A Bayesian multilevel model for fMRI data analysis

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

Bayesian approaches have been proposed by several functional magnetic resonance imaging (fMRI) researchers in order to overcome the fundamental limitations of the popular statistical parametric mapping method. However, the difficulties associated with subjective prior elicitation have prevented the widespread adoption of the Bayesian methodology by the neuroimaging community. In this paper, we present a Bayesian multilevel model for the analysis of brain fMRI data. The main idea is to consider that all the estimated group effects (fMRI activation patterns) are exchangeable. This means that all the collected voxel time series are considered manifestations of a few common underlying phenomena. In contradistinction to other Bayesian approaches, we think of the estimated activations as multivariate random draws from the same distribution without imposing specific prior spatial and/or temporal information for the interaction between voxels. Instead, a two-stage empirical Bayes prior approach is used to relate voxel regression equations through correlations between the regression coefficient vectors. The adaptive shrinkage properties of the Bayesian multilevel methodology are exploited to deal with spatial variations, and noise outliers. The characteristics of the proposed model are evaluated by considering its application to two real data sets.
Original languageUnknown
Pages (from-to)238-52
JournalComputer Methods and Programs in Biomedicine
Volume102
Issue number3
DOIs
Publication statusPublished - 1 Jan 2011

Cite this

@article{4a8c8b4d91b941bdb881c21278121c39,
title = "A Bayesian multilevel model for fMRI data analysis",
abstract = "Bayesian approaches have been proposed by several functional magnetic resonance imaging (fMRI) researchers in order to overcome the fundamental limitations of the popular statistical parametric mapping method. However, the difficulties associated with subjective prior elicitation have prevented the widespread adoption of the Bayesian methodology by the neuroimaging community. In this paper, we present a Bayesian multilevel model for the analysis of brain fMRI data. The main idea is to consider that all the estimated group effects (fMRI activation patterns) are exchangeable. This means that all the collected voxel time series are considered manifestations of a few common underlying phenomena. In contradistinction to other Bayesian approaches, we think of the estimated activations as multivariate random draws from the same distribution without imposing specific prior spatial and/or temporal information for the interaction between voxels. Instead, a two-stage empirical Bayes prior approach is used to relate voxel regression equations through correlations between the regression coefficient vectors. The adaptive shrinkage properties of the Bayesian multilevel methodology are exploited to deal with spatial variations, and noise outliers. The characteristics of the proposed model are evaluated by considering its application to two real data sets.",
keywords = "fMRI Data Analysis, Bayesian regression",
author = "{da Silva}, {Adelino Rocha Ferreira} and {DEE Group Author}",
year = "2011",
month = "1",
day = "1",
doi = "10.1016/j.cmpb.2010.05.003",
language = "Unknown",
volume = "102",
pages = "238--52",
journal = "Computer Methods and Programs in Biomedicine",
issn = "0169-2607",
publisher = "Elsevier Science B.V., Amsterdam.",
number = "3",

}

A Bayesian multilevel model for fMRI data analysis. / da Silva, Adelino Rocha Ferreira; DEE Group Author.

In: Computer Methods and Programs in Biomedicine, Vol. 102, No. 3, 01.01.2011, p. 238-52.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A Bayesian multilevel model for fMRI data analysis

AU - da Silva, Adelino Rocha Ferreira

AU - DEE Group Author

PY - 2011/1/1

Y1 - 2011/1/1

N2 - Bayesian approaches have been proposed by several functional magnetic resonance imaging (fMRI) researchers in order to overcome the fundamental limitations of the popular statistical parametric mapping method. However, the difficulties associated with subjective prior elicitation have prevented the widespread adoption of the Bayesian methodology by the neuroimaging community. In this paper, we present a Bayesian multilevel model for the analysis of brain fMRI data. The main idea is to consider that all the estimated group effects (fMRI activation patterns) are exchangeable. This means that all the collected voxel time series are considered manifestations of a few common underlying phenomena. In contradistinction to other Bayesian approaches, we think of the estimated activations as multivariate random draws from the same distribution without imposing specific prior spatial and/or temporal information for the interaction between voxels. Instead, a two-stage empirical Bayes prior approach is used to relate voxel regression equations through correlations between the regression coefficient vectors. The adaptive shrinkage properties of the Bayesian multilevel methodology are exploited to deal with spatial variations, and noise outliers. The characteristics of the proposed model are evaluated by considering its application to two real data sets.

AB - Bayesian approaches have been proposed by several functional magnetic resonance imaging (fMRI) researchers in order to overcome the fundamental limitations of the popular statistical parametric mapping method. However, the difficulties associated with subjective prior elicitation have prevented the widespread adoption of the Bayesian methodology by the neuroimaging community. In this paper, we present a Bayesian multilevel model for the analysis of brain fMRI data. The main idea is to consider that all the estimated group effects (fMRI activation patterns) are exchangeable. This means that all the collected voxel time series are considered manifestations of a few common underlying phenomena. In contradistinction to other Bayesian approaches, we think of the estimated activations as multivariate random draws from the same distribution without imposing specific prior spatial and/or temporal information for the interaction between voxels. Instead, a two-stage empirical Bayes prior approach is used to relate voxel regression equations through correlations between the regression coefficient vectors. The adaptive shrinkage properties of the Bayesian multilevel methodology are exploited to deal with spatial variations, and noise outliers. The characteristics of the proposed model are evaluated by considering its application to two real data sets.

KW - fMRI Data Analysis

KW - Bayesian regression

U2 - 10.1016/j.cmpb.2010.05.003

DO - 10.1016/j.cmpb.2010.05.003

M3 - Article

VL - 102

SP - 238

EP - 252

JO - Computer Methods and Programs in Biomedicine

JF - Computer Methods and Programs in Biomedicine

SN - 0169-2607

IS - 3

ER -