gms | German Medical Science

33rd International Congress on Electrocardiology

International Society of Electrocardiology

Multiple regression model of beat-to-beat spatial repolarization variability

Meeting Abstract

Search Medline for

  • G. Kozmann - Dept. of Information Systems, Veszprém, Ungarn
  • corresponding author presenting/speaker K. Haraszti - Dept. of Biomedical Eng., Budapest, Ungarn

33rd International Congress on Electrocardiology. Cologne, 28.06.-01.07.2006. Düsseldorf, Köln: German Medical Science; 2007. Doc06ice106

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

Published: February 8, 2007

© 2007 Kozmann 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.



Question: In this study the spatio-temporal dynamics of ventricular depolarization and repolarization have been studied on a beat-to-beat basis.

Method Used: Spatio-temporal dynamics were assessed by 192-lead body surface potential map (BSPM) records, evaluated beat-by-beat. Healthy subjects and ICD patients were enrolled in the study. All the measurements were taken in supine rest position and in each case 5 minutes were stored. In data processing premature ectopic beats were excluded, only the statistical properties of normal sinus rhythm were taken into consideration. The spatial nature of activation and repolarization variability was characterized by QRS and QRST integral maps both in the domain of measurements and 12D space of Karhunen-Loeve (KL) components. A concise characterization of spatial repolarization variability was done in terms of the nondipolarity indices (NDI). Furthermore, heart rate variability (HRV) was expressed by the Fourier transform of short term RR distance records.

Results: In healthy subjects multiple regression revealed significant within beat correlation of (the 192D) QRS and QRST integral map vectors. The regression of QRST vs. QRS integral NDI values in healthy subjects yielded R2>0.4 median value. The involvement of RR as a second independent variable into the model only slightly increased R2, i.e. the variability of RR has only minor impact of the spatial repolarization inhomogeneity. In ICD patients R2 decreased significantly (median: 0.23), due to the diminished coefficient of RR. Median value of NDI(QRST) standard deviation in normals and ICD patients was 1.62% and 3.64%, respectively.

Conclusion: According to our results high spatial integral map pattern and NDI(QRST) stability observable in healthy subjects is significantly lower in ICD patients. In arrhythmia patients the former linear association between the depolarization and repolarization spatial variability and RR is significantly diminished, residual term increased.