Table 2. Quantitative comparison of specificity and FPR on the BraTS 2020 dataset.

Methods	WT			TC			ET			Average
	Specificity↑	FPR↓		Specificity↑	FPR↓		Specificity↑	FPR↓		Specificity↑	FPR↓
Attention-UNet (9)	0.985	0.015		0.980	0.020		0.970	0.030		0.978	0.022
DAUnet (14)	0.987	0.013		0.980	0.020		0.975	0.025		0.981	0.020
SegResNet (15)	0.990	0.010		0.985	0.015		0.980	0.020		0.985	0.015
nnUNet (16)	0.988	0.012		0.982	0.018		0.976	0.024		0.982	0.018
TransBTS (20)	0.991	0.009		0.985	0.015		0.978	0.022		0.985	0.015
TransUNet (21)	0.990	0.010		0.985	0.015		0.977	0.023		0.984	0.016
UNETR (22)	0.985	0.015		0.980	0.020		0.973	0.027		0.979	0.021
SwinUNETR (23)	0.992	0.008		0.990	0.010		0.985	0.015		0.989	0.011
SwinBTS (38)	0.987	0.013		0.980	0.020		0.974	0.026		0.980	0.020
CKD-TransBTS (24)	0.990	0.010		0.985	0.015		0.978	0.022		0.984	0.016
FDiff-Fusion (39)	0.992	0.008		0.990	0.010		0.985	0.015		0.989	0.011
FCFDiff-Net	0.998*	0.002*		0.999*	0.001*		0.999*	0.001*		0.999*	0.001*

Higher specificity scores (↑) and lower FPR (↓) indicate better performance. The top result is marked with asterisk (*). BraTS, brain tumor segmentation; ET, enhancing tumor; FPR, false positive rate; TC, tumor core; WT, whole tumor.