3D Medical Image Segmentation on BTCV (test)
81.66DICEnnUNet + DTC
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| nnUNet + DTCArchitecture=CNN-based, Upsampling=Deformable Transposed Convolution2026.01 | 81.66 | 75.97 | |
| nnUNetArchitecture=CNN-based2026.01 | 81.52 | 76.1 | |
| nnMamba + DTCArchitecture=Mamba-based, Upsampling=Deformable Transposed Convolution2026.01 | 76.77 | 68.19 | |
| nnMambaArchitecture=Mamba-based2026.01 | 75.47 | 67.05 | |
| UNETR + DTCArchitecture=Transformer-based, Upsampling=Deformable Transposed Convolution2026.01 | 71.98 | 62.15 | |
| UNETRArchitecture=Transformer-based2026.01 | 69.65 | 59.18 | |
| DRDTeacher=SAT 21.57, Student=Unet3D2026.05 | 66.47 | — |