Statistical issues in fMRI for brain imaging

Functional magnetic resonance imaging is a technique developed in the last decade and used in the fields of cognitive psychology and neuroscience, amo ng others, to study the processes underlying the working of the human brain . In this paper we examine some of the statistical issues in functional mag netic resonance imaging for brain research. We start by giving a brief intr oduction to the physics of magnetic resonance imaging. Using a psychologica l experiment as a case study, we then describe questions of design and stat istical analysis. The data obtained from functional magnetic resonance imag ing studies are of a highly complex nature, displaying both spatial and tem poral correlation, as well as high levels of noise from different sources. Given this, the scope for statistics is vast, and is not limited to simple analysis of the data, once collected.