gms | German Medical Science

33rd International Congress on Electrocardiology

International Society of Electrocardiology

Multi Foci Based Initial Estimate For Cardiac Activation Time Mapping

Meeting Abstract

  • corresponding author presenting/speaker T. Oostendorp - University Medical Center Nijmegen, Nijmegen, Netherlands
  • A. van Oosterom - University of Lausanne, Lausanne, Schweiz
  • G. Fisscher - University for Health Informatics and Technology, Innsbruck, Austria

33rd International Congress on Electrocardiology. Cologne, 28.06.-01.07.2006. Düsseldorf, Köln: German Medical Science; 2007. Doc06ice055

The electronic version of this article is the complete one and can be found online at:

Published: February 8, 2007

© 2007 Oostendorp et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.



Introduction: In cardiac activation time mapping (ATM), the source is specified by depolarization times at the myocardial surface. The inverse problem for ATM is essentially non-linear, requiring the specification of an initial estimate. The latter greatly influences the quality of the final solution. In this research the performance was tested of a novel initial estimate based on uniform propagation initiated at multiple foci (MF). The results were compared with those resulting from initial estimates as previously documented in the literature: the one derived from the linear relation between activation times and QRS integral map (IM) and the one based on the critical point theorem (CP).

Methods: In MF, first the focus at the ventricular surface is identified that results in the highest correlation between model based and measured ECG. Next, iteratively, additional foci are identified that improve the correlation. The search ends if no new foci are found. Inverse computations were performed based on ECG data measured on a WPW patient. Three beats, with entirely different ECG morphology, were analyzed: 1) a fusion (WPW) beat with simultaneous activation by the Kent bundle and the Hiss bundle, 2) a beat with activation from the Kent bundle only, and 3) an ectopic beat. The location of the Kent bundle was available from catheterization.

Results: For beats 2) and 3), having a single origin, all 3 initial estimates resulted in similar solutions, with the location of the Kent bundle localized as observed during catheterization. For the fusion beat, the IM and CP did not yield a reasonable solution. Specifically, the origin of Hiss bundle initiated activation could not be found. In contrast, the MF based inverse correctly localized the early activation at middle of the left and right ventricular aspect of the septum (indicated by arrows in Figure 1 [Fig. 1]).

Conclusion: The MF based inverse permits activation time mapping in the presence of multiple foci.