNETRPre-training=From Scratch, re-implemented=true2024.02 | 93.42 | — | — | — | |
| Dense-UNetType=Voxel2020.10 | 93.36 | 0 | — | — | |
| UNETRPre-training=From Scratch, re-implemented=true2024.02 | 93.25 | — | — | — | |
| TransVWNetwork=3D UNet, Pre-training=Medical SSL2024.02 | 91.42 | — | — | — | |
| MGNetwork=3D UNet, Pre-training=Medical SSL2024.02 | 91.3 | — | — | — | |
| 3D UNetPre-training=From Scratch2024.02 | 90.7 | — | — | — | |
| 3D U-Net (Original [3])2026.06 | 90.1 | — | — | — | |
| Seg-Net 3DType=Voxel2020.10 | 87.89 | 0 | — | — | |
| DeepLab v3+Type=Image2020.10 | 85.7 | 0 | — | — | |
| KiU-NetType=Image2020.10 | 80.35 | — | — | — | |
| U-NetType=Image2020.10 | 77.23 | 0 | — | — | |
| Seg-NetType=Image2020.10 | 76.56 | 0 | — | — |