Table 1 Comparison of segmentation performance across BraTS2023 and AIIB2023 datasets.
From: CDA-mamba: cross-directional attention mamba for enhanced 3D medical image segmentation
Methods | BraTS2023 | AIIB2023 | |||||
|---|---|---|---|---|---|---|---|
WT | TC | ET | Avg | IOU | DLR | DBR | |
nnUnet3 | 92.73 | 89.54 | 83.54 | 88.60 | 87.03 | 61.29 | 50.33 |
TransUnet36 | 92.19 | 88.51 | 83.98 | 88.23 | 86.57 | 62.34 | 48.63 |
UNETR13 | 92.23 | 86.63 | 84.28 | 87.71 | 84.31 | 56.82 | 40.76 |
Swin-UNETR37 | 92.86 | 87.89 | 84.31 | 88.35 | 87.13 | 63.26 | 52.17 |
Swin-UNETR v214 | 93.38 | 89.95 | 85.22 | 89.51 | 87.49 | 64.79 | 53.25 |
MedNeXt38 | 92.49 | 87.83 | 84.05 | 88.12 | 85.78 | 57.98 | 47.43 |
SegMamba29 | 93.60 | 92.65 | 87.71 | 91.32 | 88.59 | 70.21 | 61.28 |
CDAMamba (ours) | 93.84 | 92.71 | 87.76 | 91.44 | 88.72 | 71.01 | 61.53 |
