© Oxford University Press 2001. All rights reserved. This chapter describes the various preprocessing steps necessary to take raw data from the scanner and prepare it for the 'heart' of functional magnetic resonance imaging analysis, namely statistical analysis. These preprocessing steps take the raw MR data, convert it into images that actually look like brains, then reduce unwanted noise of various types and precondition the data in order to aid the later statistics. The chapter also discusses how it is much easier to 'automate' the preprocessing steps than the statistical analysis because optimal tuning of preprocessing algorithms is less dependent on the details of any particular experiment than is the case with later statistics. It discusses the reasons for applying spatial filtering as a preprocessing step and results clearly show the blurring of activation areas as spatial filtering extent increases, even causing 'activation' outside of the brain.

Original publication

DOI

10.1093/acprof:oso/9780192630711.003.0012

Type

Chapter

Book title

Functional Magnetic Resonance Imaging: An Introduction to Methods

Publication Date

22/03/2012