A Bayesian Time-Course Model for Functional Magnetic Resonance Imaging Data

• Published in: Journal of the American Statistical Association Vol. 95; no. 451; pp. 691 - 703
• Main Author: Genovese, Christopher R.
• Format: Journal Article
• Language: English
• Published: Alexandria, VA Taylor & Francis Group 01.09.2000; American Statistical Association
• Subjects: Applications; Applications and Case Studies; Bayesian inference; Bayesian method; Biological and medical sciences; Cognition; Exact sciences and technology; Functional neuroimaging; Hierarchical models; Hierarchy; Image analysis; Induced substructures; Investigative techniques, diagnostic techniques (general aspects); Linear inference, regression; Magnetic resonance; Magnetic resonance imaging; Mathematics; Medical sciences; Miscellaneous. Technology; Modeling; Non-linear models; Parametric inference; Parametric models; Perceptual localization; Probability and statistics; Radiodiagnosis. Nmr imagery. Nmr spectrometry; Sciences and techniques of general use; Signal noise; Statistics; Time series; Working memory; Dynamic response; Validation; Monte Carlo method; Autoregression; Moving average; Magnetic resonance; Gibbs sampling; Variance analysis; Nuclear magnetic resonance imaging; Spline; Markov chain; Random field; Neurology; Randomization; Cholesky method; Imaging; Tomography; Functional magnetic resonance imaging; Hierarchical model; Bayes methods; Functional imagingN
• ISSN: 0162-1459; 1537-274X
• DOI: 10.1080/01621459.2000.10474253

Functional magnetic resonance imaging (fMRI) is a new technique for studying the workings of the active human brain. During an fMRI experiment, a sequence of magnetic resonance images is acquired while the subject performs specific behavioral tasks. Changes in the measured signal can be used to identify and characterize the brain activity resulting from task performance. The data obtained from an fMRI experiment are a realization of a complex spatiotemporal process with many sources of variation, both biological and technological. This article describes a nonlinear Bayesian hierarchical model for fMRI data and presents inferential methods that enable investigators to directly target their scientific questions of interest, many of which are inaccessible to current methods. The article describes optimization and posterior sampling techniques to fit the model, both of which must be applied many thousands of times for a single dataset. The model is used to analyze data from a psychological experiment and to test a specific prediction of a cognitive theory.