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

Table 8

Comparison between the proposed framework and previous works for identifying MCI patients based on EEG signals.

Study	Year	EEG features	Classifiers	Reported AC
[19]	2016	Spectral features	NF and KNN	88.8%
[21]	2019	Time series signal spectral and features	LC-KSVD and CLC-KSVD	88.9%
[22]	2019	Time and spectral domain features	LR and SVM	87.9%
[23]	2019	Spectral-temporal features	SVM	96.94%
[24]	2019	Spectral, statistical, and nonlinear features	SVM	96.94%
[26]	2020	AR and PE features	ELM, SVM, and KNN	97.64%
Proposed framework	2021	Spectral, functional connectivity and,	LINSVM, RBFSVM, and LR,	99.4%
		Nonlinear features	DT, RB, NB, GB, and KNN