Multiscale U-Net with Spatial Positional Attention for Retinal Vessel Segmentation

Table 5

The results of different algorithms in the CHASE_DB1 dataset.

Dataset	Methods	Year	Se (%)	Ac (%)	AUC (%)	F1-score (%)
CHASE_DB1	R2U-Net [30]	2018	77.92	95.56	97.84	81.71
	U-Net [30]	2018	82.88	95.78	97.72	77.83
	LadderNet [31]	2018	79.78	96.56	98.39	80.31
	DEU-Net [33]	2019	80.74	96.61	98.12	80.37
	IterNet [12]	2019	80.73	96.55	98.51	80.73
	AG-Net [34]	2019	81.86	97.43	98.63	N/A
	Lü et al. [37]	2020	81.35	96.17	97.82	N/A
	RVSeg-Net [39]	2020	80.69	97.26	98.33	N/A
	Proposed method	2021	81.49	97.51	99.01	83.55

The best values of Se, Ac, AUC, and F1-score are shown in bold.