A pilot experience on 5G telerobotic long-distance magnetic navigation-guided ablation for cardiac arrhythmias

Since Willem Einthoven’s electrocardiogram (ECG) breakthrough, translational and computational innovations have transformed cardiac arrhythmia management.¹ Today, the combination of telemedicine and robotic technology is promising, with demand increasing for its broad array of clinical applications.² In the past two decades, robotic magnetic navigation (RMN)-guided ablation has become an accepted method of catheter ablation for cardiac tachyarrhythmias with over 150 000 procedures performed.³ Robotic magnetic navigation technology has achieved short-distance (between operating room and control room) wired remote mapping and ablation,⁴ which serves as proof of concept to perform extended distance tele-RMN-guided procedures. Thus, the objective of this pilot study was to explore the feasibility of tele-RMN-guided catheter ablation with 5G technology in patients with cardiac arrhythmias, over long geographical distances.

From January 2023 to January 2024, 13 patients with paroxysmal supraventricular tachycardia (PSVT, n = 5), atrial tachycardia (n = 1), atrial fibrillation (AF, n = 5), and frequent premature ventricular contraction/ventricular tachycardia (PVC/VT, n = 2) were enrolled from Ruijin Hospital in China and Meshalkin National Medical Research Center of Russia. The procedures for AF ablation were performed according to widely used standardized protocols.⁵

Real-time RMN operating interface, fluoroscopic images, intracardiac electrograms, and 3D electroanatomical mapping images were transmitted via 5G technology to remote operator’s monitor (Figure 1I). Zoom transmission system was used to communicate between staff at local and remote sites. In this study, backup wireless connections were prepared to prevent internet disruption and a local operator was allowed to take control if connectivity failed. All patient data were isolated offline post-procedure. All of the anonymized RMN procedure data was preserved by the terminal equipment. The local electrophysiology (EP) staff oversaw the recording system. A 3.5 mm irrigated-tip Navistar Thermocool catheter (NaviStar™ RMT ThermoCool™, Biosense Webster Inc., Irvine, CA) was used for three-dimensional electroanatomic mapping and ablation guided by the CARTO and RMN systems. During the procedure, the remote operator was able to navigate both systems and the catheter with his mouse and keyboard. The local operator was sitting before the computer’s screen in the control room. Radiofrequency (RF) ablation application was controlled by the local operator at the instruction of the remote operator. In this study, the remote operator could neither control fluoroscopy nor start/stop a RF application.

Figure 1.

5G-based telerobotic magnetic navigation system and real case of telerobotic magnetic navigation-guided PSVT ablation. Panel I shows that real-time massages from mapping system, X-ray images, video communication system, and ECG in local control room are sent to the remote operator under 5G internet. The remote operator navigates catheter movement via stereotaxis proprietary remote-control technology. Panel II shows that two magnets locate on either side of the patient in the operating room of Shanghai Ruijin Hospital. Panel III is a remote operator’s monitor on which Carto 3 electroanatomical mapping images (A), real-time (B) and reference (C) intracardiac electrograms, and RMN operating interface (D) were integrated. The reference intracardiac electrograms (III-C) indicated an AVNRT case. Panel IV shows that a local operator (IV-A) in the control room is manipulating the Carto 3 system combined with RMN interface (IV-B). Panel V shows that a remote operator (V-A) in the meeting room of 7th Society for Cardiac Robotic Navigation located in Washington, D.C. is remotely manipulating the system and communicating with the local operator using 5G technology.

Local and remote operators collaboratively determined electrophysiological mechanisms, ablation strategies, and RF application sites/parameters (output/duration) via an audiovisual communication system. In cases of disagreement, final procedural decisions rested with the local operator while acknowledging the remote operator’s suggestion. If the remote operator’s control intermittently failed due to lost connection or significant latency, the local operator took over RMN control of the operation immediately. In this study, only the local operator was able to start/stop a RF application. If the local operator encountered any situation warranting stopping the ablation, he could stop delivery of RF energy at will and without the remote operator’s command.

The success of the procedure was defined as follows: (i) acute ablation success, with endpoints defined as: for atrioventricular nodal re-entrant tachycardia (AVNRT), AT and VT procedures, non-inducibility of arrhythmia with isoproterenol; For PVC procedures, elimination of PVCs after administering isoproterenol; for AF procedures, circumferential pulmonary vein isolation; (ii) successful tele-RMN ablation by a remote operator without conversion to a local operator; and (iii) no tele-RMN procedure-related complications.

Procedure parameters and the total latency time were analysed. Patients were scheduled for routine follow-up at 1, 3, and 6 months after ablation.

Detailed patient clinical characteristics and procedural parameters are provided in Table 1.

Table 1.

Baseline characteristics and procedural parameters of patients who underwent 5G tele-RMN-guided ablation (n = 13)

Clinical characteristics
Age (y)	55.5 ± 17.8
Male (%)	9 (69.2)
Hypertension (%)	5 (38.5)
Dyslipidaemia (%)	8 (61.5)
Diabetes (%)	3 (23.1)
Pro-BNP (pg/mL)	60.2 (45.1, 220.0)
LVEF (%)	65.0 (64.0, 71.0)
Left atrial diameter (mm)	37.1 ± 4.9
Cardiac arrhythmias
AF (%)	5 (38.5)
Paroxysmal (%)	4 (30.8)
Persistent (%)	1 (7.7)
PSVT (%)	5 (38.5)
Slow–fast AVNRT (%)	5 (38.5)
Atrial tachycardia (%)	1 (7.7)
PVC/VT (%)	2 (15.4)
Left posterior fascicle (%)	1 (7.7)
Right ventricular outflow tract (%)	1 (7.7)
Procedural parameters
Procedure time (min)	64.5 ± 16.0
RF time (min)	6.1 ± 3.3
Energy application frequency	8.0 (3.0, 52.0)
Fluoroscopic time (min)	3.3 ± 1.5
X-ray exposure dose (mGy)	34.8 (16.3, 49.5)

BNP, brain natriuretic peptide; LVEF, left ventricular ejection fraction; AF, atrial fibrillation; PSVT, paroxysmal supraventricular tachycardia; AVNRT, atrioventricular nodal re-entrant tachycardia; PVC, premature ventricular contraction; VT, ventricular tachycardia.

Among the 13 enrolled patients, two (one AF, one PSVT) underwent procedures between Ruijin Hospital (China) and Meshalkin National Medical Research Center of Russia (6814 km apart), both equipped with RMN systems. The remaining 11 subjects’ procedures were performed between Ruijin Hospital (local site) and remote conference rooms where a remote operator controlled the computer (maximum distance: 11 977 km).

This study achieved 100% acute success without any tele-RMN related or procedural complications. Mean procedure times: AF 72.4 ± 4.5 min, PSVT + AT 57.5 ± 21.2 min, PVC/VT 65.0 ± 11.3 min. Fluoroscopic time averaged 3.3 min (minimum 1.6 min in PVC/VT cases). Intraoperative image and audio signal transmission were timely and smooth, with an average total latency of 63.8 ± 23.9 ms (including both round-trip delay and image processing delay). As a representative case, a PSVT patient in Shanghai was successfully ablated by the remote operator in Washington, D.C., USA (Figure 1). No clinical arrhythmia recurrence occurred during the 6-month follow-up period.

Current telerobotic systems remain predominantly confined to short-distance applications, with their long-range operational potential remaining underdeveloped.⁶ Robotic magnetic navigation systems previously enabled short-distance (operating room and control room) wired remote ablation in EP lab. By integrating real-time information (operational interfaces, intracardiac electrograms, 3D mapping, and fluoroscopy images) into a 5G-connected remote’s computer, transcontinental telerobotic cardiac mapping and ablation were successfully achieved (Figure 1). This study developed a 5G-based RMN-guided tele-ablation platform that preserved standard and expected operational efficiency across various arrhythmias without significant procedural latency.

This pilot study adopted a conservative operator selection criterion, with local/remote physicians each having over 1000 RMN procedures’ experience. Future applications may enable remote experts to guide strategy or directly assist if local junior operators face challenges in arrhythmia diagnosis or ablation. This 5G tele-RMN-guided ablation platform circumvents geographic barriers exacerbated by regional variance in physician experience, as well as global health crises (e.g. COVID-19 pandemic travel restrictions), enabling optimal arrhythmia treatment with limited human resources.

Study limitations include: first, further prospective clinical trials with a larger volume are needed to more definitively verify the safety and efficacy of 5G-based RMN telerobotic ablation. Secondly, this is a multicentre experience from hospitals with high RMN procedure volumes. Thus, the practice standards and protocols for operators with differing experience levels, especially as related to the management of procedure-related complications, are required.

In the future, a multi-layered approach co-ordinating all technology, policy, and training aspects is essential for standardizing secure remote procedures using RMN. First, it is necessary to utilize dedicated 5G/6G networks or healthcare virtual private networks to isolate traffic from public internet risks. Secondly, zero-trust, real-time monitoring, and strict compliance should be enforced.

In conclusion, 5G-based telerobotic ablation using RMN for cardiac arrhythmias, as a novel telemedicine-based therapy choice, is feasible with efficacy and safety.

Funding

Shanghai Municipal Science and Technology Commission 24SF1906400.

Data availability

Clinical trial data are protected under patient confidentiality agreements. All other relevant data are available from the corresponding author upon reasonable request.