Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

We are developing a system for patient management in colorectal cancer, in which the need for segmentation and non-rigid registration of pre- and post-therapy images arises. Several methods for nonrigid registration have been proposed, all of which embody a 'generic' algorithm to solve registration, largely irrespective both of the kinds of images and of the application. We have evaluated several of these algorithms for this application and find their performance unsuitable for aligning pre- and post- therapy colorectal images. This leads us to identify some of the implicit assumptions and fundamental limitations of these algorithms. None of the currently available algorithms take into account the issue of scale salience and more importantly, none of the algorithms "know" enough about colorectal MRI to focus their attention for registration on those parts of the image that are clinically important. Based on this analysis, we propose a way in which we can perform registration by mobilizing the knowledge of the particular application, for example the prior shape knowledge that we have within the colorectal images as well as knowledge of the large scale non-rigid changes due to therapy. © Springer-Verlag Berlin Heidelberg 2005.

Original publication




Conference paper

Publication Date



3765 LNCS


399 - 408