Table 2 Results for each indicator on the TUS dataset.
From: Attention residual network for medical ultrasound image segmentation
Method | mIOU(%) | Recall(%) | Precision(%) | F1(%) | Accuracy(%) |
|---|---|---|---|---|---|
U-Net | 65.89 | 79.59 | 80.84 | 77.17 | 94.51 |
U-Net++ | 63.98 | 81.03 | 78.88 | 75.61 | 94.03 |
Attention U-Net | 66.70 | 80.86 | 82.16 | 77.88 | 94.61 |
Segnet | 72.29 | 85.42 | 84.10 | 82.49 | 95.54 |
TransUnet | 71.19 | 86.04 | 81.19 | 81.60 | 95.25 |
HEAT-Net | 85.36 | 87.10 | 90.04 | 87.10 | 96.01 |
MicroSegNet | 76.34 | 88.96 | 90.35 | 89.60 | 96.23 |
ARU-Net | 84.82 | 93.28 | 90.48 | 91.30 | 97.73 |
