3D Medical Image Segmentation on SUGH
47.22DiceRegFreeNet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| RegFreeNetArchitecture=CNN-based, Param=5.82, FLops=269.252026.01 | 47.22 | 0.3555 | |
| 3DUnetArchitecture=CNN-based, Param=5.75, FLops=321.882026.01 | 45.02 | 0.3325 | |
| 3DUnet++Architecture=CNN-based, Param=6.98, FLops=1335.72026.01 | 44.99 | 0.3316 | |
| SwinUNETRv2Architecture=Transformer-based, Param=72.76, FLops=846.272026.01 | 44.19 | 0.3194 | |
| SwinUNETRArchitecture=Transformer-based, Param=62.19, FLops=785.052026.01 | 44.18 | 0.3188 | |
| DAF3DArchitecture=CNN-based, Param=29.28, FLops=628.342026.01 | 43.43 | 0.3023 | |
| VNetArchitecture=CNN-based, Param=45.60, FLops=759.582026.01 | 43.04 | 0.3169 | |
| UNETR++Architecture=Transformer-based, Param=122.1, FLops=147.042026.01 | 42.37 | 0.3041 | |
| UNETRArchitecture=Transformer-based, Param=93.01, FLops=195.642026.01 | 41.76 | 0.3 | |
| UXNetArchitecture=CNN-based, Param=53.01, FLops=1497.22026.01 | 41.45 | 0.2934 |