AI Tool Helps Cardiologists Diagnose Heart Attacks Using Electrocardiogram Technology

To diagnose conditions such as heart attacks and heart rhythm disturbances, clinicians usually depend on 12-lead electrocardiograms (ECGs)—sophisticated setups that involve placing multiple electrodes around the chest and limbs to capture the heart’s electrical activity. These ECGs require specialized equipment and trained personnel, which not all clinics have. Now, a team of scientists and clinicians has demonstrated that heart conditions can be diagnosed with comparable accuracy using just three electrodes and an artificial intelligence (AI) tool.

In their findings published on August 1, 2024, in npj Digital Medicine, researchers from Scripps Research (La Jolla, CA, USA) demonstrated that their AI algorithm could recreate full 12-lead ECGs using data from only three ECG leads. Additionally, clinicians were able to detect heart attacks nearly as accurately using the AI-generated ECGs as they could with traditional 12-lead ECGs. To develop this AI tool, the team utilized over 600,000 12-lead ECGs previously collected from patients, where approximately half displayed normal rhythms and the others showed various heart conditions. They experimented to determine which combinations of two or three electrodes could be used for the AI to accurately recreate the 12-lead data.

Image: Heart conditions can be diagnosed roughly as accurately using just three electrodes and an AI tool (Photo courtesy of Scripps Research)

The team then assessed a set of 238 ECGs, half of which indicated a heart attack. Cardiologists were shown either the original 12-lead ECGs or those reconstructed by the AI from three chosen leads, without knowing which was which. The cardiologists accurately identified signs of heart attacks 81.4% of the time in the AI-generated ECGs, closely approaching the 84.6% accuracy rate of the traditional 12-lead ECGs. The researchers noted that before this algorithm can be integrated into clinical decision-making, further prospective studies are necessary with various patient groups and in diverse clinical environments. Nonetheless, if the tool maintains its efficacy, it could facilitate the use of ECGs in settings that lack specialized equipment and personnel, potentially speeding up diagnoses and treatments for patients.

“This opens up the door to patients being able to get really high-quality, time-sensitive clinical data without traveling to somewhere that has a 12-lead ECG,” said cardiologist Evan Muse, MD, PhD, the lead of cardiovascular genomics at Scripps Research Translational Institute, assistant professor of Molecular Medicine at Scripps Research and co-senior author of the new paper. “It likely means not only increased access to ECG technology, but decreased costs and improved patient safety.”