Table 3 Comparison of experimental outcomes among various networks on the SegPC-2021 dataset.
From: Medical image segmentation model based on local enhancement driven global optimization
Methods | DSC\(\uparrow\) | HD\(\downarrow\) | Cytoplasm | Nucleus |
|---|---|---|---|---|
U-Net4 | 80.54 | 35.89 | 80.76 | 80.30 |
R50 U-Net7 | 79.21 | 33.26 | 78.65 | 79.78 |
UNet++9 | 75.05 | 32.33 | 72.56 | 77.55 |
UNet3+10 | 77.31 | 34.60 | 76.03 | 78.59 |
R50 Att-UNet7 | 80.15 | 31.74 | 79.98 | 80.32 |
CBAM36 | 80.04 | 33.26 | 79.49 | 80.59 |
SENet37 | 79.00 | 35.31 | 78.43 | 79.57 |
SKNet38 | 80.52 | 30.66 | 79.95 | 81.08 |
Att-UNet8 | 80.40 | 36.04 | 79.55 | 81.25 |
TransUNet19 | 79.69 | 33.95 | 78.68 | 80.68 |
MT-UNet21 | 80.13 | 35.42 | 79.96 | 80.31 |
TransClaw23 | 79.77 | 34.45 | 79.04 | 80.50 |
SwinUNet32 | 78.22 | 35.24 | 77.32 | 79.11 |
TransDeeplab41 | 78.81 | 33.02 | 77.38 | 80.24 |
LEGO-Net(Ours) | 83.57 | 29.36 | 82.31 | 84.84 |
