Measuring spectral organization in atrial fibrillation

Atrial fibrillation (AF) is a common, clinically significant arrhythmia. When it becomes persistent and drug-refractory, catheter ablation is the preferred therapy. This paper proposes a new method for assessing the level of organization of the electrical activity in the atria during atrial fibrillation, which may enhance the long-term success of ablation therapy and identify those at risk of AF recurrence after therapy. The method uses independent component analysis (ICA) to extract the global atrial activity from the body surface 12-lead electrocardiogram (ECG). The extracted atrial signal is subjected to bandpass filtering and spectral analysis. The resulting spectrum is assessed both for the power in its dominant frequency, and for the power associated with the harmonics of the dominant frequency. The ratio of harmonic power to dominant frequency power is assigned as an index of organization. The results show that the proposed method is predictive of AF recurrence after ablation therapy. The proposed method may provide a more objective way to assess when to terminate ablation, as well as help clinicians determine how to treat post-ablation patients.