Registration and Normalization

Information about an object from different sources can be combined if a transformation allows mapping data from one source to data of the other source. In medical imaging, the two sources are image acquisition systems. If the two sources depict the same subject, this process is called registration....

Full description

Saved in:
Bibliographic Details
Published inGuide to Medical Image Analysis pp. 361 - 404
Main Author Toennies, Klaus D
Format Book Chapter
LanguageEnglish
Published United Kingdom Springer London, Limited 2017
Springer London
SeriesAdvances in Computer Vision and Pattern Recognition
Subjects
Displacement Field
Iterative Close Point Algorithm
Mutual Information
Procrustes Distance
Rigid Registration
Online AccessGet full text
ISBN9781447173182
144717318X
ISSN2191-6586
2191-6594
DOI10.1007/978-1-4471-7320-5_10

Cover

More Information
Summary:Information about an object from different sources can be combined if a transformation allows mapping data from one source to data of the other source. In medical imaging, the two sources are image acquisition systems. If the two sources depict the same subject, this process is called registration. If they depict different subjects, it is called normalization. The mapping is a geometric transformation that accounts for different positioning of a patient in two image acquisition systems. Determining a registration or normalization transformation requires redundant information in the two images, a suitable restriction of acceptable transformations, and, for iterative schemes, a criterion that rates the quality of a given transformation. Various ways to compute a registration or normalization transformation from medical images will be discussed in this chapter.
ISBN:9781447173182
144717318X
ISSN:2191-6586
2191-6594
DOI:10.1007/978-1-4471-7320-5_10