Our study has some limitations. The study was conducted in a single center and in a limited number of patients. For an accurate assessment of LAPW thickness adjacent to the esophagus, the image plane was required to be perpendicular toward the posterior LA wall. We tried to obtain such an image, but there might be a possibility of overestimation because of an oblique projection. We used only one power setting (25W) in AI guide ablation on the esophagus region and did not evaluate the effect of RF power on the outcomes. No endoscopic examination was done to detect the esophageal lesion, and asymptomatic lesion could not be ruled out. Because AF recurrence was assessed by the symptoms and Holter ECG monitoring intermittently, asymptomatic AF might have been underestimated.