Exploration of Supraventricular Conduction with respect to Atrial Fibrillation. Methodological Aspects on Selected Techniques

Detta är en avhandling från Department of Clinical Sciences, Lund University

Sammanfattning: Atrial fibrillation is the most common cardiac arrhythmia that demands medical attention. Computer-based detailed analysis of cardiac electrical signals from healthy subjects and patients with a history of atrial fibrillation has become an important tool in the further clarification of this disease. The aim of this study was to explore supraventricular conduction using the following selected techniques, to present some results, and to clarify methodological aspects.

Heart rate-stratified RR interval histograms were used to non-invasively assess the characteristics of the atrioventricular node during atrial fibrillation. An RR interval distribution with two distinct peaks was seen in most recordings, believed to represent the dual pathway physiology of the atrioventricular node.

Atrial activation patterns were analyzed using signal-averaged P waves, a technique that can extract repetitive signals from uncorrelated noise. Using several different templates, a variability of P wave morphology was identified.

In a modeling experiment, a mathematical algorithm was used to automatically classify different P wave morphologies. Different sets of parameters were tested, with the best results seen when using a decimated P wave morphology as input.

A mathematical transform was used to derive vectorcardiograms from recorded 12-lead electrocardiograms. A comparison was made between P waves extracted from derived and recorded vectorcardiograms, illustrating that parameters used to describe morphology were preserved on a group basis, but with less accurate results in the individual case.

Left atrial activation was studied using electrograms recorded from a coronary sinus catheter. Correlation function analysis was used to find similarity and time delay between pairs of electrograms. An organized and reproducible atrial conduction pattern was seen in signals recorded during sinus rhythm as well as during atrial fibrillation.

The selected methods for studying atrial fibrillation have each contributed with genuine information concerning the mechanism of this common arrhythmia. The present methodological evaluation further reinforces these results.