Table 2. To evaluate the effects of different segmentation techniques, the test results after training on the York data set are presented in the form of average (std.).

	York
	Dice index		HD (mm)
	LV	MYO	LV	MYO
U-Net	0.90 (0.03)	0.95 (0.03)	9.42 (3.7)	8.32 (3.87)
FC-DenseNet	0.91 (0.03)	0.95 (0.03)	9.15 (3.6)	8.10 (3.83)
NCDN	0.92 (0.03)	0.95 (0.02)	9.07 (3.49)	8.05 (3.62)

Notes.

Optimal values are indicated in bold.