Intracavitary signal analysis for atrial fibrillation prediction

Signal processing techniques have contributed to a new understanding of the Electrocardiogram (ECG) and its dynamic properties in a significant way. ECG signal analysis is necessary to study cardiac arrhythmias during both diagnosis and treatment. The most common type of cardiac arrhythmia is the atrial fibrillation (AF). Patients are studied by performing electrophysiological procedures in the electrophysiology room. One of the risk for the procedure is the possibility of being able to predict AFs during electrophysiological procedure. This work presents a technique for Atrial Fibrillation prediction. The main innovation of this contribution consists in the analysis of intracavitary signals of a specific disease and in the identification of a new possible predictor. The algorithm has been tested off-line for a limited (yet relevant) number of patients and it may represent a direction for the development of a reliable on line support for electro-physiological procedures.