• DocumentCode
    3592514
  • Title

    A mathematical model of the atrioventricular node during atrial fibrillation

  • Author

    Corino, V.D.A. ; Sandberg, F. ; Mainardi, L.T. ; S?¶rnmo, L.

  • Author_Institution
    Dept. of Bioeng., Politec. di Milano, Milan, Italy
  • fYear
    2010
  • Firstpage
    117
  • Lastpage
    120
  • Abstract
    The atrioventricular (AV) node plays a crucial role during atrial fibrillation (AF). The aim of this study is to present an AV node model which can be fitted to short-term ECG recordings in order to infer certain AV node characteristics. The proposed model is characterized by: i) the arrival rate of atrial impulses; ii) two different refractory periods, corresponding to dual AV nodal paths; iii) the probability of an atrial impulse choosing either of these pathways; iv) a parameter modeling prolongation of the refractory period due to different physiological reasons. The model was tested on atrial fibrillatory ECGs recorded from 33 patients; the average normalized absolute error between the normalized RR histogram and the estimated model probability density function was 0.0023 ± 0.0016, (20-ms bin size, 0-2 s interval). These preliminary results are encouraging as AV nodal properties can be noninvasively assessed by a set of statistical parameters with a simple electrophysiological interpretation.
  • Keywords
    electrocardiography; medical signal processing; physiological models; statistical analysis; AV node characteristics; atrial fibrillation; atrial impulses; atrioventricular node; electrophysiological interpretation; physiological reasons; refractory period; refractory periods; short-term ECG recordings; statistical parameters; Biological system modeling; Computational modeling; Data models; Electrocardiography; Frequency estimation; Hidden Markov models; Histograms;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computing in Cardiology, 2010
  • ISSN
    0276-6547
  • Print_ISBN
    978-1-4244-7318-2
  • Type

    conf

  • Filename
    5737923