Searchlight-based multi-voxel pattern analysis of fMRI by cross-validated MANOVA

Abstract

Multi-voxel pattern analysis (MVPA) is a fruitful and increasingly popular complement to traditional univariate methods of analyzing neuroimaging data. We propose to replace the standard 'decoding' approach to searchlight-based MVPA, measuring the performance of a classifier by its accuracy, with a method based on the multivariate form of the general linear model. Following the well-established methodology of multivariate analysis of variance (MANOVA), we define a measure that directly characterizes the structure of multi-voxel data, the pattern distinctness D. Our measure is related to standard multivariate statistics, but we apply cross-validation to obtain an unbiased estimate of its population value, independent of the amount of data or its partitioning into 'training' and 'test' sets. The estimateD can therefore serve not only as a test statistic, but as an interpretable measure of multivariate effect size. The pattern distinctness generalizes the Mahalanobis distance to an arbitrary number of classes, but also the case where there are no classes of trials because the design is described by parametric regressors. It is defined for arbitrary estimable contrasts, including main effects (pattern differences) and interactions (pattern changes). In this way, our approach makes the full analytical power of complex factorial designs known from univariate fMRI analyses available to MVPA studies. Moreover, we show how the results of a factorial analysis can be used to obtain a measure of pattern stability, the equivalent of 'cross-decoding'.

rsfMRIeffects of KB220Z™ on neural pathways in reward circuitry of abstinent genotyped heroin addicts

“…This was in large part due to the use of the multivoxel pattern analysis approach that was able to identify similar BOLD activation patterns across subjects within each group. 26 In this pilot study we also observed an attenuation of the resultant hyperactivity in the putamen of abstinent heroin-dependent subjects. Currently, albeit knowing that there is a lower D2R availability in the putamen of abstinent heroin dependent subjects, we do not fully understand the mechanism by which KB220Z ™ administration (post one–hour) induced an attenuation of this hypo state.…”

“…The Multi-Voxel Pattern Analysis (MVPA) method is a straightforward application of pattern classification techniques to fMRI datasets, where the patterns to be classified are vectors of voxel activity values (BOLD signal). 25 In the current study, we used a novel variant of the “searchlight” (SL) approach (a form of decoding) 26 to select an appropriate set of voxels in order to define multivariate features as the input for the pattern classification analysis. 27 Following earlier reported procedures by Zhang et al ., the training and test procedures were repeated 24 times.…”

A Primer on Pattern-Based Approaches to fMRI: Principles, Pitfalls, and Perspectives

“…This allows for optimization of classification while at the same time avoiding overfitting and false-positive classification. A different solution to such overfitting is to use an approach that substantially decreases the number of free parameters (e.g., Allefeld and Haynes, 2014). (A) Two different nonlinear polynomial classifiers are fit to a two-class dataset (red and green) for activity measured at two voxels, x and y.…”

MIDAS: Regionally linear multivariate discriminative statistical mapping

“…These techniques use pattern classifiers to harness the rich multivariate information present in the interactions across many voxels to obtain more powerful statistical maps. These approaches were popularized by the introduction of the searchlight methods (Kriegeskorte et al, 2006; Pereira and Botvinick, 2011; Allefeld and Haynes, 2014). Searchlight commonly applies local discriminative classifiers and creates an information map by assigning each searchlight’s classification accuracy to its center voxel.…”