• DocumentCode
    3562209
  • Title

    Detection of acute ischemia episodes from QRS angles changes using a Laplacian noise model

  • Author

    Romero, Daniel ; Martinez, Juan Pablo ; Laguna, Pablo ; Pueyo, Esther

  • Author_Institution
    Lab. Traitement du Signal et de l´Image, Univ. de Rennes I, Rennes, France
  • fYear
    2014
  • Firstpage
    629
  • Lastpage
    632
  • Abstract
    Ischemia detectors represent a useful diagnosis tool to identify acute ischemic episodes in coronary artery disease patients. In this paper, a detector of acute ischemic events based on the analysis of the QRS angles is presented. This acute ischemia detector has been developed by modelling the ischemia-induced changes in the QRS angles as an abrupt change with a certain transition time, assuming a Laplacian noise-model. The standard 12-lead electrocardiogram was used to test the proposed detector. For such proposal, we analyzed 79 patients undergoing a PCI procedure during about 5-min occlusion duration in one of the major coronary artery (LAD=25, RCA=38 and LCX=16). The three detector at the groups of patients presented good outcomes in terms of sensitivity and specificity achieving up to Se=72.7%, Sp=95.5% in the LAD group, Se=75.2%, Sp=97.2% in the RCA group and Se=72.2%, Sp=100%. in the LCX group. We conclude that the QRS angles can be used as a trigger for detecting acute myocardial ischemia although this must be further validated with other contexts in which ischemic events occur more gradually.
  • Keywords
    diseases; electrocardiography; patient diagnosis; 12-lead electrocardiogram; Laplacian noise model; QRS angle changes; QRS angles analysis; acute ischemia episodes detection; acute myocardial ischemia; coronary artery disease; ischemia detector; ischemia-induced changes; time 5 min; Abstracts; Context; Detectors; Electrocardiography; Laplace equations; Lead; Monitoring;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computing in Cardiology Conference (CinC), 2014
  • ISSN
    2325-8861
  • Print_ISBN
    978-1-4799-4346-3
  • Type

    conf

  • Filename
    7043121