A Statistical Approach to Motion Compensated Cone Beam Reconstruction

Mark Lyksborg, Mads Fogtmann Hansen, Rasmus Larsen

    Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

    406 Downloads (Pure)


    One of the problems arising in radiotherapy planning is the quality of CT planning data. In the following attention is giving to the cone-beam scanning geometry where reconstruction of a 3D volume based on 2D projections, using the classic Feldkamp-Davis-Kress (FDK) algorithm requires a large number of projections to be adequate. Since the patients are breathing freely during a scan, the number of projections with similar respiration may be to low. In the following we use an iterative reconstruction combined with the simultaneous estimation of the motion field, to improve reconstruction in these situations. Using a simulated dataset we demonstrate that this combination outperforms the FDK but due to ill possessedness of the motion estimation it is only on par with the sole iterative method.
    Original languageEnglish
    Title of host publicationProceedings of IEEE International Symposium on Biomedical Imaging: : From Nano to Macro
    Publication date2010
    ISBN (Print)978-1-4244-4125-9
    Publication statusPublished - 2010
    Event2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro - Rotterdam, Netherlands
    Duration: 14 Apr 201017 Apr 2010
    Conference number: 7


    Conference2010 IEEE International Symposium on Biomedical Imaging
    Internet address

    Bibliographical note

    Copyright 2010 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.


    Dive into the research topics of 'A Statistical Approach to Motion Compensated Cone Beam Reconstruction'. Together they form a unique fingerprint.

    Cite this