Table 3.

Investigating the performance of different configurations of the proposed method (mean ± standard deviation). Best results are in bold.

Model	ACC (%)	DSC (%)	IoU (%)	SEN (%)	SPE (%)
Baseline	$0.9844\pm 0.0098$	$0.7556\pm 0.1170$	$0.6195\pm 0.1268$	$0.7762\pm 0.1479$	$0.9920\pm 0.0041$
Baseline + DWT	$0.9847\pm 0.0092$	$0.7641\pm 0.1134$	$0.6361\pm 0.1202$	$0.7888\pm 0.1446$	$0.9927\pm 0.0040$
Baseline + LPDGC	$0.9859\pm 0.0088$	$0.7696\pm 0.1117$	$0.6415\pm 0.1127$	$0.7892\pm 0.1426$	$0.9934\pm 0.0036$
Baseline + FAM	$0.9860\pm 0.0080$	$0.7703\pm 0.1092$	$0.6389\pm 0.1164$	$0.7974\pm 0.1391$	$0.9939\pm 0.0032$
Baseline + DWT + LPDGC	$0.9865\pm 0.0074$	$0.7742\pm 0.1070$	$0.6420\pm 0.1151$	$0.8062\pm 0.1403$	$0.9942\pm 0.0030$
Baseline + DWT + FAM	$0.9867\pm 0.0062$	$0.7785\pm 0.1058$	$0.6510\pm 0.1142$	$0.8148\pm 0.1395$	$0.9944\pm 0.0028$
LungINFseg (w/o augmentation)	$0.9874\pm 0.0057$	$0.7853\pm 0.1050$	$0.6522\pm 0.1136$	$0.8229\pm 0.1374$	$0.9948\pm 0.0025$
LungINFseg (with augmentation)	$\mathbf{0.9892}\mathbf{\pm }\mathbf{0.0054}$	$\mathbf{0.8034}\mathbf{\pm }\mathbf{0.1021}$	$\mathbf{0.6877}\mathbf{\pm }\mathbf{0.1095}$	$\mathbf{0.8310}\mathbf{\pm }\mathbf{0.1323}$	$\mathbf{0.9952}\mathbf{\pm }\mathbf{0.0022}$