Recently developed artificial intelligence software can determine whether firefighters may be about to experience a potentially fatal cardiac event, researchers say.

A team of researchers from the National Institute of Standards and Technology, the University of Rochester, and Google recently published a report that details their machine-learning model.

Using AI technology and existing electrocardiogram data for 112 firefighters, the researchers created the Heart Health Monitoring, or H2M, model.

Individual heartbeats from the EKGs were classified either as normal or abnormal, with the latter possibly indicating irregular heart rhythms. Such rhythms can prompt the heart to stop pumping blood, often because of a heart attack, and trigger sudden cardiac arrest – the leading cause of on-the-job death among firefighters. 

The H2M model correctly identified abnormal EKG samples with nearly 97% accuracy.

In a press release, NIST researcher Chris Brown says sudden cardiac events “are by far the No. 1 killer of firefighters.” The release notes that about 40% of on-duty fatalities can be attributed to sudden cardiac arrest events, which “kill on-duty firefighters at twice the rate of police officers and four times the rate of other emergency responders.”

The researchers say they hope the model ultimately can be used in a portable heart monitor that firefighters could wear while on the job.

“This technology can save lives,” NIST researcher Wai Cheong Tam said in the release. “It could benefit not only firefighters, but other first responders and additional populations in the general public.”