Signal Processing And Clinical Application Of Surface ECG AF

BackgroundAtrial fibrillation (AF) is one of the most common clinical arrhythmia. Radiofrequency catheter ablation plays an important role in the treatment of AF, but the recurrence reaches up to20-40percent. Although there’re several clinical variables related to the recurrence, we cann’t predict it precisely. As one of the quantitative indicators of atrial remodeling, atrial fibrilatory frequency(AFF) can be used to predict ablation outcome and help with patient selection for catheter ablation.ObjectivesGet AFF through signal processing of surface ECG of AF patients based on Independent Component Analysis (ICA). Investigate the relations between surface ECG AFF and clinical variables. Evaluate the role of surface ECG AFF to predict clinical outcomes of catheter ablation in patients with atrial fibrillation.MethodsRetrospective study patients with atrial fibrillation who had taken radiofrequency catheter ablation at the Beijing Union Medical College Hospital Department of Cardiology, from March2012to March2013. The clinical information of these patients, including patient history, ECG data and echocardiography date, were collected. The AFF of surface ECG and endocardial electrograms before radiofrequency catheter ablation were assessed respectively and analyzed with respect to clinical variables and catheter ablation outcome.ResultsCatheter ablation was performed in36patients (27men, aging60±10years), including10(27.8%) paroxysmal atrial fibrillation and26(72.2%) persistent atrial fibrillation. The surface ECG AFF is6.36±1.17Hz (4-8.4Hz).23patients accomplished6-months follow up(18men, aging58 ±10years), including9(39.1%) paroxysmal atrial fibrillation and14(60.9%)persistent atrial fibrillation.7(30.4%) patients had AF recurrence. The AFF of surface ECG and endocardial electrograms are highly linearly related (r=0.953, P<0.001). The surface ECG AFF was related to age (r=-0.459,P=0.006), type of AF(P=0.002), LVDD(r=-0.345, P=0.042), hypertension (p=0.027),. According to ROC curve, AFF was divided into2groups by6.45Hz.It could be used to predict recurrence6months after catheter ablation, with sensitivity85.7%and specificity68.7%.ConclusionICA is a reliable method in signal processing of atrial fibrillation. AFF is related to clinical features, such as age, gender and type of AF. AFF is related to recurrence after catheter ablation. The higher is AFF, the higher is recurrence rate. It could help with patient selection for catheter ablation.