Table 3. The classification accuracies with respect three types of features, including type1) intensity-unadjusted image, type2) histogram-equalized image, and type3) LBP-TOP maps.

		accuracy	specificity	senSitivity	auc
AD VS. NC	TYPE1	63.09%	FAILED	FAILED	FAILED
	TYPE2	80.98%	81.5%	74.5%	0.843
	TYPE3	82.84%	82.7%	80.4%	0.874
MCI VS. NC	TYPE1	43.27%	FAILED	FAILED	FAILED
	TYPE2	52.93%	50.00%	55.20%	0.529
	TYPE3	61.53%	63.50%	61.50%	0.642

The performances are shown in in terms of sensitivity, specificity, AUC, and accuracy rate. Specificity, sensitivity and AUC are not shown for failed tests.