Neural network approaches and their reproducibility in the study of verbal working memory and Alzheimer’s disease

Abstract

As clinical and cognitive neurosciences mature, the need for sophisticated neuroimaging analysis becomes more apparent. Multivariate analysis techniques have recently received increasing attention because they have attractive features that cannot be easily realized by the more commonly used univariate, voxel-wise, techniques. Multivariate approaches evaluate correlation/covariance of activation across brain regions, rather than proceeding on a voxel-by-voxel basis. Thus, their results can be more easily interpreted as a signature of neural networks. Univariate approaches, in contrast, cannot directly address functional connectivity in the brain. Apart from this conceptual difference, the covariance approach can also result in greater statistical power when compared with univariate techniques, which are forced to employ very stringent, and often overly conservative, corrections for voxel-wise multiple comparisons. Multivariate techniques also lend themselves much better to prospective application of results from the analysis of one dataset to entirely new datasets. We provide two examples that illustrate different uses of multivariate techniques in cognitive and clinical neuroscience. We hope this contribution helps facilitate wider dissemination of these techniques in the research community.

Differences between chronological and brain age are related to education and self-reported physical activity

“…We used the scaled subprofile modeling (SSM) (Moeller et al, 1987; Spetsieris and Eidelberg, 2010) approach to calculate physiological brain age (BA) based on regional grey matter volumes from cortical and subcortical locations using the principal components analysis (PCA) toolbox (http://groups.google.-com/group/gcva) (Habeck et al, 2005; Habeck and Stern, 2007). Briefly, the eighty-four grey matter volume regions of interest were subjected to a principal component analysis.…”

Three-fold cross-validation of parkinsonian brain patterns

“…* The number of MSA subjects in the train set for fold 2 is 13 and total patients are 55. For test sets (2,3) and (1,2) MSA subjects are 27 and total subjects are 111. ** Composite accuracy and PPV values are the same.…”

“…By contrast, voxelwise effects using mass univariate analysis are not direct representations of functional connectivity and are not as reproducible because they can include more global or cohort dependent non-disease related statistical variance which can overshadow disease related features [1], [3], [4]. Various newer MRI based approaches are still in early stages of development [21].…”

“…However, in parkinsonism there are numerous unrelated conditions that are indistinguishable in early stages [2] and differences from expected norm values are minimal. In these cases there is an advantage in using multivariate methods of analysis that examine the covariance of voxels once mean effects are removed rather than mass univariate methods of analysis that rely on statistical differences in global mean normalized voxel values to determine areas of disease involvement [1], [3]- [6]. The scaled subprofile model of principal P. G. Spetsieris, V. Dhawan and D. Eidelberg component analysis (SSM/PCA) is a multivariate approach that has revealed invaluable information regarding disease pathways in parkinsonism and other disorders and has provided powerful computational tools for single subject diagnosis [7]- [12].…”

Metabolic networks for assessment of therapy and diagnosis in Parkinson's disease

“…3 This approach is based on principal components analysis and has been described in detail elsewhere. 4,5 In addition to creating the means to identify and validate specific metabolic covariance patterns related to disease, this computational method can be used to quantify these patterns in individual subjects. This application is particularly suitable for the study of novel therapies 6 , and in the investigation of presymptomatic individuals at risk for disease.…”