• DocumentCode
    674515
  • Title

    Classification of delayed enhancement scar islands by means of their local subendocardial transmurality

  • Author

    Merino-Caviedes, S. ; Cordero-Grande, Lucilio ; Sevilla, Teresa ; Perez, Teresa ; Martin-Fernandez, Marcos ; Alberola-Lopez, Carlos

  • Author_Institution
    Lab. de Procesado de Imagen, Univ. de Valladolid, Valladolid, Spain
  • fYear
    2013
  • fDate
    22-25 Sept. 2013
  • Firstpage
    555
  • Lastpage
    558
  • Abstract
    Delayed Enhancement Magnetic Resonance (DE-MR) has clinical relevance in the diagnosis and prognosis of several cardiomyopathies with different etiologies, since they often lead to different scar configurations. However, to the best of our knowledge very little research has been devoted to the computational characterization of the scar islands detected by DE-MR and its location within the myocardial wall. Our purpose is to classify between subendocardial or transmural (S/T) scar islands and mid-wall or subepicardial (M/E) ones by means of their local subendocardial transmurality (LSTM).The results suggest that inspecting local magnitudes such as LSTM within isolated scar islands in DE-MR provides useful information with respect to the diagnosis of different types of cardiomyopathies and might help in the development of computer assisted diagnostic frameworks for nonischemic cardiomyopathies.
  • Keywords
    biomedical MRI; diseases; muscle; MRI; cardiomyopathy diagnosis; cardiomyopathy prognosis; computational characterization; computer assisted diagnostic frameworks; delayed enhancement magnetic resonance imaging; delayed enhancement scar island classification; etiologies; local subendocardial transmurality; midwall scar islands; myocardial wall; nonischemic cardiomyopathies; scar configurations; subendocardial scar islands; subepicardial scar islands; transmural scar islands; Abstracts; Biomedical imaging; Gold; Heart rate; Magnetic resonance imaging; Markov processes;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computing in Cardiology Conference (CinC), 2013
  • Conference_Location
    Zaragoza
  • ISSN
    2325-8861
  • Print_ISBN
    978-1-4799-0884-4
  • Type

    conf

  • Filename
    6713437