AI algorithm helps detect abnormal electrograms to improve AF ablation procedures
A novel expertise-based machine learning software algorithm consistently enabled operators to perform ablation on regions exhibiting abnormal electrograms in patients with atrial fibrillation, researchers reported.
“Often, you don’t have the answer of the best approach” to an AF ablation, “particularly for the substrate,” Jean-Paul Albenque, MD, cardiac electrophysiologist at Clinique Pasteur in Toulouse, France, told Healio. “The new artificial intelligence software gives the possibility to understand more wisely that atrial fibrillation ablation does not stop with just pulmonary vein isolation.”
Source: Adobe Stock
Albenque and colleagues conducted a prospective, multicenter, nonrandomized proof-of-concept study to determine whether the algorithm (VX1, Volta Medical) could generate dispersion maps and promote consistent ablation outcomes.
“Multiple groups have reported on the usefulness of ablating in atrial regions exhibiting abnormal electrograms during AF,” Albenque and colleagues wrote. “Still, previous studies have suggested that ablation outcomes are highly operator- and center-dependent. ... We compared the acute and long-term outcomes after ablation in regions exhibiting dispersion between a primary and satellite centers. We also compared outcomes to a control group in which dispersion-guided ablation was performed visually by trained operators.”
The study included 85 patients (mean age, 70 years; 67% men) with persistent AF, of whom 29% had long-standing persistent AF.
The rate of AF termination was 92% in the primary center and 83% in the satellite centers (P = .31), according to the researchers.
The rate of freedom from AF with or without antiarrhythmic drugs was 86% after one procedure and 89% after a mean of 1.3 procedures per patient (P = .4), the researchers found.
The rate of freedom from any atrial arrhythmia with or without antiarrhythmic drugs was 54% after one procedure and 73% after a mean of 1.3 procedures per patient (P < .001), according to the researchers.
Outcomes did not differ between the primary center and the satellite centers, nor between the patients who had procedures guided by the algorithm and the control group.
“Procedures using this technology can be performed by different physicians, and most interestingly, it did not appear to have a long learning curve,” Albenque told Healio. “Maybe this technology can be used to improve the results of AF ablation. And it is very simple, using only a computer, and can be employed in any [electrophysiology] lab.”
Volta Medical has begun enrolling patients in the TAILORED-AF randomized trial, which will compare ablation guided by the algorithm with a standard anatomical ablation approach.
For more information:
Jean-Paul Albenque, MD, can be reached at Clinique Pasteur, 31076 Toulouse, France.
Collapse