Table 3. Studies on lung segmentation and localization proposed by different researchers.

Study	Dataset	Model	Dice Coef.	Accuracy
Xue et al. (2020)	Shenzhen	Sample Selection + Joint Optimisation	–	0.9253
Rajaraman et al. (2018)	-Shenzhen -Montgomery -Private	Customized VGG16	–	0.917
Zhang et al. (2021)	-Shenzhen (S) -Montgomery (M) -Combination S + M	DEFU-Net	0.9154 0.9227 0.9667	– – 0.9804
Balık & Kaya (2022)	-Kaggle Dataset	U-NET	–	0.92
Sharma et al. (2022)	Kaggle Dataset (Chest Xray Masks and Labels)	U-NET U-NET+	0.9488 0.9235	0.9635 0.9610
Stirenko et al. (2018)	Shenzhen	DCNN		0.74
Rajaraman et al. (2021a)	Shenzhen Tuberculosis X-ray (TBX11K)	VGG16-CXR-U-Net VGG16-CXR-U-Net (AT)	0.5189 0.7552	- -
Rajaraman et al. (2022)	Shenzhen	Stacked Ensemble	0.5743	–
Ngoc et al. (2022)	Shenzhen	Light-weight U-Net	0.7252	–
Proposed model	Shenzhen	RNGU-NET	0.9721	0.9856