Graphical-model-based multivariate analysis of functional magnetic-resonance data

Neuroimage. 2007 Apr 1;35(2):635-47. doi: 10.1016/j.neuroimage.2006.11.040. Epub 2007 Jan 25.

Abstract

We propose a method for the analysis of functional magnetic-resonance (fMR) data, based on a Bayesian-network representation. Our method identifies multivariate linear/nonlinear voxel-activation pattern differences across groups, which may provide information complementary to that resulting from a general linear model (GLM)-based analysis. In addition, we describe a model-stabilization method based on data resampling, which may be helpful in the presence of small numbers of subjects, or when data are noisy.

Publication types

  • Research Support, N.I.H., Extramural
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Brain / physiology*
  • Brain / physiopathology
  • Dementia / physiopathology*
  • Female
  • Humans
  • Linear Models
  • Magnetic Resonance Imaging / statistics & numerical data*
  • Male
  • Multivariate Analysis