Researchers use an AI-guided screening approach to detect previously unrecognized atrial fibrillation

Mayo Clinic researchers used artificial intelligence (AI) to assess patients’ electrocardiograms (ECGs) as part of a targeted screening strategy for atrial fibrillation, a common heart rhythm disorder. Atrial fibrillation is an irregular heartbeat that can lead to blood clots that can travel to the brain and cause a stroke, but it is vastly underdiagnosed. In the decentralized, digital study, the AI ​​identified new cases of atrial fibrillation that would not have come to clinical attention during routine care.

Previous research had already developed an AI algorithm to identify patients with a high probability of previously unknown atrial fibrillation. Algorithm for detecting atrial fibrillation in normal sinus rhythm from ECG is licensed from Anumana Inc., an AI-focused health technology company, by inference and Mayo Clinic.

We believe screening for atrial fibrillation has great potential, but currently the yield is too low and the cost is too high to make widespread screening a reality. This study demonstrates that an AI-ECG algorithm can help target screening to patients most likely to benefit.”

Peter Noseworthy, MD, cardiac electrophysiologist, Mayo Clinic and lead study author

About the study

The study recruited 1,003 patients for ongoing monitoring and used an additional 1,003 usual care patients as real-world controls. The findings, published in The Lancetshowed that AI can indeed identify a subset of high-risk patients who would benefit more from additional intensive cardiac monitoring to detect atrial fibrillation, supporting an AI-guided targeted screening strategy.

ECGs are commonly done for a variety of diagnoses, but because atrial fibrillation can be transient, the chances of catching an episode on a single 10-second ECG trace are low. Patients can undergo continuous or intermittent cardiac monitoring approaches that have higher detection rates, but they require too many resources to apply to everyone and can be time-consuming and costly for patients.

This is where the AI-guided ECG can help. The AI ​​algorithm can identify patients who, although they have a normal rhythm on the day of the ECG, may be at increased risk for undetected episodes of atrial fibrillation at other times. These patients may then undergo further monitoring to confirm the diagnosis.

“Traditional screening programs select patients based on their age (65 or older) or the presence of conditions such as high blood pressure. These approaches make sense because advanced age is one of the risk factors most important causes of atrial fibrillation. However, it is not possible to repeatedly perform intensive cardiac monitoring in more than 50 million elderly adults across the country,” says Xiaoxi Yao, Ph.D., researcher in health outcomes at the Department of Cardiovascular Medicine and the Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery Dr. Yao is the study’s lead author.

“The study shows that an AI algorithm can select a subgroup of older adults who might benefit more from intensive monitoring. If this new strategy is widely implemented, it could reduce uncontrolled atrial fibrillation. diagnosed and prevented strokes and death in millions of patients around the world,” says Dr Yao.

Next steps

The next stage of this research will be a multicenter hybrid trial focusing on the effectiveness of AI-ECG workflow implementation in various clinical settings and patient populations.

“We hope this approach will be particularly useful in resource-limited settings where the rate of undetected atrial fibrillation may be particularly high and the resources to detect it may be limited. However, more work is needed to overcome barriers to implementation, and further studies need to assess targeted screening strategies in these settings,” says Dr. Noseworthy.

“Now that we have demonstrated that AI-based atrial fibrillation screening is possible, we will also need to show that patients with atrial fibrillation detected by screening benefit from treatment to prevent strokes,” says the Dr. Noseworthy. “Our ultimate goal is to prevent I believe the current study has brought us a little closer.


Journal reference:

Noseworthy, PA, et al. (2022) Artificial Intelligence-Guided Screening for Atrial Fibrillation Using Sinus Rhythm Electrocardiogram: A Prospective Non-Randomized Interventional Trial. The Lancet.

Leave a Reply