TABLE 10.

Comparison of performance of binary classification MCI against HC with state of the art methods using rs-fMRI.

Dataset	Feature measures	Classifier	Accuracy (%)	References
MCI: 31, HC: 31	Functional activity co-variations of ROIs	SVM	62.90	Eavani et al., 2013
MCI: 31, HC: 31	Group sparse representation	SVM	66.13	Wee et al., 2014
MCI: 31, HC: 31	SDFN	SVM	70.97	Leonardi et al., 2013
MCI: 31, HC: 31	Deep auto encoder and HMM	SVM	72.58	Suk et al., 2016

FC, functional connectivity; ALFF, amplitude of low-frequency fluctuation; AUC, area under the curve; DMN, default mode network; SN, salience network; LDA, linear discriminant analysis; ROC, receiver operating characteristic; ROI, region of interest; AAL, automated anatomical labeling; DMN, default mode network; SDFN, sliding window-based dynamic functional network; HMM, hidden markov model.