Table 7.

Results of BCR-UNet and other methods on STARE dataset.

Models	ACC	SEN	SPE	AUC	F1	IOU	MCC
NFN+ (37)	0.9727	0.8096	0.9843	0.9844	-	-	-
SCS-Net (38)	0.9736	0.8207	0.9839	0.9877	-	-	-
Deng et al. (39)	0.9732	0.8272	0.9847	-	0.8196	-	-
U-Net (5)	0.9713	0.7726	0.9876	0.9813	0.8026	0.6709	0.7887
Attention UNet (41)	0.9718	0.7553	0.9896	0.9807	0.8008	0.6687	0.7881
SD-UNet (24)	0.9719	0.7913	0.9865	0.9816	0.8094	0.6806	0.7957
MultiResUNet (27)	0.9730	0.7837	0.9883	0.9730	0.8137	0.6870	0.8017
DRNet (20)	0.9724	0.7855	0.9878	0.9805	0.8110	0.6830	0.7975
BCR-UNet	0.9743	0.8308	0.9860	0.9873	0.8302	0.7103	0.8168

Bold values is the highest scores for the metrics.