Table 3 Experiment results of lung feild segmentation
Family | Methods | Year | Dice(%) | IoU(%) | Precision(%) | Recall(%) |
|---|---|---|---|---|---|---|
CNN | UNet [1]\(^*\) | 2015 | 95.10±1.33 | 90.72±2.37 | 96.38±2.18 | 90.66±3.18 |
UNet++ [2]\(^*\) | 2018 | 93.48±2.49 | 88.11±4.10 | 95.87±1.73 | 95.00±1.78 | |
CENet [4]\(^*\) | 2019 | 96.53±2.81 | 92.60±4.53 | 96.76±2.08 | 94.81±1.95 | |
Atten_UNet [3]\(^*\) | 2018 | 95.20±2.36 | 91.39±2.48 | 97.42±1.56 | 91.54±1.07 | |
XlSor [37]\(^*\) | 2019 | 97.54 | - | 97.40 | 97.73 | |
CA-Net [5]\(^*\) | 2020 | 95.95±1.50 | - | - | - | |
Transformer | SwinUnet [26]\(^*\) | 2021 | 95.58 | 93.31 | 96.93 | 94.34 |
TransUNet(ViT) [10]\(^*\) | 2021 | 96.89 | 93.98 | 98.19 | 95.63 | |
TransUNet(R50) [10]\(^*\) | 2021 | 97.03 | 94.23 | 98.37 | 95.02 | |
iU-Net(Ours) | 2022 | 97.21 | 94.35 | 98.54 | 96.75 |