Brain Tumor Segmentation on BraTS 2019
94WT Segmentation ScoreReFRM3D
Evaluation Results
| Method | Links | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ReFRM3D2025.12 | 94 | 92.6 | 93.6 | — | — | — | — | — | — | — | — | — | — | — | |
| ReFRM3D2025.12 | 94 | 92.6 | 93.6 | — | — | — | — | — | — | — | — | — | — | — | |
| GMLN-BTSParams (M)=4.58⋆2025.07 | 91.3 | 84.9 | 92 | 89.4 | — | — | — | — | — | — | — | — | — | — | |
| SegFormer3DParams (M)=4.512025.07 | 89.6 | 82.4 | 91.8 | 87.9 | — | — | — | — | — | — | — | — | — | — | |
| Barzegar and Jamzad2025.12 | 88.7 | 89 | 90.1 | — | — | — | — | — | — | — | — | — | — | — | |
| Ullah et al.2025.12 | 87.2 | 86.7 | 90.1 | — | — | — | — | — | — | — | — | — | — | — | |
| SwinUNETRParams (M)=62.192025.07 | 87.2 | 83.1 | 74.5 | 81.6 | — | — | — | — | — | — | — | — | — | — | |
| Li et al.2025.12 | 83.4 | 80.2 | 86.7 | — | — | — | — | — | — | — | — | — | — | — | |
| Xu et al.2025.12 | 82 | 74 | 90 | — | — | — | — | — | — | — | — | — | — | — | |
| nnFormerParams (M)=150.502025.07 | 81.4 | 81.6 | 73.1 | 78.7 | — | — | — | — | — | — | — | — | — | — | |
| SuperLightNetParams (M)=2.972025.07 | 81.2 | 80.9 | 91.3 | 84.5 | — | — | — | — | — | — | — | — | — | — | |
| Liu et al.2025.12 | 81.1 | 76.7 | 89.2 | — | — | — | — | — | — | — | — | — | — | — | |
| Tong and Wang2025.12 | 77.4 | 76.7 | 88.6 | — | — | — | — | — | — | — | — | — | — | — | |
| Liu et al.suffix=a2025.12 | 76.7 | 81.1 | 89.2 | — | — | — | — | — | — | — | — | — | — | — | |
| UNETRParams (M)=92.492025.07 | 74.9 | 64.2 | 53.9 | 64.4 | — | — | — | — | — | — | — | — | — | — | |
| ADRUwAMS2026.04 | — | — | — | — | — | — | — | — | 90.6 | 82.79 | 72.939 | 2.0552 | 3.2217 | 23.7989 | |
| BCPL/U=25/2252026.03 | — | — | — | — | 77.06 | 18.1 | 66.02 | 5.33 | — | — | — | — | — | — | |
| Brats 2019 WinnerImage size=4× 128× 128× 128, Encoder endpoint=128× 16× 16× 16, Loss Function=L(dice), Optimizer=Adam, Learning rate=1e−4, No of Epochs=4052025.02 | — | — | — | — | 88.796 | 4.6181 | — | — | — | — | — | — | — | — | |
| CAMLL/U=25/2252026.03 | — | — | — | — | 78.35 | 13.34 | 67.02 | 3.85 | — | — | — | — | — | — | |
| Cascaded 3D U-Net and 3D U-Net++2026.04 | — | — | — | — | — | — | — | — | 86.7 | 83.4 | 80.2 | — | — | — | |
| CSE-Light-UNETRL/U=25/2252026.03 | — | — | — | — | 79.73 | 11.65 | 68.76 | 2.25 | — | — | — | — | — | — | |
| dual supervision guided attentional network2026.04 | — | — | — | — | — | — | — | — | 88.2 | 77.1 | 72.7 | 8.09 | 10.3 | 6.6 | |
| Fully SupervisedL/U=250/02026.03 | — | — | — | — | 79.93 | 11.44 | 69.33 | 2.26 | — | — | — | — | — | — | |
| heuristic approach for segmentation2026.04 | — | — | — | — | — | — | — | — | 85.98 | 77.28 | 71.53 | — | — | — | |
| Inception-v3Image size=128× 128× 128× 4, Encoder endpoint=8×8×8×256, Loss Function=L(dice)+ L(CF), Optimizer=Adam, Learning rate=0.0001, No of Epochs=100, Time Taken=4min2025.02 | — | — | — | — | 98.07 | 11.1141 | — | — | — | — | — | — | — | — | |
| Inception-v4Image size=128× 128× 128× 4, Encoder endpoint=8×8×8×256, Loss Function=L(dice)+ L(CF), Optimizer=Adam, Learning rate=0.0001, No of Epochs=100, Time Taken=4min2025.02 | — | — | — | — | 98.07 | 11.1141 | — | — | — | — | — | — | — | — | |
| Light-UNETRL/U=25/02026.03 | — | — | — | — | 75.91 | 12.93 | 64.8 | 3.86 | — | — | — | — | — | — | |
| MLRPL/U=25/2252026.03 | — | — | — | — | 78.1 | 12.5 | 67.51 | 3.79 | — | — | — | — | — | — | |
| Multiscale lightweight 3D segmentation with attention mechanism2026.04 | — | — | — | — | — | — | — | — | 89.94 | 83.49 | 77.91 | 5.45 | 6.56 | 4.03 | |
| RAAGR2-Net2026.04 | — | — | — | — | — | — | — | — | 88.4 | 81.4 | 76.3 | — | — | — | |
| ResNetImage size=128× 128× 128× 4, Encoder endpoint=8×8×8×256, Loss Function=L(dice)+ L(CF), Optimizer=Adam, Learning rate=0.0001, No of Epochs=100, Time Taken=4min2025.02 | — | — | — | — | 98.66 | 10.8453 | — | — | — | — | — | — | — | — | |
| SS-NetL/U=25/2252026.03 | — | — | — | — | 78.79 | 13.65 | 67.96 | 3.96 | — | — | — | — | — | — | |
| Swinbts2026.04 | — | — | — | — | — | — | — | — | 89.75 | 79.28 | 74.43 | — | — | — | |
| UA-MTL/U=25/2252026.03 | — | — | — | — | 74.13 | 16.23 | 62.45 | 4.42 | — | — | — | — | — | — | |
| UNETImage size=128× 128× 128× 4, Encoder endpoint=8×8×8×256, Loss Function=L(dice)+ L(CF), Optimizer=Adam, Learning rate=0.0001, No of Epochs=100, Time Taken=4min2025.02 | — | — | — | — | 98.84 | 10.625 | — | — | — | — | — | — | — | — |