Rigid body registration

This chapter opens with a discussion of how images are transformed via the process of resampling. This chapter is about rigid registration of images and therefore it describes the parameterization of rigid body transformations as a subset of the more general affine transformations. Image registration is important in many aspects of functional image analysis. In imaging neuroscience, particularly for fMRI, the signal changes due to any hæmodynamic response can be small compared to apparent signal differences that can result from subject movement. Subject head movement in the scanner cannot be completely eliminated, so retrospective motion correction is performed as a preprocessing step. This is especially important for experiments where subjects may move in the scanner in a way that is correlated with the different conditions. It also focuses on the methods of rigid body registration, in both intra and intermodality contexts. Intramodality registration implies registration of images acquired using the same modality and scanning sequence or contrast agent, whereas intermodality registration allows the registration of different modalities. © 2004 Elsevier Inc. All rights reserved.