Queen of Hearts: the FDA approves an AI to read ECGs and detect hidden heart attacks, though its cost savings remain unproven

By Zendoric · September 12, 2026.

The US Food and Drug Administration (FDA) has granted a "De Novo" classification —the pathway reserved for devices unlike anything already approved— to Powerful Medical's STEMI AI ECG model, a medical technology company based in New York. The system, dubbed "Queen of Hearts," analyzes electrocardiograms (ECGs) from patients with chest pain to detect acute arterial blockages, the cause of the most severe heart attacks. According to Stat News, the specialist outlet that broke the story, fewer than ten AI devices receive this classification each year: it is a regulatory rarity, not a routine formality.

The system's clinical value lies not in detecting the obvious, but the opposite. The ECG pattern that reveals a severe heart attack is called STEMI (ST-segment elevation, the electrical signature of a blocked coronary artery) and the model identifies it quickly. The problem, according to the source, is that up to 50% of patients with blocked arteries do not show that pattern on the first ECG. That misreading can delay treatment by four to five hours, according to Powerful Medical itself, a margin that in cardiology is the difference between saving a heart and losing it. The other side of the problem is the reverse: between 25% and 40% of severe heart attack alerts turn out to be false alarms that needlessly mobilize teams of specialists who prepare the patient for surgery, with the cost that entails.

The epidemiological context explains the urgency. Heart disease is the leading cause of death in the United States and caused roughly one in three deaths in 2023, according to the Centers for Disease Control and Prevention (CDC). But the alarmism deserves qualification: a study published in 2025 in the Journal of the American Heart Association documents that heart attack mortality fell by nearly 90% between 1970 and 2022, partly thanks to medical innovation. "Queen of Hearts" is not arriving in virgin territory, but into a market already crowded with legacy software and other AI tools that aim to detect everything from atrial fibrillation to genetic heart conditions, as Stat News also reports.

The regulatory and economic piece deserves attention of its own. Before this classification, the algorithm already held "breakthrough" device designation, which eases access to supplemental Medicare payments for expensive technology under a more flexible regime that, according to Stat News, expires in 2028. And in Houston, Memorial Hermann-Texas Medical Center is planning a pilot program to test whether the algorithm generates real savings. In other words: the FDA's endorsement certifies that the model works on clinical paper, but the business case —how much money and how many lives it saves in practice— is still being built.

Our reading is that this is exactly the kind of AI application that matters most in the short term, and precisely because it is unflashy. There is no pretense here of replacing the doctor: the tool is designed for paramedics and emergency physicians, who are the first to see the patient and do not always have a cardiologist's specialized training to recognize those atypical "STEMI equivalents," as Powerful Medical's own chief executive calls them to Stat News. It is pattern recognition applied to a well-defined medical signal, with strict regulatory oversight and a narrow use case: the antithesis of the artificial general intelligence narrative. And that narrowness is precisely what makes it credible.

At the same time, it is worth keeping the short-term limits in view: the economic savings have yet to be proven, Medicare's flexible payment window has a 2028 expiration date —forcing Powerful Medical and hospitals like the one in Houston to generate real evidence before the subsidy runs out— and the AI-for-ECG market is saturated enough that differentiation is not guaranteed. None of this detracts from the advance: if the promise holds up, it means hours of treatment gained for up to half the patients the system misclassifies today. Multiplied across hundreds of equally narrow applications —in cardiology, oncology, neurology— this is precisely the kind of incremental, auditable, regulator-approved progress that, compounded over years, can genuinely bring us closer to the underlying horizon we defend at Zendoric: not a miracle cure overnight, but the sustained erosion of the planet's leading cause of death, one timely diagnosis at a time.