Assessing common classification methods for the identification of abnormal repolarization using indicators of T-wave morphology and QT interval

  • Authors:

  • Saeed Shakibfar;Claus Graff;Lars H. Ehlers;Egon Toft;Jørgen K. Kanters;Johannes J. Struijk

  • Affiliations:

  • Center for Sensory Motor Interaction (SMI), Department of Health Science and Technology, Aalborg University, Aalborg, Denmark;Medical Informatics (MI), Department of Health Science and Technology, Aalborg University, Aalborg, Denmark;Department of Business Studies, Aalborg University, Aalborg, Denmark;Center for Sensory Motor Interaction (SMI), Department of Health Science and Technology, Aalborg University, Aalborg, Denmark;Department of Cardiology P, Gentofte University Hospital, Gentofte, Denmark and Danish National Research Foundation, Center for Cardiac Arrhythmia (DARC), University of Copenhagen, Copenhagen, Den ...;Medical Informatics (MI), Department of Health Science and Technology, Aalborg University, Aalborg, Denmark

  • Venue:
  • Computers in Biology and Medicine
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Various parameters based on QTc and T-wave morphology have been shown to be useful discriminators for drug induced I"K"r-blocking. Using different classification methods this study compares the potential of these two features for identifying abnormal repolarization on the ECG. A group of healthy volunteers and LQT2 carriers were used to train classification algorithms using measures of T-wave morphology and QTc. The ability to correctly classify a third group of test subjects before and after receiving d,l-sotalol was evaluated using classification rules derived from training. As a single electrocardiographic feature, T-wave morphology separates normal from abnormal repolarization better than QTc. It is further indicated that nonlinear boundaries can provide stronger classifiers than a linear boundaries. Whether this is true in general with other ECG markers and other data sets is uncertain because the approach has not been tested in this setting.