Level-Set Based Carotid Artery Segmentation for Stenosis Grading

C.M. van Bemmel, L.J. Spreeuwers, M.A. Viergever, W.J. Niessen

    Research output: Contribution to conferencePaperpeer-review

    15 Citations (Scopus)

    Abstract

    A semi-automated method is presented for the determination of the degree of stenosis of the internal carotid artery (ICA) in 3D contrast-enhanced (CE) MR angiograms. Hereto, we determined the central vessel axis (CA), which subsequently is used as an initialization for a level-set based segmentation of the stenosed carotid artery. The degree of stenosis is determined by calculating the average diameters of cross-sectional planes along the CA. For twelve ICAs the degree of stenosis was determined and correlated with the scores of two experts (NASCET criterion). The Spearman’s correlation coefficient for the proposed method was 0.96 (p<0.001), versus 0.89 and 0.88 (p<0.001) for the manual scores, and a smaller bias and tighter confidence bounds for the automated method were found.
    Original languageEnglish
    Pages36-43
    Number of pages8
    DOIs
    Publication statusPublished - 2002
    Event5th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2002 - Tokyo, Japan
    Duration: 25 Sept 200228 Sept 2002
    Conference number: 5

    Conference

    Conference5th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2002
    Abbreviated titleMICCAI
    Country/TerritoryJapan
    CityTokyo
    Period25/09/0228/09/02
    OtherSeptember 25-28, 2002

    Keywords

    • IR-65109
    • EWI-14094

    Fingerprint

    Dive into the research topics of 'Level-Set Based Carotid Artery Segmentation for Stenosis Grading'. Together they form a unique fingerprint.

    Cite this