Atrial fibrillation (AF) is the most common arrhythmia in clinics worldwide. Radiofrequency or balloon cryoablation of circumferential pulmonary vein (PV) vestibule is the standard method for the treatment of AF [1]. However, in recent years, radical isolation strategies, such as left atrial roof and isthmus ablation, as well as the fragmentation potential ablation have been widely applied, the damage to the surrounding tissues caused by ablation energy is under intensive focus [2]. Hitherto, the cardiac electrophysiologists have paid enough attention to the injury of esophagus and phrenic nerves, and hence, several studies have explored the anatomical correlation between the left atrium (LA)-PV ablation targets of AF and adjacent esophagus and phrenic nerve [3‐7]. In our previous studies by MSCT, we have showed a close relation of the right phrenic nerves to superior caval vein, right atrium, and right superior PV, and the adjacency of the left phrenic nerves to left atrial appendage and coronary venous tributaries [7]. Recently Felix Bourier et al. [8] reported a feasible method combining the 3D-mapping system integration module and CT-derived anatomy, showing that even the three‐dimensional image integration also can show the right phrenic nerves into fluoroscopic imaging clearly. Anatomically, the ablation targets of AF are adjacent to the lung, airway, and pulmonary artery (PA), such that the ablation energy may damage the bronchi, PAs, and lung parenchyma, resulting in the symptoms of cough, dyspnea, hemoptysis, and even death in the patients [5, 9]. Animal studies have shown that ablation energy may cause acute bronchial inflammation and mucosal injury [10]. Also, rare clinical cases of atrial-bronchial fistula caused by the ablation of AF are reported. [9] The clinical trial of sustained treatment of paroxysmal atrial fibrillation (STOP-AF) [11] showed that the incidence of persistent cough caused by cryoablation was 17%, and its subsequent post-approval clinical trial [12] showed that the incidence of hemoptysis caused by ablation was 1%. Verma et al. reported a patient with a formation of ice crystals in the left main bronchus after 50 s of ablation of the left superior PV [13]. Furthermore, the PV ablation resulted in ice crystal formation and internal hemorrhage in the main bronchus was reported in seven patients [14].

The high tissue resolution of multi-slice spiral computed tomography (MSCT) angiography can visually and accurately display the anatomy of LA and its surrounding tissues. The powerful volume rendering (VR) and multi-planar reformation (MPR) technologies of MSCT can not only display the spatial anatomical correlation between the heart and adjacent structures, but also achieve accurate quantitative measurements. Moreover, MSCT imaging is simple, quick, non-invasive, and importantly, can reconstruct and accurately measure the images at any angles in vivo, thereby avoiding tissue shrinkage, distortion, and vascular volume changes caused by the immersion of the formalin solution in autopsy. It is no doubt that MSCT is more accurate to measure the spatial anatomical distances between two adjacent structures of the heart as compared to the autopsies. Recently, we have successfully achieved the non-invasive phrenic nerve imaging in patients using MSCT angiography [7], but few studies had described the anatomical adjacency between the ablation targets of AF and the bronchi as well as the PAs. In this study, MSCT angiography was used to evaluate the spatial anatomical relation between ablation targets of AF, such as four PVs and ablation lines on the LA roof, and the adjacent bronchi and PAs in vivo, which might provide a promising guidance for cardiac electrophysiologists to reduce the complications of AF ablation.

In recent years, with the rapid development of radiofrequency catheter and balloon cryoablation techniques for AF, the damage caused by the ablation energy to the tissues surrounding the ablation targets has attracted attention and has been investigated in-depth with respect to cardiac morphology and anatomy. Previous studies showed that tissue structures < 5 mm away from ablation targets had a high risk of damage [15]. Anatomically, the AF ablation targets, such as PV orifice and LA roof, are closely adjacent to the trachea/bronchial tree and PAs and their branches. Therefore, the ablation energy might damage the adjacent bronchi and PAs, resulting in cough, dyspnea, hemoptysis, and other symptoms [5, 9]. Verma et al. [14] observed in 7 patients with the intraoperative real-time bronchoscopy that ice crystals and internal hemorrhage were formed within the left primary bronchi during the cryoablation of the left superior PVs. Kuck et al. [16] reported that the incidence of chest pain, cough, and hemoptysis in AF cryoablation and radiofrequency ablation was 2.9% and 1.9%, 0.8% and 0.5%, 0.3% and 0.5%, respectively.

Recently, the temporal and spatial resolution of MSCT has been greatly improved, which in turn, displayed the structures of the heart, mediastinum, and blood vessels more clearly and accurately. The unique 3D reconstruction technique of MSCT displayed the anatomy of the heart and adjacent tissues from variable angles and directions, thereby providing a powerful tool for exploring the anatomy correlation of the heart for arrhythmia ablation. The unique MPR post-processing technique of MSCT allows for an accurate measurement of the shortest spatial distance between two adjacent tissues in vivo. In the present study, we used the MPR post-processing technique of MSCT to quantitatively evaluate the anatomical adjacency of the AF ablation targets, such as four PV orifices and the LA roof, to the surrounding airway and the PA system, in order to provide guidance for cardiac electrophysiologists before ablation to reduce the complications of AF intra-operatively and post-operatively.

MSCT angiography is a feasible method to assess the anatomical correlation between the ablation targets of AF and adjacent airway and PA systems. The PAs had a closer anatomical relationship with the LA roof and PV orifices than the bronchi. Compared with the inferior PVs, the superior PVs, especially the right superior PVs, were more closely related to the bronchi and PAs. The ablation of the middle and posterior position of the LA roof might have higher possibilities of risking the adjacent bronchi. Coronary MSCT angiography is a valuable tool to evaluate the anatomical adjacency of the ablation targets, which provides reference for cardiac electro-physiologists to reduce the AF ablation complications.