Table VII.

SUMMARY OF LEAVE-ONE-OUT CROSS-VALIDATION RESULTS.

Methods	Model	AUC	ACC	SEN	SPE
CT images based CNN	R-IMG	0.61	0.58	0.88	0.28
	C-IMG	0.68	0.68	0.71	0.65
	R-Hybrid	0.71	0.73	0.62	0.84
	C-Hybrid	0.74	0.73	0.78	0.68
	M-IMG (20)	0.83	0.80	0.69	0.84
	M-Hybrid (20)	0.84	0.83	0.93	0.73
	M-IMG (80)	0.78	0.77	0.68	0.87
	M-Hybrid (80)	0.78	0.79	0.65	0.93
	M-IMG (vote)	0.82	0.77	0.94	0.60
	3D-IMG (20)	0.70	0.73	0.61	0.84
	3D-IMG (80)	0.79	0.72	0.78	0.65
GLCM based CNN	R-GLCM	0.74	0.73	0.69	0.77
	C-GLCM	0.80	0.79	0.78	0.81
	M-GLCM	0.89	0.82	0.72	0.94
	3D-GLCM	0.93	0.90	0.90	0.90

ACC, SEN and SPE are short for accuracy, sensitivity and specificity.