Multivariate pattern analysis in functional brain imaging

Abstract

The non-invasive recording of brain activity with functional brain imaging greatly advances our understanding of human cognition. At the meantime, more powerful multivariate analysis methods are being developed to compensate the limited capability of traditional univariate approaches. In this review, I will introduce the development of these multivariate methods for functional magnetic resonance imaging (fMRI), the dominant brain imaging technique used in cognitive neuroscience society. The physiological basis of this analysis approach and its future directions will be discussed as well.