FDA Clears Queen of Hearts AI for STEMI ECG Triage
Powerful Medical, a New York digital-health company, received FDA De Novo clearance for Queen of Hearts, an artificial intelligence system that detects STEMI and STEMI equivalents on electrocardiograms,
PromptCrates Editorial
Staff Writer
Powerful Medical, a New York digital-health company, received FDA De Novo clearance for Queen of Hearts, an artificial intelligence system that detects STEMI and STEMI equivalents on electrocardiograms, Cardiovascular Business reported on 4 September 2026. The clearance positions the tool for first-contact triage where minutes still decide whether a blocked artery is recognized before cath-lab activation delays compound.
What Queen of Hearts is designed to catch
STEMI—ST-elevation myocardial infarction—is the classic ECG emergency, but STEMI equivalents can be subtler and easier to miss under time pressure. Queen of Hearts is trained to flag both classes so emergency clinicians and first-contact teams can escalate faster. Clinical collaboration with Stephen W. Smith, MD, of Hennepin County Medical Center shaped the pattern library around real-world tracings rather than textbook cartoons alone.
The product sits inside Powerful Medical's PMcardio platform, which is meant to embed AI reads into existing ECG workflows instead of forcing a separate viewer. That embedding strategy mirrors broader FDA digital-health experiments that emphasize integration over novelty—context also covered in our reporting on the FDA TEMPO pilot with four companies. Hospitals evaluating Queen of Hearts will care less about a demo screenshot and more about whether the alert lands where nurses and physicians already look.
Regulatory history gave the De Novo a running start. Queen of Hearts already held a CE mark for European markets and won FDA breakthrough device designation in March 2025. Breakthrough status does not equal clearance, but it does create a tighter dialogue with reviewers on evidence plans—useful for ECG AI where false negatives and false positives both carry clinical cost.
Evidence from TCT and JACC interventions
Supporting data presented around TCT 2025 and discussed in JACC: Cardiovascular Interventions associated Queen of Hearts with improved detection performance and fewer false positives relative to conventional interpretation pathways studied by the sponsors. Exact endpoint definitions belong in the peer-reviewed tables, yet the directional claim matches what emergency departments want: catch more true occlusions without flooding the cath lab with noise.
Robert Herman, MD, PhD, Powerful Medical's chief medical officer, has framed the clinical problem as delay at first contact—the moment when an ECG is acquired but expertise is unevenly distributed across shifts and sites. Felix Bauer, the company's COO, has similarly emphasized operational triage: getting a reliable second read into the hands of teams that cannot always wait for a specialist over-read. Those quotes matter because they set expectations. Queen of Hearts is sold as triage support, not as an autonomous cath-lab activation button.
Health systems should still demand local validation. ECG vendors, lead placement habits, and population mix differ across regions. A De Novo decision unlocks US marketing; it does not guarantee identical positive predictive value in every ED. Sensible rollouts start with shadow mode, measure alert burden, and only then wire alerts into STEMI pager workflows.
Why FDA digital health context matters now
Cardiology AI clearances are accelerating, and regulators are experimenting with lifecycle oversight models for adaptive software. The TEMPO-style pilots show FDA interest in tighter manufacturer dialogue after authorization, not only at the single decision letter. Queen of Hearts enters that environment as a high-stakes ECG triage claim: if the model drifts or site performance decays, the clinical downside is immediate, a theme that also shows up in broader capability-gating debates covered in our frontier cyber gating roundup.
For buyers, the checklist is concrete. Confirm how PMcardio stores and displays model versioning. Ask whether STEMI-equivalent definitions match your cardiology group's activation criteria. Review liability language for alert overrides. And train staff that a cleared AI can still miss edge cases—especially paced rhythms, early repolarization variants, and noisy prehospital tracings.
Competitors will respond. ECG AI is a crowded category, and a De Novo for STEMI equivalents raises the marketing bar for anyone claiming chest pain triage. Expect rival vendors to publish head-to-head reader studies and hospitals to demand prospective metrics on door-to-activation time, false activation rates, and clinician trust scores. Queen of Hearts now has a US regulatory story; sustaining it will depend on post-market evidence, not only the September clearance headline.
Patients should understand the human loop remains. An AI flag can speed recognition, but reperfusion still requires clinical judgment, cath-lab capacity, and transfer logistics. The promise of Queen of Hearts is fewer missed equivalents at the first ECG—not a replacement for cardiology teams. If hospitals pair the tool with training and audit, the De Novo could translate into minutes saved. If they treat clearance as autopilot, the same alerts can create alarm fatigue instead of better care.
Procurement committees should also budget for change management. ECG AI that only lives in a pilot dashboard rarely moves door-to-balloon metrics; the wins show up when triage nurses, EMS partners, and cardiology fellows share a common alert language. Powerful Medical's PMcardio embedding story will be judged on whether Queen of Hearts becomes part of that shared language without adding another ignored beep to an already noisy bay.
