Table 1.

Performance comparison with state-of-the-art methods on BUSI dataset.

Methods	Dice (%)	IoU (%)	Precision (%)	Recall (%)
U-Net	82.7 ± 0.6	70.6 ± 0.7	83.5 ± 0.6	82.0 ± 0.7
U-Net +  +	84.1 ± 0.6	72.6 ± 0.6	84.9 ± 0.6	83.4 ± 0.6
Attention U-Net	85.3 ± 0.5	74.4 ± 0.6	86.1 ± 0.5	84.6 ± 0.6
TransUNet	86.2 ± 0.7	75.8 ± 0.8	89.1 ± 0.6	83.6 ± 0.9
Swin-UNet	87.1 ± 0.5	77.3 ± 0.5	88.0 ± 0.6	86.3 ± 0.6
UNeXt	84.6 ± 0.6	73.3 ± 0.7	85.3 ± 0.6	83.9 ± 0.7
UNETR +  +	86.7 ± 0.6	76.6 ± 0.6	88.5 ± 0.5	85.2 ± 0.6
MedNeXt	87.4 ± 0.5	78.1 ± 0.5	88.1 ± 0.5	86.8 ± 0.5
USEANet	87.6 ± 0.5	78.3 ± 0.5	88.6 ± 0.6	86.9 ± 0.5
Wave-GMS	87.9 ± 0.4	78.7 ± 0.5	88.8 ± 0.5	87.2 ± 0.5
CEA-Net (Ours)	88.3 ± 0.5	79.2 ± 0.6	89.3 ± 0.5	87.8 ± 0.5

Values are reported as mean ± standard deviation over five-fold cross-validation.