Medical Image Segmentation on GlaS
96.91DiceCascaded-MERIT
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Cascaded-MERITArchitecture=Cascaded-MERIT2026.03 | 96.91 | — | — | — | — | — | — | — | — | |
| Deco-Mamba-V1Architecture=Deco-Mamba-V12026.03 | 96.91 | — | — | — | — | — | — | — | — | |
| Deco-Mamba-V0Architecture=Deco-Mamba-V02026.03 | 96.61 | — | — | — | — | — | — | — | — | |
| PVT-EMCAD-B2Architecture=PVT-EMCAD-B22026.03 | 96.38 | — | — | — | — | — | — | — | — | |
| Swin-UMambaArchitecture=Swin-UMamba2026.03 | 95.85 | — | — | — | — | — | — | — | — | |
| VM-UNetArchitecture=VM-UNet2026.03 | 94.25 | — | — | — | — | — | — | — | — | |
| PAG-TransYnetArchitecture=PAG-TransYnet2026.03 | 94.2 | — | — | — | — | — | — | — | — | |
| Mamba-UnetArchitecture=Mamba-Unet2026.03 | 93.41 | — | — | — | — | — | — | — | — | |
| EoSeg2026.06 | 93.27 | — | 87.79 | — | — | — | — | — | — | |
| AutoSAM2023.06 | 92.82 | — | 87.08 | — | — | — | — | — | — | |
| UHR-Net2026.04 | 92.7 | — | 87 | — | 94.7 | 91.5 | — | — | — | |
| EViT-UNet2026.06 | 92.44 | — | 86.5 | — | — | — | — | — | — | |
| SpectraFlow2026.05 | 92.12 | — | 85.63 | — | 93.28 | 91.2 | — | — | — | |
| MedAdaptor-SAMPrompting Strategy=GT points2023.06 | 92.02 | — | 85.88 | — | — | — | — | — | — | |
| Med-URWKV†Type=RWKV, Source=Ours, Params=24.41 M2025.06 | 91.97 | — | 85.76 | — | — | — | — | — | — | |
| ESPNet2026.04 | 91.9 | — | 85 | — | 94.3 | 89.6 | — | — | — | |
| Rich-U-Net2026.03 | 91.84 | — | 84.93 | 1.1515 | — | — | — | — | — | |
| Fully-SupervisedLabel Ratio (%)=100 (Full)2025.10 | 91.7 | — | — | — | — | — | 0.273 | 6.875 | 19.62 | |
| SelfReg + SwinUNet2026.06 | 91.62 | — | 85.29 | — | — | — | — | — | — | |
| ConDSeg2026.04 | 91.6 | — | 85.1 | — | 93.5 | 90.5 | — | — | — | |
| Lightweight decoder h(g(I))2023.06 | 91.51 | — | 84.8 | — | — | — | — | — | — | |
| ConDSeg2026.05 | 91.38 | — | 84.96 | — | 93.17 | 90.24 | — | — | — | |
| 3P-SEGBackbone=Hardnet-852023.06 | 91.19 | — | 84.34 | — | — | — | — | — | — | |
| Med-URWKV-SType=RWKV, Source=Ours, Params=53.07 M2025.06 | 91.13 | — | 84.42 | — | — | — | — | — | — | |
| Med-URWKV-TType=RWKV, Source=Ours, Params=14.33 M2025.06 | 91.02 | — | 84.24 | — | — | — | — | — | — | |
| EMCAD2026.03 | 90.97 | — | 84.9 | 1.7589 | — | — | — | — | — | |
| Swin-UMambaType=Mamba, Source=MICCAI’24, Params=55.06 M2025.06 | 90.91 | — | 84.28 | — | — | — | — | — | — | |
| Attention-UNetType=CNN, Source=MIDL’18, Params=34.88 M2025.06 | 90.78 | — | 83.78 | — | — | — | — | — | — | |
| RWKV-UNetType=RWKV, Source=Arxiv’25, Params=17.40 M2025.06 | 90.57 | — | 83.53 | — | — | — | — | — | — | |
| UCTransNetArchitecture=UCTransNet2026.03 | 90.18 | — | — | — | — | — | — | — | — | |
| UCTransNet2026.06 | 90.18 | — | 82.96 | — | — | — | — | — | — | |
| UNet++Type=CNN, Source=TMI’19, Params=47.19 M2025.06 | 90.13 | — | 82.95 | — | — | — | — | — | — | |
| UCTransNet2023.06 | 89.84 | — | 82.24 | — | — | — | — | — | — | |
| VM-UNetType=Mamba, Source=Arxiv’24, Params=34.62 M2025.06 | 89.8 | — | 82.5 | — | — | — | — | — | — | |
| UCTransNetType=ViT, Source=AAAI’22, Params=66.24 M2025.06 | 89.64 | — | 82.3 | — | — | — | — | — | — | |
| Swin-UNetArchitecture=Swin-UNet2026.03 | 89.58 | — | — | — | — | — | — | — | — | |
| SwimUNet2026.06 | 89.58 | — | 82.06 | — | — | — | — | — | — | |
| InvCoSSParadigm=Continual Self-supervised Learning, Access=Data-free2025.12 | 89.56 | — | — | 45.24 | — | — | — | — | — | |
| Path.Paradigm=Static Pre-training Baseline2025.12 | 89.51 | — | — | 46.79 | — | — | — | — | — | |
| TopoSemiSegLabel Ratio (%)=202025.10 | 89.5 | — | — | — | — | — | 0.51 | 9.825 | 30.462 | |
| OursLabel Ratio (%)=202025.10 | 89.4 | — | — | — | — | — | 0.392 | 7.925 | 26.175 | |
| FCN-Hardnet85Backbone=Hardnet-852023.06 | 89.37 | — | 82.09 | — | — | — | — | — | — | |
| MedCoSSParadigm=Continual Self-supervised Learning, Access=Data-replay2025.12 | 89.13 | — | — | 46.69 | — | — | — | — | — | |
| CaSSLeParadigm=Continual Self-supervised Learning, Access=Data-free2025.12 | 89.12 | — | — | 47.93 | — | — | — | — | — | |
| CMUNeXt2026.04 | 89.1 | — | 81.2 | — | 89.2 | 90.1 | — | — | — | |
| Zig-RiRType=RWKV, Source=TMI’25, Params=24.58 M2025.06 | 89.04 | — | 81.13 | — | — | — | — | — | — | |
| MedT2023.06 | 88.85 | — | 78.93 | — | — | — | — | — | — | |
| UNeXt2026.03 | 88.83 | — | 82.51 | 3.2347 | — | — | — | — | — | |
| Att-UNetArchitecture=Att-UNet2026.03 | 88.8 | — | — | — | — | — | — | — | — | |
| AttUNet2026.06 | 88.8 | — | 80.69 | — | — | — | — | — | — | |
| MRUNet2026.06 | 88.73 | — | 80.89 | — | — | — | — | — | — | |
| PMTLabel Ratio (%)=202025.10 | 88.7 | — | — | — | — | — | 0.698 | 9.98 | 34.805 | |
| UNeXtType=CNN, Source=MICCAI’22, Params=1.47 M2025.06 | 88.56 | — | 80.61 | — | — | — | — | — | — | |
| TransUNetArchitecture=TransUNet2026.03 | 88.4 | — | — | — | — | — | — | — | — | |
| OursLabel Ratio (%)=102025.10 | 88.4 | — | — | — | — | — | 0.501 | 7.85 | 30.525 | |
| XNetLabel Ratio (%)=202025.10 | 88.4 | — | — | — | — | — | 0.735 | 10.188 | 35.298 | |
| TransUNet2026.06 | 88.4 | — | 80.4 | — | — | — | — | — | — | |
| Medical SAM32026.01 | 88.2 | — | 80.7 | — | — | — | — | — | — | |
| EWCParadigm=Continual Self-supervised Learning, Access=Data-free2025.12 | 88.16 | — | — | 50.81 | — | — | — | — | — | |
| ACC-UnetType=CNN, Source=MICCAI’23, Params=16.77 M2025.06 | 87.92 | — | 79.63 | — | — | — | — | — | — | |
| DTAN2026.04 | 87.9 | — | 78.5 | — | 85.8 | 90.2 | — | — | — | |
| DTAN2026.05 | 87.9 | — | 78.55 | — | 88.51 | 90.23 | — | — | — | |
| Joint SSLParadigm=Static Pre-training Baseline, Decoder=Shared2025.12 | 87.86 | — | — | 51.2 | — | — | — | — | — | |
| Joint SSLParadigm=Static Pre-training Baseline, Decoder=Separate2025.12 | 87.83 | — | — | 52.08 | — | — | — | — | — | |
| TopoSemiSegLabel Ratio (%)=102025.10 | 87.8 | — | — | — | — | — | 0.551 | 8.3 | 35.845 | |
| URPCLabel Ratio (%)=202025.10 | 87.8 | — | — | — | — | — | 0.759 | 14.35 | 42.587 | |
| SAMPrompting Strategy=AutoSAM output mask prompt2023.06 | 87.71 | — | 79.92 | — | — | — | — | — | — | |
| ERParadigm=Continual Self-supervised Learning, Access=Data-replay2025.12 | 87.61 | — | — | 53.1 | — | — | — | — | — | |
| UNet++Architecture=UNet++2026.03 | 87.56 | — | — | — | — | — | — | — | — | |
| UNet++2026.06 | 87.56 | — | 79.13 | — | — | — | — | — | — | |
| MGFuseSegType=CNN, Source=BIBM’23, Params=22.27 M2025.06 | 87.41 | — | 78.97 | — | — | — | — | — | — | |
| XNetLabel Ratio (%)=102025.10 | 87.4 | — | — | — | — | — | 0.843 | 14.238 | 40.912 | |
| U-Net++2023.06 | 87.36 | — | 79.03 | — | — | — | — | — | — | |
| CASF-Net2026.05 | 87.21 | — | 78.43 | — | 91.35 | 85.94 | — | — | — | |
| CASF-Net2026.04 | 87.2 | — | 78.4 | — | 91.3 | 85.9 | — | — | — | |
| PMTLabel Ratio (%)=102025.10 | 87.2 | — | — | — | — | — | 0.798 | 13.92 | 39.85 | |
| U-Net++2026.05 | 86.97 | — | 77.61 | — | 89.65 | 85.56 | — | — | — | |
| PackNetParadigm=Continual Self-supervised Learning, Access=Data-free2025.12 | 86.96 | — | — | 52.81 | — | — | — | — | — | |
| U-Net++2026.04 | 86.9 | — | 77.6 | — | 89.6 | 85.5 | — | — | — | |
| Trans-U-Net2026.03 | 86.69 | — | 80.06 | 4.8793 | — | — | — | — | — | |
| UA-MTLabel Ratio (%)=202025.10 | 86.6 | — | — | — | — | — | 1.123 | 18.038 | 53.014 | |
| DCSAU-Net2026.05 | 86.51 | — | 77.63 | — | 93.08 | 82.55 | — | — | — | |
| DCSAU-Net2026.04 | 86.5 | — | 77.6 | — | 93 | 82.5 | — | — | — | |
| EMLabel Ratio (%)=202025.10 | 86.5 | — | — | — | — | — | 1.255 | 17.275 | 58.673 | |
| X-rayParadigm=Static Pre-training Baseline2025.12 | 86.45 | — | — | 47.82 | — | — | — | — | — | |
| MTLabel Ratio (%)=202025.10 | 86.3 | — | — | — | — | — | 2.126 | 29.963 | 64.275 | |
| TFSParadigm=Static Pre-training Baseline2025.12 | 86.2 | — | — | 51.88 | — | — | — | — | — | |
| U-Net2023.06 | 86.05 | — | 75.12 | — | — | — | — | — | — | |
| Attn U-Net2026.05 | 85.94 | — | 76.63 | — | 91.8 | 82.2 | — | — | — | |
| MedT2026.06 | 85.92 | — | 75.47 | — | — | — | — | — | — | |
| PVT-EMCAD-b1Type=CNN, Source=CVPR’24, Params=15.41 M2025.06 | 85.71 | — | 76.01 | — | — | — | — | — | — | |
| DoubleAANet2026.05 | 85.6 | — | 83.28 | — | 88.14 | 90.16 | — | — | — | |
| U-Net2026.05 | 85.51 | — | 75.83 | — | 90.34 | 82.82 | — | — | — | |
| U-Net2026.04 | 85.5 | — | 75.8 | — | 90.3 | 82.8 | — | — | — | |
| U-NetArchitecture=U-Net2026.03 | 85.45 | — | — | — | — | — | — | — | — | |
| U-Net2026.06 | 85.45 | — | 74.78 | — | — | — | — | — | — | |
| TransUNetType=ViT, Source=Arxiv’21, Params=92.23 M2025.06 | 85.29 | — | 75.95 | — | — | — | — | — | — | |
| URPCLabel Ratio (%)=102025.10 | 84.9 | — | — | — | — | — | 1.155 | 19.588 | 54.832 | |
| TGANet2026.04 | 84.7 | — | 71.8 | — | 86.9 | 80.2 | — | — | — | |
| CTParadigm=Static Pre-training Baseline2025.12 | 84.52 | — | — | 50.81 | — | — | — | — | — |