Artificial intelligence can now predict deadly heart arrhythmias up to two weeks in advance, potentially transforming cardiac care.
A new study published in the European Heart Journal reveals that AI could play a crucial role in preventing sudden cardiac death. Researchers from Inserm, Paris Cité University, and the Paris public hospital group (AP-HP), working with colleagues in the United States, have developed an artificial neural network modeled after the human brain.
The AI algorithm analyzed data from over 240,000 ambulatory electrocardiograms (ECGs) across six countries and successfully identified patients at high risk of experiencing life-threatening arrhythmias within two weeks—with accuracy exceeding 70%.
Sudden cardiac death claims more than 5 million lives globally each year, often striking without warning in individuals with no prior heart disease diagnosis.
How the AI Works
Engineers from Cardiologs (Philips group), in collaboration with Paris Cité and Harvard universities, developed the neural network to mimic brain functions. The algorithm analyzed millions of hours of heartbeat data from ECGs collected in the USA, France, UK, South Africa, India, and Czechia.
Through this analysis, researchers identified new weak signals that predict arrhythmia risk, particularly focusing on the time needed to electrically stimulate and relax heart ventricles during complete cardiac cycles.
“By analyzing electrical signals for 24 hours, we can identify subjects susceptible to developing serious heart arrhythmias within the next two weeks. If left untreated, these arrhythmias can progress to fatal cardiac arrest,” explains Dr. Laurent Fiorina, the study’s first author. Dr. Fiorina is a researcher at the Paris Cardiovascular Research Centre (PARCC), a cardiologist at Cardiovascular Institute Paris-Sud, and medical director for artificial intelligence at Philips.
Impressive Results and Future Applications
While still in the evaluation phase, the neural network detected at-risk patients in 70% of cases and correctly identified no-risk patients in 99.9% of cases.
This algorithm could eventually monitor at-risk patients in hospitals. With further refinement, it might be integrated into ambulatory Holter monitors or even smartwatches.
“We’re proposing a paradigm change in sudden death prevention,” notes Eloi Marijon, Inserm research director at PARCC and cardiology department head at Georges Pompidou European Hospital. “Until now, we’ve tried to identify medium and long-term risks but couldn’t predict what might happen in the minutes, hours, or days before cardiac arrest. Now, thanks to AI, we can predict these events in the very short term and potentially intervene before it’s too late.”
The team plans to conduct prospective clinical studies to test the model under real-world conditions. Dr. Fiorina emphasizes, “It’s essential for this technology to be evaluated in clinical trials before being used in medical practice. But we’ve already shown that AI has the potential to radically transform the prevention of serious arrhythmias.”
Reference: “Near-term prediction of sustained ventricular arrhythmias applying artificial intelligence to single-lead ambulatory electrocardiogram” by Laurent Fiorina, Tanner Carbonati, Kumar Narayanan, Jia Li, Christine Henry, Jagmeet P Singh and Eloi Marijon, 30 March 2025, European Heart Journal.
