Scientists have discovered that heart conditions can be diagnosed using artificial intelligence (AI) with fewer electrocardiogram (ECG) leads.
A study by Scripps Research in California, published on August 1, 2024, in npj Digital Medicine, shows their AI algorithm can create full 12-lead ECGs using data from just three leads. This means doctors can identify heart attacks with almost the same accuracy as with traditional 12-lead ECGs, which use multiple electrodes on the chest and limbs.
Dr. Evan Muse, a cardiologist and co-senior author of the study, stated, “This could allow patients to get high-quality, timely clinical data without needing a full 12-lead ECG setup, increasing access to ECG technology, reducing costs, and improving patient safety.”
The research team used data from over 600,000 ECGs to develop the AI tool. They found that combining two limb leads with one chest lead provided the best results. The AI was tested with 238 ECGs, half showing heart attacks. Cardiologists couldn’t tell the AI-generated ECGs from the original ones and identified heart attacks with nearly the same accuracy.
Dr. Giorgio Quer, co-senior author, emphasized the importance of ensuring cardiologists could accurately interpret the AI-generated data. Further studies are needed before the algorithm can be used in clinical decision-making, but it shows promise for use in various settings with less specialized equipment.
Source: theengineer
