Table 4.

Algorithms for the classification of arteries and veins.

Authors	Method	Database(s) used	No. of images	Performance (ACC)
Zamperini et al. [66]	Supervised classifiers	Non-public dataset	42	0.9313
Relan et al. [67]	GMM-EM clustering	Non-public dataset	35	0.92
Dashtbozorg et al. [68]	Graph-based classification	DRIVE; INSPIRE-AVR [86]; VICAVR [137]	138	0.874; 0.883; 0.898
Estrada et al. [69]	Graph-based framework, global likelihood model	Non-public dataset; 1:2:DRIVE; INSPIRE-AVR	110	0.910; 1:0.935, 2:0.917; 0.909
Relan et al. [70]	LS-SVM classification	Non-public dataset; DRIVE	90	0.9488; 0.894