A new algorithm for measuring atrial fibrillation organization

Subjective and quantitative measures of atrial fibrillation (AF) organization exist. The authors' introduce a new quantitative method and compare it to 2 previous algorithms. In the authors' method, 2 atrial electrograms are processed with 2 linear adaptive filters to predict 1 electrogram from the other. The mean-squared error (MSE) of the prediction quantifies the non-linearity or disorganization between the electrograms. From atrial epicardial recordings, MSE during sinus rhythm and atrial tachycardias were similar, but showed more organization than AF with different levels of vagal stimulation (VS). All 3 algorithms for quantifying AF organization showed low VS AF to be more organized than high VS AF, with MSE showing the most significant difference compared to the other 2 methods. MSE discriminates between AF and organized rhythms and is sensitive to different degrees of AF organization.