Generation of realistic atrial to atrial interval series during atrial fibrillation.

The aim of the this study is to describe a methodological architecture for the generation of realistic atrial to atrial activation intervals (AA) during atrial fibrillation (AF), which can be used to investigate the role of the fibrillatory process in the ventricular response during AF. In this study, a methodology for the generation of AA interval series with a desired probability density function and autocorrelation function is presented. The methodology was evaluated on 2000 AA interval series from 20 endocardial recordings. The results showed that synthetic AA series presented the same statistical characteristics as the real AA series, with a correlation higher than 0.94 (P < 0.01) for all measured statistical parameters. In addition, the role of each statistical characteristic of the AA interval series in the ventricular response during AF is examined using a mathematical model of the atrioventricular node. The statistical characteristics of the AA series influenced the position of more probable RR intervals and the shape of the RR histogram, demonstrating the importance of an accurate characterization and generation of AA interval series during AF. The use of the present methodology may help in understanding the role of the atrial fibrillatory process in the ventricular response during AF.