This is an electrogram mapping study that tests the ability of a new signal processing technique called electrogram morphology recurrence mapping to locate drivers in participants with atrial fibrillation. This study will also test the hypothesis that there are non-invasive markers that can predict driver location. Electrogram mapping will be performed in participants undergoing ablation for persistent atrial fibrillation. These participants will have a cardiac MRI and 12-lead ECG prior to the ablation procedure.

Atrial fibrillation (AF) has been described as a new emerging epidemic currently affecting around 1% of the population. Risk of AF increases with age and as the elderly continue to comprise a larger percentage of the population in the United States, the health burden of AF continues to increase. Some estimates project nearly 16 million cases of AF in the US alone by 2050. The economic burden of AF has been estimated to be between $6-26 billion. AF is a significant risk factor for stroke if left untreated as well as mortality. The diagnosis and treatment of AF have therefore become an important aspect of cardiovascular medicine.

Current pharmacological treatments for AF have modest efficacy and suffer from potentially life-threatening side effects. Current ablative and surgical approaches to AF, though somewhat more successful, use an anatomic, 'one-size fits all' strategy (with some minor variations) that does not address the specific mechanisms underlying this complex arrhythmia. The pulmonary veins have been shown to be a common location of AF triggers. Thus, the most common approach for catheter ablation involves the isolation of the pulmonary veins, such as the wide area circumferential pulmonary vein ablation strategy. The lack of consideration of specific sources may be, at least in part, responsible for post- ablation arrhythmias occurring in up to 40% of patients who undergo pulmonary vein ablation. However, while ablation/surgical success has been shown to increase with more extensive ablation/surgery in the atria, the latter comes at the cost of significant "collateral damage" to the atrial myocardium as well as to surrounding structures, with an accompanying increase in the risk of complications and a decrease in atrial function. The time and cost to perform these procedures is also significant. Thus, catheter and surgical ablation strategies tailored to the specific arrhythmia origin of a patient's AF would have a significant impact on the ability to treat this arrhythmia.

The investigator's laboratory has developed a novel electrogram morphology recurrence analysis to observe the nonlinear dynamics of AF activation. This technique to map AF activation is based on the hypothesis that AF electrograms with morphologies that are highly recurrent and that have short cycle lengths represent the locations of stable AF sources. The investigator's preliminary data show that when the highest recurrence percentage or shortest recurrence cycle length occurs in the left atrium in patients undergoing pulmonary vein-based ablation, the patients has a much higher chance to stay in sinus rhythm post-ablation than when the highest recurrence percentage or shortest recurrence cycle length occur in the right atrium. Cycle length of all activations was not a predictor of outcomes. High resolution electrical mapping of AF in the canine atria showed that the shortest recurrence cycle length sites correspond to areas containing rotors. Thus, these preliminary data provide strong evidence that morphology recurrence mapping is able to identify the locations of AF sources.

Inclusion Criteria:

• Male and female, At least 21 years of age.
• Subjects with persistent AF undergoing a first catheter ablation procedure.

Exclusion Criteria:

• Inability to sign consent.
• Patients with a life expectancy <1 year.
• Previous ablation of the atria
• Pregnant women and women that are breast feeding.
• Subjects with contraindication to MRI such as cochlear implants, ocular foreign body (e.g. metal shavings), brain aneurysm clip, implanted neural stimulator, implanted cardiac pacemaker or defibrillator.
• Subjects with a history of kidney problems (GFR < 30m/min) or have had a kidney and/or liver transplant in the past 4 weeks will be excluded from the study or undergo the MRI exam without the use of a contrast agent, per standard MR exclusion criteria.