3D Medical Image Segmentation on BraTS (test)
1.5Memory Usage (GB)RefineFormer3D
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| RefineFormer3DParams (M)=2.94, FLOPs (G)=191.2, Input resolution=1 x 4 x 128^3, Environment=RTX 5080, Ryzen 9 9950X, PyTorch 2.8, CUDA 12.82026.02 | 1.5 | 8.35 | 296 | |
| DS-UNETR++Params (M)=42.6, FLOPs (G)=70.1, Input resolution=1 x 4 x 128^3, Environment=RTX 5080, Ryzen 9 9950X, PyTorch 2.8, CUDA 12.82026.02 | 2.4 | 62.4 | 1,498 | |
| UNETRParams (M)=92.5, FLOPs (G)=153.5, Input resolution=1 x 4 x 128^3, Environment=RTX 5080, Ryzen 9 9950X, PyTorch 2.8, CUDA 12.82026.02 | 3.3 | 82.5 | 2,145 | |
| nnFormerParams (M)=149.6, FLOPs (G)=421.5, Input resolution=1 x 4 x 128^3, Environment=RTX 5080, Ryzen 9 9950X, PyTorch 2.8, CUDA 12.82026.02 | 12.6 | 148 | 5,248 | |
| SwinUNETRParams (M)=62.8, FLOPs (G)=572.4, Input resolution=1 x 4 x 128^3, Environment=RTX 5080, Ryzen 9 9950X, PyTorch 2.8, CUDA 12.82026.02 | 19.7 | 228.6 | 7,612 |