• DocumentCode
    561845
  • Title

    Automatic quantification of cardiac scar extent from late gadolinium enhancement magnetic resonance imaging

  • Author

    Corsi, Cristiana ; Tarroni, Giacomo ; Tornani, Alessandro ; Severi, Stefano ; Lamberti, Claudio

  • Author_Institution
    DEIS, Univ. of Bologna, Bologna, Italy
  • fYear
    2011
  • fDate
    18-21 Sept. 2011
  • Firstpage
    405
  • Lastpage
    408
  • Abstract
    Late gadolinium enhancement cardiac magnetic resonance imaging (LGE-CMRI) is the technique of choice to detect myocardial scars and assess myocardial viability. In clinical practice, this analysis is performed qualitatively or by manually tracing the enhanced area in each acquired slice. The purpose of this study was to test and validate a technique for automated localization and quantification of scar extent. CMRI data in patients with previous myocardial infarction were analyzed using custom software from which the myocardium was automatically identified from steady-state free precession images and registered on LGE-CMRI data. Scar tissue was defined as myocardium with signal intensity ≥80% of its maximum and quantified on each slice. Scar location and extent were assessed and compared with expert analysis. Preliminary results showed that automatic localization of scar from LGE-CMRI is feasible and scar quantification is accurate and reliable.
  • Keywords
    biological tissues; biomedical MRI; cardiovascular system; image enhancement; medical expert systems; medical image processing; LGE-CMRI data; automated localization; automated quantification; automatic localization; automatic quantification; cardiac scar extent; custom software; expert analysis; late gadolinium enhancement cardiac magnetic resonance imaging; manual tracing; myocardial infarction; myocardial scar detection; myocardial viability; myocardium; scar location; scar quantification; scar tissue; signal intensity; steady-state free precession images; Educational institutions; Image edge detection; Image segmentation; Level set; Magnetic resonance imaging; Motion segmentation; Myocardium;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computing in Cardiology, 2011
  • Conference_Location
    Hangzhou
  • ISSN
    0276-6547
  • Print_ISBN
    978-1-4577-0612-7
  • Type

    conf

  • Filename
    6164588