Alzheimer’s Disease Diagnosis Based on Cortical and Subcortical Features

Table 12

Algorithm performance comparison over OASIS and ADNI MRI data.

Approach	Year	Dataset	Classifier	Modalities	AD vs HC
					ACC	SEN	SPEC
Cuingnet et al. [23]	2011	ADNI	SVM	MRI	NA	81%	95%
Cho et al. [25]	2012	ADNI	LDA	MRI	NA	82%	93%
Chyzhyk et al. [17]	2012	OASIS	Kernel-LICA-DC	MRI	74.25	96	52.5
Lama et al. [22]	2017	ADNI	RELM	MRI	77.88	68.85	83.54
Jha and Kwon [19]	2017	OASIS	Sparse autoencoder	MRI	91.6	98.09	84.09
Islam and Zhang [20]	2017	OASIS	Ensemble of deep convolutional neural networks	MRI	93.18	NA	93
Farhan et al. [21]	2014	OASIS	Ensemble of classifier	MRI	93.75	100	87.5
Khajehnejad et al. [18]	2017	OASIS	Semisupervised classifier	MRI	93.86	94.65	93.22
Jha et al. [16]	2018	ADNI	ELM	MRI	90.26	90.27	90.20
		OASIS			95.27	96.59	93.03
Proposed method	2018	NRCD	Softmax classifier	MRI	99.34	98.14	100
	2018	OASIS	Softmax classifier	MRI	98.40	93.75	100