Characterisation of atrial flutter variants based on the analysis of spatial vectorcardiographic trajectory from standard ECG

Abstract

After atrial fibrillation, atrial flutter is the most common atrial tachyarrhythmia. Its diagnosis relies on the twelve lead electrocadriogram analysis of the distinctive waves in several leads. Nonetheless, the accurate identification of the type of atrial flutter still requires an invasive procedure. The maneuver for healing atrial flutter consists on ablating a section of the anatomy of the atria, to stop the macroreentrant circuit to keep happening, allowing the signal to travel to the ventricles in stead of staying at the atria. The region to ablate directly depends on the place at which the macroreentrant circuit is located, which at the same time depends on the type of atrial flutter. Being able to noninvasively detect the atrial flutter variant would produce a great advantage when healing this illness. The hypothesis stated at this dissertation is based on the slow conduction regions as the key factor to distinguish the atrial flutter class. This and unveiling further relations between cardiac illnesses and their signal’s alter ego are the purpose of this research project. With such aim, different methods are developed based on the vectorcardiographic representation of electrocardiograms from patients suffering from different atrial flutter types. These methods consist on the characterisation of vectorcardiographic signals from different standpoints. Besides, a mathematical model is implemented to create a large database with synthetic vectorcardiographic signals allowing to test the validity of the utilised methods. The results prove the importance of slow regions in the vectorcardiographic representation of the patient’s signals to characterise the atrial flutter type non-invasively. Furthermore, the analysis of the outcome of the different methods reveal a wide variety of features relating characteristics of the vectorcardiographic signal to the anatomy and physiology of this cardiac disease. Hence, not only results supporting the hypothesis were successful (taking into account some limitations), but also a variegated assortment of results unmasked remarkable relations among the vectorcardiographic signal and the characteristics of the atrial flutter disease.