Automatic Diagnosis of Mild Cognitive Impairment Based on Spectral, Functional Connectivity, and Nonlinear EEG-Based Features

Table 6

The intersection of the selected features in all iterations of 10-fold crossvalidation.

Feature set	Selected features
Functional connectivity	All features
Spectral	All features
Nonlinear	DFA (Fp1, Fp2, F4, Fz, T4, T6, Pz, O1, O2)
	Higuchi (Fp2, F8, T4, P4, Pz, O2, C4, Cz, O2)
	Correlation dimension (Fp1, Fp2, F3, Fz, T3, T4, T6, P3, C3, C4, O1, O2)
	Lyapunov exponent (F3, F4, Fz, T5, P4, O2, C3, C4, Cz, O1, O2)
	C0-complexity (Fp1, F4, T3, P3, P4, Pz, C4, O1)
	Kolmogorov entropy (Fp1, Fp2, F3, F8, T3, T5, T6, Pz, C4, O1, O2)
	Shannon entropy (Fp1, Fp2, F3, Fz, T3, T6, T4, T5, O1, Cz)
	Approximate entropy (Fp1, Fp2, F4, F3, Fz, F7, T6, P3, P4, O2, C3, Cz)