AI with EKG can detect a form of heart disease, research shows

Dive Brief:

Use of artificial intelligence (AI) with an electrocardiogram (EKG) can create an inexpensive, noninvasive test for asymptomatic left ventricular dysfunction, a precursor to heart failure, a study by the Mayo Clinic has found.
An AI/EKG test offers an alternative to expensive imaging tests such as echocardiograms and CT and MRI scans as well as tests that measure natriuretic peptide levels in the blood (BNP) that have sometimes yielded disappointing results, the Mayo researchers said.
The accuracy of the AI/EKG test compared favorably with other common screening tests such as prostate-specific antigen for prostate cancer, mammography for breast cancer and cervical cytology for cervical cancer, the study found. The results were published in Nature Medicine.

Dive Insight:

Asymptomatic left ventricular dysfunction, which affects seven million Americans, is characterized by weak pumping in which the heart can’t push enough blood into circulation. It increases a patient’s risk of overt heart failure and mortality but is treatable when identified.

Congestive heart failure affects more than five million Americans and generates more than $30 billion in healthcare expenditures in the United States, the Mayo Clinic said.

There currently is no inexpensive, noninvasive test for asymptomatic left ventricular dysfunction, and BNP tests are the best existing screening but require a blood draw and can yield disappointing results, the hospital said.

"The ability to acquire an ubiquitous, easily accessible, inexpensive recording in 10 seconds – the EKG – and to digitally process it with AI to extract new information about previously hidden heart disease holds great promise for saving lives and improving health," study author Paul Friedman, chair of the Midwest Department of Cardiovascular Medicine at Mayo Clinic, said in a press release.

The Mayo Clinic researchers used the hospital’s stored digital data to create and test a neural network. They screened 625,326 EKG and transthoracic echocardiograms to identify the population to be studied for analysis.

The study concluded that AI applied to a standard EKG reliably detects asymptomatic left ventricular dysfunction. The test also showed that patients with a positive AI screen who did not have ventricular dysfunction were at four times the risk of developing future ventricular dysfunction, compared with those with a negative screen. This is most likely due to the test identifying early, subtle EKG changes that occur before heart muscle weakness, the researchers said.