Atrial fibrillation (AFib) affects more than 6 million Americans, and that number is projected to rise to more than 12 million by 2030. That growing patient population demands more precise and effective approaches to treatment, and physician leaders at Ochsner Health in Louisiana are seeing promising results with a recently introduced AI-driven approach for AFib ablation procedures.
Michael L. Bernard, MD, PhD, is a cardiac electrophysiologist and when attending Heart Rhythm 2024 in Boston, he saw a presentation by the researchers who conducted the Tailored vs. Anatomical Ablation Strategy for Persistent Atrial Fibrillation (TAILORED-AF) randomized controlled trial. The trial results were later published in Nature Medicine.
“It was a large, multi-country trial with more than 350 patients and strong clinical outcomes. Results showed that success rates improved significantly when the AI software was used alongside standard practices, and that really sparked my interest,” said Dr. Bernard. He is the section head of electrophysiology at Ochsner Health, which is part of the AMA Health System Member Program that provides enterprise solutions to equip leadership, physicians and care teams with resources to help drive the future of medicine.
Dr. Bernard was intrigued by how AI-driven software analyzed the complex electrical behavior of AFib in real time to detect patterns that human eyes and other software cannot. The tech can help physicians with real-time analysis of three-dimensional heart maps.
Impressed—and inspired—Dr. Bernard returned to Ochsner Health in New Orleans with a mission. He would champion the investigation and eventual adoption of this technology at John Ochsner Heart & Vascular Institute. The health system’s implementation of AI-powered mapping software reflects a physician-led approach to digital health innovation and builds on previous work with a separate cardiac ablation procedure.
From AI implementation to digital health adoption and EHR usability, the AMA is fighting to make technology work for physicians, ensuring that it is an asset to doctors. That includes recently launching the AMA Center for Digital Health and AI to give physicians a powerful voice in shaping how AI and other digital tools are harnessed to improve the patient and clinician experience.
Health AI oversight, safety and physician feedback
After discussing the researchers’ findings with his colleagues, Dr. Bernard asked the software developers to see the technology in use. In October 2024, he and Sammy Khatib, MD, also a cardiac electrophysiologist and the cardiology chair at Ochsner Health, traveled to Phoenix to observe actual cases.
“Seeing it in real life was an invaluable process,” Dr. Bernard said. “After that, we were on board.” They began the process of talking with their physician colleagues at Ochsner Health about the technology.
Vetting included safety and regulatory reviews by Ochsner Health’s legal, IT and purchasing committees. “It underwent additional layers of scrutiny and governance review processes because of the AI component,” he explained. Strong clinical data was considered. The randomized controlled trial “resulted in very high disease-free rates that have rarely been seen in atrial fibrillation ablation studies.”
As published in Nature Medicine, for patients who had AFib for six months or longer, 88% who got the AI-driven tailored treatment were free of AFib after one procedure. That compares with the 65% AFib-free rate achieved in the conventional procedure arm.
Ochsner Health’s governance structure includes an AI steering committee that reviews new technology for privacy, safety and ethical issues, and each new initiative is given a “clinical champion” who is accountable for outcomes. A 70-member Epic Academy facilitates training and collects physician feedback to ensure that new tools enhance workflows rather than disrupt them.
It helped that the technology fit seamlessly into Ochsner’s existing infrastructure, which includes the largest electrophysiology laboratory space in Louisiana and Mississippi. The AI system functions as a simple software plug-in to existing mapping tools, so integration was unobtrusive. “We just have one additional computer in the room,” Dr. Bernard explained. “There was no major change in our workflow.”
How this health AI tool helps doctors
Physician engagement was essential from the start. A small team of six staff members and two cardiac electrophysiology fellows were charged with maximizing the use of the technology, which included an unexpected benefit: real-time alerts during procedures that direct physicians to zones within the heart that indicate more analysis is needed.
“I’m learning from AI what’s a high-value target versus what’s not,” Dr. Bernard said. “I didn’t expect to learn that from a machine.” The technology has also improved clinical insight.
“I’ve been fighting AFib throughout my whole career,” he said. “The software is identifying sites in cases where the usual techniques don’t work.”
Patients have been receptive too, with most having a favorable response to the introduction of AI in their treatment. “It’s been positive more than it’s been negative,” Dr. Bernard said. “In fact, some people ask for it after they hear about it.”
The AI-driven software “doesn’t replace what we do” as physicians or cardiac electrophysiologists, Dr. Bernard said. “It just helps us interpret better by giving us information that we didn’t have before.”
Learn more about AMA policy on augmented intelligence (AI), a conceptualization of artificial intelligence that focuses on AI’s assistive role, emphasizing that its design enhances human intelligence rather than replaces it.
“Eye-opening” results for health AI
Ochsner Health is among about 20 centers in the U.S. to have adopted this AI-driven software in treatment of AFib and worked to integrate it responsibly into physician workflows.
The Ochsner Health experience “shows that we can use AI products in medicine safely and they can be helpful,” Dr. Bernard said. “It is a collaboration of human hands and AI that delivers good results.”
Early results have been promising, and Dr. Bernard cited one of his earliest success stories.
“One of the first patients had been in AFib for six months,” he said. “He’d undergone multiple failed ablations. We used the AI-guided approach, and he has now been arrhythmia-free for three months. When I thought I didn’t have a chance, this thing really delivered.”
That’s just one case, he acknowledged, “but it’s been eye-opening, and seeing that rhythm normalize in real time was incredibly rewarding.”
Dr. Bernard and his colleagues are studying patients and the outcomes of their AI-assisted AFib ablation procedures. The data will be tracked for short-term and longitudinal results, then used to mirror the tactics of the clinical trial over one or two years.
The most rewarding aspect of the experience for Dr. Bernard is seeing improved outcomes in patients who had few remaining options.
“When you’ve done everything you know to do and you still can’t fix the problem, this gives us another chance,” he said. “It gives patients with AFib hope for a restored quality of life.”
Ochsner Health’s patient data and outcomes may help shape how AI-guided ablation is used nationwide. “If we’re getting good results with the first iteration of this technology, it’s easy to imagine what future iterations are going to do,” Dr. Bernard said. “It really is an exciting time to be part of this field.”
Explore further with the AMA STEPS Forward® toolkit, “Governance for Augmented Intelligence,” which is a comprehensive eight-step guide for health care organizations to establish a governance framework to implement, manage and scale AI solutions.
