A groundbreaking AI-powered ECG model has emerged as a game-changer in the early detection of severe heart attacks, offering a more accurate and efficient approach than traditional triage methods. This innovative technology, developed by researchers at AZORG Hospital in Belgium, has the potential to revolutionize emergency cardiac care and save lives.
The study, published in JACC: Cardiovascular Interventions and presented at TCT 2025 in San Francisco, focused on ST-segment elevation myocardial infarction (STEMI), a critical type of heart attack where a major coronary artery is blocked. Quick action to restore blood flow is crucial, but delays often occur, especially in non-specialized hospitals and rural areas. These delays can have devastating consequences, with mortality rates three times higher when the time to reperfusion exceeds 90 minutes.
Lead author Dr. Robert Herman emphasizes the dual benefit of AI-driven ECG interpretation: "It can identify true heart attacks early and reduce unnecessary activations, bringing a balanced approach to emergency care." By improving triage accuracy, this technology streamlines emergency response, reduces strain on clinical teams, and ensures prompt intervention for patients in need.
In a large-scale real-world evaluation, researchers analyzed data from 1,032 patients with suspected STEMI across three diverse primary PCI centers between 2020 and 2024. The AI ECG model, named "Queen of Hearts," was trained to detect acute coronary occlusion, including STEMI equivalents, and differentiate them from benign mimics. The results were impressive: the model detected 553 out of 601 confirmed STEMIs, outperforming standard triage which identified only 427. Furthermore, the AI ECG model had a significantly lower false-positive rate of 7.9% compared to 41.8% for standard triage, representing a remarkable fivefold reduction.
Dr. Timothy D. Henry, senior author of the study, highlights the potential impact: "AI-enhanced STEMI diagnosis can shorten time to treatment and reduce false activations, especially benefiting non-PCI centers in optimizing STEMI patient transfers."
However, the study also sparks controversy. In an editorial comment, Dr. Mohamad Alkhouli from the Mayo Clinic commends the researchers for addressing a complex aspect of interventional cardiology but cautions that the AI model, originally developed for occluded arteries, requires further validation across diverse patient populations. He emphasizes the challenge of integrating AI into high-stakes clinical settings, where it must complement human judgment.
This study opens a new chapter in the integration of AI into cardiac care, raising important questions about the role of technology in emergency medicine. As we navigate this exciting yet complex landscape, one thing is clear: the potential for AI to save lives is immense, but it must be approached with careful consideration and further research.
