Artificial Intelligence Outperforms Human Clinicians in Detecting Hidden Heart Attacks During Emergency ECG Screening

New study from ESC 2026 shows AI-based ECG interpretation is twice as accurate as human clinicians in detecting certain types of heart attacks.

By: AXL Media

Published: Mar 20, 2026, 8:36 AM EDT

Source: Information for this report was sourced from European Society of Cardiology

Revolutionizing Diagnosis for Non-Traditional Heart Attacks

The landscape of emergency cardiac care is facing a potential shift following new data showing that artificial intelligence can identify life threatening arterial occlusions more effectively than conventional human led methods. In patients presenting with acute coronary syndrome, the absence of a specific ECG sign known as an ST elevation often complicates the diagnostic process, leading to dangerous delays in treatment. According to findings presented in Lisbon, a new smartphone based AI algorithm has demonstrated the ability to bridge this gap, offering a more precise tool for early intervention.

Clinical Performance and Statistical Superiority

The prospective study, led by Doctor Federico Nani of Central Hospital Bolzano, evaluated ,1490 patients who exhibited symptoms of heart attacks but lacked the characteristic ST elevation on their initial scans. While human led ECG interpretation correctly identified occlusive myocardial infarctions in only 42% of cases, the AI based algorithm achieved a correct identification rate of 84%. With a specificity of 99% and a negative predictive value of 98%, the tool proved to be exceptionally reliable at both detecting active heart attacks and ruling out those not requiring immediate invasive surgery.

Addressing the Challenges of Emergency Triage

Standard diagnostic pathways currently rely heavily on monitoring troponin levels and performing coronary angiography, processes that can take significant time in a high pressure emergency room setting. Doctor Nani explained that many patients without an ST elevation still suffer from total arterial blockages, but clinicians often struggle to recognize these "hidden" heart attacks quickly. The AI tool's ability to provide a high accuracy reading from the very first ECG allows for the optimization of patient management, potentially directing high risk individuals to the catheterization lab much sooner than traditional protocols allow.