Computational mapping identifies localized mechanisms for ablation of atrial fibrillation.

Atrial fibrillation (AF) is the most common heart rhythm disorder in the Western world and a common cause of hospitalization and death. Pharmacologic and non-pharmacologic therapies have met with limited success, in part due to an incomplete understanding of the underlying mechanisms for AF. AF is traditionally characterized by spatiotemporally disorganized electrical activation and, although initiating triggers for AF are described, it is unclear whether AF is sustained by spatially meandering continuous excitation (re-entrant waves), localized electrical sources within the atria, or some other mechanism. This has limited therapeutic options for this condition. Here we show that human AF is predominantly caused by a small number (1.8 ± 0.9) of localized re-entrant waves or repetitive focal beats, that remain stable with limited spatial migration over prolonged periods of time. Radiofrequency ablation that selectively targeted the sites of these sources was able to immediately terminate fibrillation and eliminate the arrhythmia with high success. Our results show that human AF, despite apparent spatiotemporal disorganization, is often perpetuated by a few spatially-constrained and temporally conserved sources whose targeted ablation can eliminate this complex rhythm disorder.