[Retracted] An Automated Deep Learning Model for the Cerebellum Segmentation from Fetal Brain Images

Table 3

Segmentation fetal cerebellum evaluation.

Methods	Losing variable	Precision	Recall	DSC	F1-score	HD	Accuracy	rate	Proceeding time (seconds)
Attention U-Net	FTL	0.84	0.79	0.84	0.85	40.05	0.92	4.20-21	0.50
	CL	0.89	0.82	0.87	0.84	35.87	0.94	6.12-87	0.45
	DL	0.86	0.85	0.82	0.88	34.25	0.91	5.24-87	0.42
U-Net++	FTL	0.91	0.84	0.91	0.87	28.67	0.89	7.24-65	0.35
	CL	0.89	0.87	0.89	0.82	25.50	0.85	4.20-30	0.40
	DL	0.92	0.88	0.92	0.84	24.26	0.92	6.54-40	0.32
U-Net	FTL	0.94	0.90	0.87	0.90	22.56	0.94	5.23-01	0.30
	CL	0.91	0.92	0.88	0.91	30.02	0.92	3.20-62	0.32
	DL	0.89	0.89	0.84	0.89	20.26	0.93	2.50-64	0.38
ResU-Net-c	FTL	0.90	0.94	0.91	—	18.2	—	3.02-01	0.32
	CL	0.95	0.9	0.91	—	17.8	—	9.52-01	0.30
	DL	0.93	0.92	0.92	—	17.9	—	—	0.34
Proposed ReU-Net	FTL	0.94	0.95	0.94	0.92	18.24	0.97	4.50-05	0.25
	CL	0.96	0.93	0.92	0.93	15.25	0.95	10.45-21	0.20
	DL	0.94	0.92	0.93	0.94	15.78	0.98	—	0.23