Imagine a world where we could predict and prevent sudden cardiac deaths with the help of artificial intelligence. That’s exactly what a recent groundbreaking study has explored. Published on March 30, 2025, in the European Heart Journal, researchers from Inserm, Université Paris Cité, AP-HP, and their American partners have developed a neural network that mimics the human brain. This sophisticated algorithm examined data from over 240,000 ambulatory electrocardiograms, identifying patients at risk of severe arrhythmias that could lead to cardiac arrest within two weeks, with an impressive accuracy of over 70%.
Sudden cardiac death is a major concern, accounting for more than 5 million deaths worldwide each year. It often strikes without warning, affecting even those without a history of heart disease. The ability of AI to foresee life-threatening arrhythmias, which can result in fatal cardiac arrests, marks a significant advancement in medical science. Engineers from Cardiologs, a Philips entity, worked alongside Université Paris Cité and Harvard University to develop this neural network, aiming to boost prevention measures.
The study analyzed millions of heartbeats from six countries: the United States, France, the United Kingdom, South Africa, India, and the Czech Republic. Researchers focused on the heart’s electrical activity during a complete cardiac cycle, looking for subtle signs of arrhythmia risk.
Laurent Fiorina, who led the study at the Paris Cardiovascular Research Center and serves as a cardiologist at the Institut cardiovasculaire Paris Sud, shared, “We discovered that by analyzing 24-hour electrical signals, we could identify individuals likely to develop severe cardiac arrhythmia within the next two weeks. If untreated, this could lead to fatal cardiac arrest.”
The neural network, while still under evaluation, showed it could detect high-risk patients in 70% of cases and low-risk individuals in 99.9% of cases. This technology could soon be part of hospital monitoring systems and integrated into devices like Holter monitors or smartwatches.
Eloi Marijon, a research director at Inserm, explains, “What we’re proposing here is a change in how we prevent sudden deaths. Until now, we’ve tried to identify at-risk patients over the medium and long term, but predicting what might happen in the minutes, hours, or days before a cardiac arrest was beyond our reach. Today, thanks to AI, we can predict these events in the very short term and potentially intervene before it’s too late.”
Future clinical studies are planned to test the model’s effectiveness in real-world settings. Fiorina emphasizes, “It’s crucial for this technology to be evaluated in clinical trials before becoming part of medical practice. But we’ve already shown that AI has the potential to transform how we prevent severe arrhythmias.”
