An Efficient Combination among sMRI, CSF, Cognitive Score, and APOE ε4 Biomarkers for Classification of AD and MCI Using Extreme Learning Machine

Table 6

10-fold cross-validated classification performance for MCIs vs. MCIc.

Features measure	ELM				SVM-RBF			SVM-linear
	ACC%	SEN%	SPE%	AUC	ACC%	SEN%	SPE%	ACC%	SEN%	SPE%
Cortical thickness	63.71	71.10	38.37	0.64	57.18	65.53	60.38	50.05	71.18	68.13
Surface area	61.43	72.28	74.80	0.69	54.25	63.22	76.17	58.27	72.23	64.85
Volume	67.85	84.40	87.73	0.74	59.17	78.28	67.55	62.14	74.45	57.73
CSF	71.37	78.84	68.57	0.72	64.73	60.09	72.78	67.33	78.82	68.78
APOE + MMSE	76.09	87.13	72.09	0.79	65.53	68.81	67.43	68.08	72.21	66.33
Concatenation (all_features_set)	83.38	93.01	75.77	0.85	71.37	81.16	76.75	75.71	78.38	84.67