A Novel Workflow Combining Artificial Intelligence-Twelve Lead Electrocardiographic Analysis and Real-Time Mapping for Hemodynamically Unstable Ventricular Tachycardia.

BACKGROUND

Activation mapping of scar-mediated ventricular tachycardias (VT) are often limited by hemodynamic instability. Substrate mapping and 12-lead ECG localization of VT exit site is often limited by variable scar complexity. Emphasis on other mapping strategies in sinus rhythm would improve catheter ablation success in such cases.

OBJECTIVE

The aim was to investigate the level of agreement between artificial intelligence (AI) based 12-lead ECG localization of VT exit site and mapping utilizing pace-mapping and mapping of channels of slow conduction within the scar tissue of hemodynamically unstable VT.

METHODS

This was a single-center proof-of-concept study that included patients who underwent catheter ablation procedure of hemodynamically unstable scar-mediated VT. The performance of AI-assisted ECG analysis of VT exit site (Vektor Medical, San Diego, CA) was compared with sites of successful ablation based on substrate mapping in sinus rhythm.

RESULTS

A total of nine hemodynamically unstable VT rhythms were induced in four patients. In the seven VTs were AI-assisted ECG analysis was used; there was a 100% level of agreement with the site of successful ablation based on substrate mapping. Ablation targeting those sites resulted in non-induction of all induced and mapped VTs. None of the four patients had device therapy for recurrent VT or all-cause mortality at 6 months of follow-up.

CONCLUSION

A multi-strategic approach utilizing digitalized analysis of 12-lead ECG of VT exit site, pace-mapping and mapping of channels of slow conduction in scar tissue, has the potential to enhance successful catheter ablation of hemodynamically unstable VT.