When minutes matter: heart attack and diagnostic uncertainty

One of the key signs on an ECG is ST-segment elevation, which may signal coronary artery disease. In acute coronary syndrome, i.e., a heart attack, a patient has roughly 90 minutes for an intervention to be performed without permanent consequences. Current detection, however, is imprecise: among patients flagged as positive, data suggest that up to around 40 % are false positives, which burdens the system. Worse still, some true heart attacks are identified too late, leading to delayed treatment and serious health impacts.

AI in the service of physicians: PMcardio

Powerful Medical has developed the PMcardio solution, which uses artificial intelligence to evaluate ECGs in any clinical setting. The models analyze a wide range of heart conditions and also take into account the patient’s current status. Uniquely, the system can handle paper records as well, which is important where digitization is not yet complete. The goal is more accurate triage and faster decision-making at the bedside.

Real-world validation shows the benefits: at one Belgian cardiovascular center, the number of false positives in suspected heart attacks was reduced by 68 %, while all true heart attacks were captured. A U.S. center reports similar results, with false positives reduced while maintaining high detection accuracy. The technology received an EU EIC grant, obtained MDR medical certification in 2022, and has been on the market for more than a year; in the meantime, it has processed over 300,000 ECG records. A team of more than fifty people works on the project, based mainly in Bratislava, and collaborates with centers in Europe, the USA, and Israel.