Table 5 The segmentation results (mean ± std) on the external validation dataset GUSI-B. The best results are highlighted in bold.
Metric | Ours | U-net | AAU-Net | R2U-Net | LGANet | FAT-Net | Swin-UNet | Transfuse |
|---|---|---|---|---|---|---|---|---|
Accuracy | 98.22 ± 0.22 | 98.02 ± 0.19 | 98.06 ± 0.24 | 96.21 ±0.38 | 97.93±0.19 | 97.93±0.21 | 97.53±0.19 | 98.02±0.41 |
Precision | 88.10 ± 2.81 | 86.83±5.57 | 85.10±6.04 | 73.36±1.91 | 86.97±5.87 | 87.62±4.57 | 85.07±0.77 | 87.00±7.14 |
Recall | 90.49±0.70 | 90.13 ± 5.22 | 94.30 ±4.51 | 84.12 ± 8.82 | 88.68±5.86 | 87.48±5.63 | 84.49±1.91 | 90.32±5.73 |
IoU | 80.60± 1.87 | 78.78±1.07 | 80.28±1.53 | 64.16±4.65 | 77.67±1.15 | 77.44±2.30 | 73.59±1.93 | 78.91±3.01 |
Dice | 89.25 ± 1.14 | 88.12±0.67 | 88.98±0.91 | 78.07±3.47 | 87.43±0.73 | 87.26±1.45 | 84.77±1.28 | 88.19±1.89 |
Specificity | 98.91 ± 0.30 | 98.72±0.63 | 98.39±0.66 | 97.27±0.48 | 98.74±0.68 | 98.86±0.50 | 98.69±0.06 | 98.71±0.79 |
| ||||||||
