© 2024 Connecticut Public

FCC Public Inspection Files:
Public Files Contact · ATSC 3.0 FAQ
Play Live Radio
Next Up:
0:00 0:00
Available On Air Stations

AI could help detect irregular heart rhythms in EKGs that humans can't see


Using artificial intelligence, cardiologists have developed a way to predict who will develop atrial fibrillation, or A-Fib. It's a common type of arrhythmia that can be dangerous if left untreated. NPR's Allison Aubrey reports.

ALLISON AUBREY, BYLINE: If you've ever had an EKG, or electrocardiogram, you know they're quick and painless. Tiny electrodes are placed on your chest, and your heart's electrical signals display as little waves and squiggles on a screen. Dr. Neal Yuan of the San Francisco VA Medical Center says this gives him lots of information to help make a diagnosis.

NEAL YUAN: We look at all those squiggles, and then we say, well, we've got these rules for what sort of squiggle patterns look like what. And we have certain ideas for certain diagnoses based on certain patterns.

AUBREY: This may sound straightforward. The EKG has been around about a hundred years, and doctors know how to spot the obvious things - say, a heart attack or active A-Fib. But inside these little squiggles and waves, there's lots of information that doctors just can't easily see. But Dr. Yuan says technology can help.

YUAN: The machine can learn from seeing millions of ECGs. And it doesn't forget, and it, you know, doesn't grow tired (laughter), unlike, you know, humans.

AUBREY: He says each EKG produces about 20,000 numbers to decipher, which can overwhelm the human brain. But a machine can crunch these quickly. So as part of the new study, funded by the National Institutes of Health, he and some collaborators at Cedars-Sinai fed millions of data points from EKGs into a computer.

YUAN: What deep learning and machine learning allows us to do is it can hash through all of that information in the 20,000 different numbers.

AUBREY: And identify complicated relationships. In his study, the goal was to identify who is at risk of A-Fib. So they had the machine assess the EKGs of patients who'd had A-Fib in the last month compared to those who had not to look for subtle differences.

YUAN: So it essentially takes in an ECG, and then it makes a guess based off those 20,000 numbers. And then it learns whether that guess is right or wrong, and then it adjusts its model to make a better guess next time.

AUBREY: Turns out the model they developed actually helped them predict who would develop A-Fib.

YUAN: I'm really excited about it.

AUBREY: Their new study, published in the medical journal JAMA Cardiology, is the first step to bringing this to clinical practice.

YUAN: We are at the forefront of this wave right now, right? And it's definitely coming.

AUBREY: Used in the right ways, he says AI can help doctors do their jobs better. Allison Aubrey, NPR News. Transcript provided by NPR, Copyright NPR.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.

Allison Aubrey is a correspondent for NPR News, where her stories can be heard on Morning Edition and All Things Considered. She's also a contributor to the PBS NewsHour and is one of the hosts of NPR's Life Kit.

Stand up for civility

This news story is funded in large part by Connecticut Public’s Members — listeners, viewers, and readers like you who value fact-based journalism and trustworthy information.

We hope their support inspires you to donate so that we can continue telling stories that inform, educate, and inspire you and your neighbors. As a community-supported public media service, Connecticut Public has relied on donor support for more than 50 years.

Your donation today will allow us to continue this work on your behalf. Give today at any amount and join the 50,000 members who are building a better—and more civil—Connecticut to live, work, and play.