Table 2.

Details of training and validation set for the classification problem.

Dataset	Type	Total no. of X-Ray images per class	Training set	Validation set
TBX11 K	Healthy	3800	2660	1140
	Sick but not TB	3800	2660	1140
	TB	800	560	240
Shenzhen	Healthy	167	116	51
	TB	160	112	48
Montgomery	Healthy	31	21	10
	TB	42	29	13
DA + DB	Healthy	54	37	17
	TB	34	23	11
NLM (shenzhen and montgomery), Belarus, NIAID, SNA	Healthy	3500	2100	700
	TB	3500	2100	700