Table 9.

Summary of 2.5D networks deep learning-based methods.

Study	Year	Datasets	No. of Patients	Performance				Image Modalities	Time/ Equipment	Category
				Dice (%)	ASD (mm)	95HD (mm)	RMSE (mm)
Qiu [5]	2019	In-house PDDCA	109 40	$88.10$ $93.28\pm 1.44$	—	— $1.43\pm 0.56$	$0.58$ —	CT	2.5 min/GPU	DL
Lei [90]	2021	StructSeg2019	50	$91.10\pm 2.90\left(L\right)$; $91.70\pm 1.50\left(R\right)$	—	$2.81\pm 0.45\left(L\right)$; $2.70\pm 0.40\left(R\right)$	—	CT	2 min/GPU	DL
		StructSeg2019 + PDDCA + In-house	50 + 48 + 67	$90.00\pm 4.20$		$6.54\pm 19.14$
Liang [91]	2020	PDDCA In-house	48 96	$94.10\pm 0.70$ $91.10\pm 1.00\left(L\right)$; $91.40\pm 2.00\left(R\right)$	$0.28\pm 0.14$ $0.76\pm 0.13\left(L\right)$; $0.86\pm 0.14\left(R\right)$	—	—	CT	—/GPU	DL
Qiu [92]	2020	In-house PDDCA	109 40	$97.53\pm 1.65$ $95.10\pm 1.21$	$0.21\pm 0.26$ $0.14\pm 0.04$	$2.40\pm 4.61$ $1.36\pm 0.45$	—	CT	1.5 min/GPU	DL