Abstract Background Little is known about radiofrequency (RF) lesion volume (rLV). Most of animal or ex-vivo studies quantify the size of RF lesions size using width and/or depth, while electrophysiologists use manufacturers’ indirect indices (such as Ablation Index or Lesion Size Index). Purpose We aimed to evaluate rLV by high resolution 9-Tesla magnetic resonance imaging (MRI), in ex-vivo swine model. We studied the variability of rLV and extra-cardiac effects in two groups of swine, one treated with low energy (LE)-long application RF (20W-50 sec) and another treated with high energy (HE)-short application RF (50 W-20 sec). Methods This was a prospective observational study on 18 piglets, that were anesthetized, mechanically ventilated, and divided into LE (n = 9) and HE (n = 9) groups (Figure 1). An irrigated contact force catheter was used to perform RF applications in 4 largely separated spots: 2 on right atrium and 2 on left atrium wall. RF parameters were monitored during RF. Each animal was sacrificed after 5-7 days, and RF lesion samples were excised and acquired by 9 MRI. Results 70 RF applications were performed (36 in LE group and 34 in HE group). MRI was performed on 26 samples of LE group and 28 of HE group (Figure 2). The median RF lesion volume was 89.3 mm3 (IQR 47.9-175.9). Regarding lesion volume, the HE groups showed 35% higher volume than LE group (mean lesion volume 86.2 mm3 vs 116.3 mm3, p = 0.119). HE group also showed more complications (0 vs 7, p=0.02) and more extracardiac lesion (14 vs 18, p=0.613). Conclusions There was a wide and unpredictable variability of RF lesion on atrial wall, even under controlled experimental conditions, which could explain arrhythmias recurrences. The variability seems greatest, although not statistically significant, during HE applications. The dose/effect relationship needs to be carefully studied for treatments of cardiac arrhythmias for both RF and other forms of energy available today.Study flow-chartComparison between groups