Catheter ablation for atrial fibrillation (AF) is an effective rhythm control method that shows superior rhythm outcome than antiarrhythmic drug (AAD) treatment in drug-resistant AF. However, AF catheter ablation still has a substantial recurrence rate. The current AAD use guidelines for AF management focus on the safety of drug use. However, if the AAD efficacy evaluation system using computer modeling reflecting the individual anatomy, electrophysiology, and histological characteristics of patients is practical, it will help to select a more effective AAD type or dose. The purpose of this study is to conduct a prospective randomized clinical study on the efficacy and safety of computer modeling for optimal AAD selection in patients with recurrent AF after catheter ablation. The investigator will evaluate the efficacy of AAD simulations by comparing virtual AAD effect guided therapy and empirical AAD use in patients with recurrent AF after AF catheterization. The investigator will test the virtual AAD effects in the computer simulations integrated by cardiac images and 3D electrophysiological maps obtained during de novo AF ablation. The investigator will compare the effects of the most potent AAD selected by virtual AAD simulation with those of empirical AAD.