Table 7.

Artificial intelligence classifiers and biomarker input features for highly cited local literature.

Study	PYa	Database	Classifier	Input features
Hinrichs et al [84]	2009	ADNIb	Spatially augmented LPboostingc	MRId+FDG-PETe
Zhang et al [77]	2011	ADNI	Multiple-kernel SVMf	MRI+PET+CSFg
Zhang and Shen [79]	2012	ADNI	M3Th	MRI+PET+CSF
Young et al [80]	2013	ADNI	SVM+GPi	MRI+FDG-PET+CSF+APOEj
Gray et al [82]	2013	ADNI	Random forest	MRI+FDG-PET+CSF+APOE
Moradi et al [78]	2015	ADNI	LDSk+random forest	MRI+aggregate biomarker
Liu et al [83]	2015	ADNI	SAEl+softmax regression+SVM	MRI+FDG-PET
Sorensen et al [85]	2016	ADNI+AIBLm+Metropolit	SVM+logistic regression	MRI+CSF
Lee et al [81]	2019	ADNI	CNNn	MRI+CSF+APOE

^aPY: publication year.

^bADNI: Alzheimer’s Disease Neuroimaging Initiative.

^cLPboosting: linear programming boosting.

^dMRI: magnetic resonance imaging.

^eFDG-PET: fluorodeoxyglucose positron emission tomography.

^fSVM: support vector machine.

^gCSF: cerebrospinal fluid.

^hM3T: multimodal multitask.

ⁱGP: Gaussian process.

^jAPOE: apolipoprotein E.

^kLDS: low density separation.

^lSAE: stacked autoencoder.

^mAIBL: Australian Imaging, Biomarker & Lifestyle.

ⁿCNN: convolutional neural network.