. Author manuscript; available in PMC: 2018 Jun 5.

Published in final edited form as: Nat Neurosci. 2017 Feb 23;20(3):365–377. doi: 10.1038/nn.4478

Table 2.

Named multivariate models in translational neuroimaging

Name	Initials	Predictive of	Features	Algorithms	Refs
Ordinal regression characteristic index of dementia	ORCHID	Progression of AD	sMRI	Ordinal regression with Gaussian process	38
Spatial pattern of abnormality for recognition of early Alzheimer’s disease	SPARE-AD	Early detection of AD	sMRI	SVM with RAVENS methods	124
Alzheimer’s disease pattern similarity	AD-PS	Risk assessment for AD	sMRI	Regularized logistic regression	125
Structural MRI-based brain amyloidosis score	sMRI-BAS	Amyloid beta positive	sMRI	Partial least squares	126
Structural abnormality index	STAND	Diagnosis of AD	sMRI	SVM	127
Parkinson’s disease-related pattern	PDRP	PD status	FDG-PET	Spatial covariance analysis	95
Parkinson’s disease-related cognitive pattern	PDCP	Cognitive dysfunction in PD	FDG-PET	Spatial covariance analysis	128
Parkinson’s disease-related tremor pattern	PDTP	Tremor in PD	FDG-PET	Spatial covariance analysis	129
Multiple-system atrophy-related pattern	MSARP	MSA status	FDG-PET	Spatial covariance analysis	130
Corticobasal degeneration-related pattern	CBDRP	CBD status	FDG-PET	Spatial covariance analysis	131

Note: these models can be applied prospectively to new individuals and data sets to generate predictions about clinical status, symptoms and future outcomes. Their levels of diagnosticity, generalizability and ease of application vary, as do the amounts of prospective testing done to date. However, each of these constitutes a research product that can be shared and characterized across laboratories. CBD, corticobasal degeneration; MSA, multiple-system atrophy; PD, Parkinson’s disease; RAVENS, regional analysis of volumes examined in normalized space; SVM, support vector machine.