AI model detects heart failure signs via standard ECG

An international research study led by Italian institutions has developed an Artificial Intelligence model capable of detecting signs of heart failure through a standard electrocardiogram (ECG). The study, published in the European Heart Journal – Quality of Care and Clinical Outcomes, utilized a deep neural network trained on 84,895 ECGs and NT-proBNP tests.

The AI can identify micro-electrical signals invisible to the human eye that correlate with high levels of NT-proBNP, a key biomarker for diagnosing heart failure. In tests conducted on 679 patients in Cosenza and Trieste, the algorithm demonstrated an accuracy rate of nearly 90% in identifying high-risk values.

This technology could potentially allow for rapid, low-cost diagnosis using a simple ECG, bypassing the need for more expensive and time-consuming blood tests or echocardiograms. Researchers involved include representatives from the University of Calabria, Humanitas University, University of Naples Federico II, University of Trieste, University of Zurich, Imperial College London, King’s College London, and Seoul National University.