Zebrafish Embryo Vessel Segmentation Using a Novel Dual ResUNet Model

Table 1

Quantitative metrics for ISVs segmentation.

	Accuracy (%)				Dice
	Med.	Mean	Std.	value	Med.	Mean	Std.	value
FCN	85.23	83.21	10.31		0.870	0.843	0.131
DeconvNet	83.62	82.77	10.20		0.852	0.839	0.140
SegNet	85.61	84.62	8.55		0.858	0.841	0.132
U-Net	84.67	83.01	8.62		0.871	0.856	0.123
CU-Net	88.30	85.45	6.22		0.885	0.879	0.092
Dual UNet-14	86.51	85.14	7.34		0.880	0.870	0.091
Dual ResUNet-10	88.11	86.48	5.47		0.891	0.875	0.068
Dual ResUNet-18	89.61	88.80	3.94		0.901	0.885	0.050
Dual ResUNet-14	90.34	88.91	3.21		0.904	0.892	0.044