The purpose of this study is to evaluate our previously proposed model-based, non-invasive approach to assess atrioventricular (AV) node refractory periods during atrial fibrillation (AF) by comparing its estimates to the invasively assessed AV node refractory periods. Patients referred to the hospital for pulmonary vein isolation (PVI), being in AF upon arrival, are included in the study. Using our approach, we estimate the AV node refractory periods from the ECG before and during the PVI procedure. The refractory periods is estimated using both single and dual pathway models; Bayes Information Criterion is employed to select the most appropriate model. Following conversion to sinus rhythm, a S1S2 atrial pacing protocol is applied to invasively assess AV node refractory periods. Preliminary results suggest that the method for non-invasive assessment of AV node refractoriness is accurate, however, more data is needed to better establish performance.
IntroductionDuring atrial fibrillation (AF), the atrioventricular (AV) node is continuously bombarded with atrial electrical impulses. The AV node is the natural barrier which restricts conduction of atrial impulses into the His-Purkinje system and is therefore very important during AF since it to some extent protects the ventricles of the heart from the extremely fast atrial rate. Electrophysiologic factors such as refractoriness of the AV node and concealed conduction influence the AV nodal blocking of impulses. Even though these properties play a prominent role in ventricular rate control, they are not routinely evaluated in clinical practice. The existence of dual AV nodal pathways, i.e., a fast pathway (FP) with a longer refractory period and a slow pathway (SP) with shorter refractory period [1] is a further complicating factor. During normal sinus rhythm atrial impulses are generally conducted though the FP, but during AF conduction may occur through the SP due to its shorter refractory period.We have previously proposed a model-based approach for joint analysis of atrial and ventricular activity in the ECG during AF [2-4]. The method is based on a functional model of the AV node where the model parameters reflects SP and FP refractoriness. The method has been used to analyze 12-lead resting ECG data from 14 patients in a phase II, open-label, sequential-group, doseescalation trial of tecadenoson administered intra venously alone and in combination with esmolol [5] as well as longterm ECG data from 60 patients in a randomized, crossover study comparing the efficacy of four different rate control drugs [6]. The results showed that changes in AV node refractoriness, as reflected in the model parameter estimates, were in agreement with previous clinical findings. This suggests that the model parameters estimated from ECG during AF may be used as a surrogate for the AV node refractory period invasively assessed through a pacing protocol [7].The purpose of the present study is to evaluate our model-based, non-invasive approach to assess AV node refractory p...