Medical Image Segmentation on ISIC 2017
98.7Dice ScoreMFSNet
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MFSNetTraining time=3hr 51min 20sec2022.03 | 98.7 | 97.4 | 99.9 | 99.9 | — | |
| Navarro et al.2022.03 | 93.8 | 84.6 | — | — | — | |
| Double U-NetTraining time=4hr 18min 07sec2022.03 | 91.3 | 91.8 | 96.3 | 97.4 | — | |
| U-RWKVparams=2.97, FLOPs=7.282025.07 | 90.13 | — | — | — | — | |
| TransUnetparams=93.23, FLOPs=32.232025.07 | 90.1 | — | — | — | — | |
| U-Netparams=34.53, FLOPs=65.522025.07 | 89.87 | — | — | — | — | |
| CMUNeXtparams=3.14, FLOPs=7.412025.07 | 89.85 | — | — | — | — | |
| GuiDINO-W2026.03 | 89.81 | 82.32 | — | — | 13.14 | |
| CMU-Netparams=49.93, FLOPs=91.252025.07 | 89.7 | — | — | — | — | |
| UNeXtparams=1.47, FLOPs=0.572025.07 | 89.6 | — | — | — | — | |
| MSCB-UNet2026.04 | 89.6 | 81.2 | — | — | — | |
| Att-UNetparams=4.91, FLOPs=9.452025.07 | 89.57 | — | — | — | — | |
| nnWNet2026.03 | 89.44 | 82.62 | — | — | 12.73 | |
| U-RWKV-sparams=0.46, FLOPs=1.022025.07 | 89.41 | — | — | — | — | |
| H2Former2026.03 | 89.36 | 82.56 | — | — | 13.04 | |
| ConvUNeXtparams=3.51, FLOPs=7.252025.07 | 89.35 | — | — | — | — | |
| UHR-Net2026.04 | 89.2 | 81.8 | — | — | — | |
| nnUNet2026.03 | 89.18 | 82.3 | — | — | 13.38 | |
| TinyUnetparams=0.48, FLOPs=1.672025.07 | 89.03 | — | — | — | — | |
| BGDiffSeg2026.04 | 88.7 | 79.7 | — | — | — | |
| Ozturk et al.2022.03 | 88.6 | 78.3 | 85.4 | 98.1 | — | |
| SwinUNet2026.03 | 88.38 | 81.24 | — | — | 14.17 | |
| ConDSeg2026.04 | 88.3 | 80.9 | — | — | — | |
| SwinUnetparams=27.15, FLOPs=5.912025.07 | 87.71 | — | — | — | — | |
| Al et al.2022.03 | 87.1 | 77.1 | 85.4 | 96.7 | — | |
| MedTparams=1.37, FLOPs=2.412025.07 | 86.72 | — | — | — | — | |
| RWKV-UNet-SParams.=9.70M, FLOPS=7.64G2025.01 | 86.38 | — | — | — | — | |
| RWKV-UNet-SParams.=9.70M, FLOPS=7.64G2025.01 | 86.38 | — | — | — | — | |
| Xie et al.2022.03 | 86.2 | 78.3 | 87 | 96.4 | — | |
| TransUNetParams.=105.3M, FLOPS=33.42G2025.01 | 86.09 | — | — | — | — | |
| TransUNetParams.=105.3M, FLOPS=33.42G2025.01 | 86.09 | — | — | — | — | |
| Deco-Mamba-V1Architecture=Deco-Mamba-V12026.03 | 86.01 | — | — | — | — | |
| Med-URWKV†Type=RWKV, Source=Ours, Params=24.41 M2025.06 | 85.91 | 77.69 | — | — | — | |
| SegDINO2026.03 | 85.76 | 77.6 | — | — | 20.8 | |
| Cascaded-MERITArchitecture=Cascaded-MERIT2026.03 | 85.67 | — | — | — | — | |
| RWKV-UNet-TParams.=3.15M, FLOPS=3.70G2025.01 | 85.59 | — | — | — | — | |
| RWKV-UNet-TParams.=3.15M, FLOPS=3.70G2025.01 | 85.59 | — | — | — | — | |
| Saha et al.2022.03 | 85.5 | 77.2 | 82.4 | 98.1 | — | |
| Swin-UMambaArchitecture=Swin-UMamba2026.03 | 85.47 | — | — | — | — | |
| RWKV-UNetParams.=17.13M, FLOPS=14.50G2025.01 | 85.32 | — | — | — | — | |
| RWKV-UNetParams.=17.13M, FLOPS=14.50G2025.01 | 85.32 | — | — | — | — | |
| Tschandl et al.2022.03 | 85.3 | 77 | — | — | — | |
| Deco-Mamba-V0Architecture=Deco-Mamba-V02026.03 | 85.14 | — | — | — | — | |
| Swin-UNetArchitecture=Swin-UNet2026.03 | 85.01 | — | — | — | — | |
| FAT-NetArchitecture=FAT-Net2026.03 | 85 | — | — | — | — | |
| FAT-Net2026.04 | 85 | 76.5 | — | — | — | |
| DCSAU-Net2026.04 | 85 | 76.1 | — | — | — | |
| Med-URWKV-TType=RWKV, Source=Ours, Params=14.33 M2025.06 | 84.91 | 76.44 | — | — | — | |
| RWKV-UNetType=RWKV, Source=Arxiv’25, Params=17.40 M2025.06 | 84.89 | 76.26 | — | — | — | |
| Swin-UMambaType=Mamba, Source=MICCAI’24, Params=55.06 M2025.06 | 84.86 | 76.38 | — | — | — | |
| CENet2026.04 | 84.8 | 76.4 | — | — | — | |
| R50-UNetParams.=28.78M, FLOPS=9.72G2025.01 | 84.78 | — | — | — | — | |
| R50-UNetParams.=28.78M, FLOPS=9.72G2025.01 | 84.78 | — | — | — | — | |
| PVT-CASCADEParams.=35.27M, FLOPS=8.20G2025.01 | 84.77 | — | — | — | — | |
| PVT-CASCADEParams.=35.27M, FLOPS=8.20G2025.01 | 84.77 | — | — | — | — | |
| UNextParams.=1.47M, FLOPS=0.58G2025.01 | 84.68 | — | — | — | — | |
| UNextParams.=1.47M, FLOPS=0.58G2025.01 | 84.68 | — | — | — | — | |
| FCBFormerParams.=51.96M, FLOPS=41.22G2025.01 | 84.56 | — | — | — | — | |
| FCBFormerParams.=51.96M, FLOPS=41.22G2025.01 | 84.56 | — | — | — | — | |
| Mamba-UnetArchitecture=Mamba-Unet2026.03 | 84.56 | — | — | — | — | |
| Zig-RiRType=RWKV, Source=TMI’25, Params=24.58 M2025.06 | 84.55 | 75.92 | — | — | — | |
| Med-URWKV-SType=RWKV, Source=Ours, Params=53.07 M2025.06 | 84.51 | 76.1 | — | — | — | |
| H-VmnetType=Mamba, Source=Neurocomputing’25, Params=6.44 M2025.06 | 84.48 | 76.01 | — | — | — | |
| TransUNetType=ViT, Source=Arxiv’21, Params=92.23 M2025.06 | 84.44 | 75.87 | — | — | — | |
| Rolling-UNet-MParams.=7.10M, FLOPS=8.31G2025.01 | 84.42 | — | — | — | — | |
| Rolling-UNet-MParams.=7.10M, FLOPS=8.31G2025.01 | 84.42 | — | — | — | — | |
| PVT-EMCAD-B2Architecture=PVT-EMCAD-B22026.03 | 84.33 | — | — | — | — | |
| Unver et al.2022.03 | 84.3 | 74.8 | 90.8 | 92.7 | — | |
| UNeXtType=CNN, Source=MICCAI’22, Params=1.47 M2025.06 | 84.27 | 75.65 | — | — | — | |
| TransUNetArchitecture=TransUNet2026.03 | 84.1 | — | — | — | — | |
| PVT-EMCAD-B2Params.=26.76M, FLOPS=5.64G2025.01 | 84.05 | — | — | — | — | |
| PVT-EMCAD-B2Params.=26.76M, FLOPS=5.64G2025.01 | 84.05 | — | — | — | — | |
| PAG-TransYnetArchitecture=PAG-TransYnet2026.03 | 83.99 | — | — | — | — | |
| Att-UNetParams.=8.73M, FLOPS=16.78G2025.01 | 83.96 | — | — | — | — | |
| Att-UNetParams.=8.73M, FLOPS=16.78G2025.01 | 83.96 | — | — | — | — | |
| VM-UNetType=Mamba, Source=Arxiv’24, Params=34.62 M2025.06 | 83.68 | 75.19 | — | — | — | |
| PVT-EMCAD-b1Type=CNN, Source=CVPR’24, Params=15.41 M2025.06 | 83.56 | 74.75 | — | — | — | |
| U-KANParams.=25.36M, FLOPS=8.08G2025.01 | 83.46 | — | — | — | — | |
| U-KANParams.=25.36M, FLOPS=8.08G2025.01 | 83.46 | — | — | — | — | |
| UCTransNetArchitecture=UCTransNet2026.03 | 83.46 | — | — | — | — | |
| U-KAN2026.03 | 83.41 | 74.62 | — | — | 23.57 | |
| Swin-UnetType=ViT, Source=ECCV’22, Params=27.15 M2025.06 | 83.27 | 74.2 | — | — | — | |
| U-NetParams.=7.77M, FLOPS=12.16G2025.01 | 82.94 | — | — | — | — | |
| U-NetParams.=7.77M, FLOPS=12.16G2025.01 | 82.94 | — | — | — | — | |
| U-Net2026.04 | 82.8 | 73.7 | — | — | — | |
| MGFuseSegType=CNN, Source=BIBM’23, Params=22.27 M2025.06 | 82.79 | 74.11 | — | — | — | |
| ACC-UnetType=CNN, Source=MICCAI’23, Params=16.77 M2025.06 | 82.35 | 73.81 | — | — | — | |
| UNet++Params.=9.16M, FLOPS=34.71G2025.01 | 82.27 | — | — | — | — | |
| UNet++Params.=9.16M, FLOPS=34.71G2025.01 | 82.27 | — | — | — | — | |
| SegNetTraining time=4hr 04min 17sec2022.03 | 82.1 | 69.6 | 80.1 | 95.4 | — | |
| UCTransNetType=ViT, Source=AAAI’22, Params=66.24 M2025.06 | 82.08 | 73.55 | — | — | — | |
| VM-UNetArchitecture=VM-UNet2026.03 | 81.67 | — | — | — | — | |
| U-NetArchitecture=U-Net2026.03 | 81.59 | — | — | — | — | |
| Attention-UNetType=CNN, Source=MIDL’18, Params=34.88 M2025.06 | 81.42 | 72.51 | — | — | — | |
| MISSFormerType=ViT, Source=TMI’23, Params=35.45 M2025.06 | 81.32 | 72.54 | — | — | — | |
| UNet++Type=CNN, Source=TMI’19, Params=47.19 M2025.06 | 80.9 | 71.72 | — | — | — | |
| Att-UNetArchitecture=Att-UNet2026.03 | 80.82 | — | — | — | — | |
| UNetType=CNN, Source=MICCAI’15, Params=24.89 M2025.06 | 80.82 | 71.66 | — | — | — | |
| UNet++Architecture=UNet++2026.03 | 80.64 | — | — | — | — | |
| Goyel et al.2022.03 | 79.3 | 87.1 | 89.9 | 95 | — |