Q&A: AI transforming electrophysiology, but ‘preserving the humanism’ necessary
Click Here to Manage Email Alerts
Key takeaways:
- AI has the potential to greatly improve the workflow of cardiac electrophysiologists.
- Clinicians will need to ably navigate AI-based systems without losing their critical thinking skills.
Few areas in medicine have been impacted by AI as much as cardiac electrophysiology. While still in early stages, it is already influencing how electrophysiologists detect disease and predict outcomes.
At HRX 2024, the Heart Rhythm Society’s meeting on digital health advancements, experts in electrophysiology and AI convened for a session on what AI can accomplish in electrophysiology and how it can be appropriately implemented clinically.
One of the session’s panelists, Jagmeet P. Singh, MD, DPhil, FACC, FHRS, professor of medicine at Harvard Medical School, cardiac electrophysiologist and founding director of the Resynchronization and Advanced Cardiac Therapeutics Program at Massachusetts General Hospital Heart Center, a member of the Healio | Cardiology Today Editorial Board and the author of Future Care, a book on AI and other technological advancements in medicine, spoke with Healio about how AI is used in cardiac electrophysiology, what institutions need to do to make sure it is implemented properly and the importance of “preserving the humanism” in an increasingly digital world.
Healio: To what extent is AI currently used in electrophysiology?
Singh: Much of AI in electrophysiology is still investigational. Although its role in predictive analytics toward diagnosing new-onset atrial fibrillation, ventricular tachycardia, sudden death and heart failure is possible, its clinical validation and use in clinical care are still ongoing. The same goes with augmented reality approaches for interventional procedures in electrophysiology — they are being tested and studied.
Healio: How is current AI technology able to help electrophysiologists do their jobs?
Singh: So, currently much of AI being deployed is primarily to support administrative tasks. Ambient technology is being used to generate outpatient clinic notes, helping reduce the administrative time and burn. Generative AI is also being used to help unchain us from some of the mundane tasks like prescription refills and preauthorizations. Most of these are still in pilot phases. In addition to this, we are investigating the role of AI to facilitate responses to our patients’ queries electronically.
Healio: What, if any, concerns are there about how use of AI might evolve in electrophysiology?
Singh: I think the biggest concern is that algorithms being used for clinical care should be unbiased and truly representative of the population they are being used in. Making sure that we are not enhancing disparities, while maintaining privacy and security, is going to be very important. The black box issue and distrust of patients and physicians of AI is also a concern. We need to be transparent and gain the trust of all, as AI technologies and predictive analytics evolve. And then, large language models have a propensity to confabulate and can misinform us. We need to make sure that we are not blindly trusting these outputs, and it may take some time and refinement before they could potentially be used to help us deliver clinical care.
Healio: What are the best ways to implement AI in electrophysiology clinical programs?
Singh: We need to recognize that medicine is evolving. The way we receive and deliver care is going to morph over the coming years, and it is imperative that we make virtual care and AI-facilitated approaches a part of the curriculum. We have to ensure that our next generations become great knowledge navigators, but don’t lose their critical thinking. We have to recognize that we are in the age of intelligence, and we should be adjusting our clinical programs to adapt to that.
Healio: What were some of the other important insights learned from this session at HRX?
Singh: I think we realized the importance of collaboration — and working together to create datasets to solve many of the problems medicine faces. We discussed the importance of predictive analytics and generative AI, but also recognized the gaps that exist. One of the things we spent time discussing is how AI can help us begin to focus on value over volume. Also, preserving the humanism within our clinical practice is of paramount importance. We have to make sure we create a synergy between the digital and human touch as we advance personalized care.
Reference:
- Singh JP, et al. Digital innovation, AI and the future of cardiac electrophysiology. Presented at: HRX 2024; Sept. 5-7, 2024; Atlanta.
For more information:
Jagmeet P. Singh, MD, DPhil, FACC, FHRS, can be reached at jsingh@mgh.harvard.edu.
Collapse
Healio Interviews
Click Here to Manage Email Alerts