Table 3. Segmentation performance of different possible combinations of three enhancement methods (LP, WL, FR) and three segmentation methods (GC, TV, LS) on the STARE dataset.

LP, WL and FR denote local phase, wavelet and Frangi’s eigenvalue based enhancement filters respectively. GC, TV and LS denote graph cut, total variation and level set based segmentation methods respectively. Se: sensitivity; Sp: specificity; Acc: accuracy; AUC: area under the curve.

Enhancement	Segmentation	Se	Sp	Acc	AUC	Seg. Time (s)	Entire Time (s)
LP	GC	0.786	0.975	0.951	0.881	0.5	3.5
	TV	0.777	0.960	0.945	0.869	15.2	18.2
	LS	0.775	0.950	0.937	0.863	18.7	21.7
WL	GC	0.640	0.985	0.960	0.813	0.5	4.5
	TV	0.634	0.967	0.948	0.800	15.2	19.2
	LS	0.620	0.960	0.943	0.792	18.7	22.7
FR	GC	0.626	0.976	0.951	0.801	0.5	18.5
	TV	0.423	0.995	0.912	0.709	15.2	33.2
	LS	0.633	0.964	0.938	0.799	18.7	36.7