TABLE II.

Comparison of the Vessel Segmentation Results for Superficial Vascular Plexus en face Scans (Mean ± Std. Dev.), Using All Thirty Scans From rose-1 Training Set for Supervised Training. Bold Numbers Indicate the Best Mean Value

Method	AUC	ACC	GMEAN	KAPPA	DSC	FDR
R2U-Net [48]	0.931 ± 0.038	0.917 ± 0.026	0.821 ± 0.062	0.708 ± 0.069	0.757 ± 0.054	0.158 ± 0.042
U-net [27]	0.938 ± 0.037	0.922 ± 0.025	0.834 ± 0.061	0.727 ± 0.067	0.773 ± 0.052	0.147 ± 0.042
nnU-Net [49]	0.935 ± 0.040	0.924 ± 0.026	0.836 ± 0.056	0.734 ± 0.069	0.779 ± 0.054	0.141 ± 0.034
CS-Net [29]	0.932 ± 0.037	0.920 ± 0.027	0.832 ± 0.053	0.724 ± 0.069	0.771 ± 0.052	0.155 ± 0.023
OCTA-Net [25]	0.944 ± 0.031	0.922 ± 0.027	0.828 ± 0.048	0.726 ± 0.067	0.772 ± 0.051	0.142 ± 0.023
ACRROSS(CS-Net)	0.937 ± 0.037	0.921 ± 0.026	0.830 ± 0.062	0.721 ± 0.069	0.768 ± 0.055	0.151 ± 0.039
ACRROSS(U-net)	0.945 ± 0.033	0.924 ± 0.026	0.852 ± 0.053	0.743 ± 0.067	0.788 ± 0.051	0.165 ± 0.030