Rigid registration of CT, MR and cryosection images using a GLCM framework

Morten Bro-Nielsen, J. Troccaz (Editor), E. Grimson, R. Mosges

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

    177 Downloads (Pure)

    Abstract

    The majority of the available rigid registration measures are based on a 2-dimensional histogram of corresponding grey-values in the registered images. This paper shows that these features are similar to a family of texture measures based on grey level co-occurrence matrices (GLCM). Features from the GLCM literature are compared to the current range of measures using images from the visible human data set. The voxel-based rigid registration of cryosection and CT images have not been reported before. The tests show that mutual information is the best general measure, but some GLCM features are better for specific modality combinations. This paper discusses existing and some new voxel similarity measures for image registration. Elaborate tests are used to evaluate the different measures and compare them. Finally, a registration algorithm based on voxel similarity measures is described and some results are presented
    Original languageEnglish
    Title of host publicationCVRMed-MRCAS '97. 1st Joint Conf, Comp Vision, VR and Robotics in Medicine and Medical Robotics and Comp-Assisted Surgery Proc
    PublisherSpringer Verlag
    Publication date1997
    Pages171-180
    Publication statusPublished - 1997
    EventCVRMed-MRCAS '97. 1st Joint Conf, Comp Vision, VR and Robotics in Medicine and Medical Robotics and Comp-Assisted Surgery Proc -
    Duration: 1 Jan 1997 → …

    Conference

    ConferenceCVRMed-MRCAS '97. 1st Joint Conf, Comp Vision, VR and Robotics in Medicine and Medical Robotics and Comp-Assisted Surgery Proc
    Period01/01/1997 → …

    Keywords

    • medical image processing
    • image texture
    • matrix algebra
    • image registration
    • biomedical NMR
    • computerised tomography

    Fingerprint

    Dive into the research topics of 'Rigid registration of CT, MR and cryosection images using a GLCM framework'. Together they form a unique fingerprint.

    Cite this