A non-invasive method to predict electrical cardioversion outcome of persistent atrial fibrillation

Atrial fibrillation (AF) is the most common cardiac arrhythmia with episodes that may terminate spontaneously in the first stages of the disease. On the other hand, when the arrhythmia is not self-terminating, normal sinus rhythm (NSR) restoration use to be required to reduce the risk of stroke and improve cardiac output. Electrical cardioversion (ECV) is the most effective alternative to revert AF back to sinus rhythm. However, because of its collateral effects and the high risk of AF recurrence, it is clinically important to predict NSR maintenance after ECV before it is attempted. This work presents a non-invasive method able to predict the ECV outcome of persistent AF. In this respect, the atrial activity (AA) organization degree has been computed, both in time and wavelet domains, using a non-linear regularity index, such as sample entropy (SampEn). The main hypothesis considers that AF recurrence can be greater in those patients who present a more disorganized AA. Considering only the time-domain organization analysis, 90% (19 out of 21) sensitivity and 79% (11 out of 14) specificity was obtained, whereas, with only the wavelet-domain organization analysis, 81% (17 out of 21) sensitivity and 86% (12 out of 14) specificity was reported. By combining suitably both organization strategies, 95% (20 out of 21) sensitivity and 93% (13 out of 14) specificity was obtained and the ECV outcome in 33 out of 35 patients (94%) was correctly predicted. These results show that the proposed AA organization schemes and their suitable combination are promising candidates for predicting successful cardioversion and NSR maintenance following ECV in persistent AF patients. Nevertheless, further studies employing larger ECV databases are required to provide confidence and reliability to these methods.