A major safety study led by UC Davis Health tested an artificial intelligence (AI)-powered electrocardiogram (ECG) model to see how well it could detect severe heart attacks.
The findings showed that the Queen of Hearts AI-based ECG platform outperformed standard triage in the emergency department and two other locations. It identified ST-elevation myocardial infarction (STEMI) heart attacks more accurately and with far fewer false alarms.
The results were published in JACC: Cardiovascular Interventions. The findings were also presented at the 2025 Transcatheter Cardiovascular Therapeutics (TCT) conference.
"These results demonstrate the potential of AI-based applications to transform emergency cardiovascular care," said Bryn Mumma, a UC Davis Health professor of emergency medicine and primary investigator of the study. "By improving triage accuracy, we can ease the burden on clinical teams, minimize false positives and ensure that patients who truly need urgent intervention receive it without delay."
"These results demonstrate the potential of AI-based applications to transform emergency cardiovascular care."-Bryn Mumma
What is a STEMI heart attack?
A STEMI is a severe type of heart attack caused by a completely blocked artery.
The best treatment is to quickly restore blood flow using a procedure called percutaneous coronary intervention, or PCI. When treatment takes longer than 90 minutes, the risk of death is about three times higher.
"STEMI is a life-threatening emergency where every minute matters," explained Mumma. "Accurate diagnosis and rapid treatment to restore blood flow are critical and any tool that speeds up this process can help us save more lives."
UC Davis Medical Center recently received a Get With The Guidelines: Coronary Artery Disease STEMI Receives Gold Plus award from the American Heart Association. Hospitals earn the award by demonstrating a commitment to treating patients according to the most up-to-date, research-based guidelines for STEMI care.
UC Davis is one of only three sites for study
The heart attack study evaluated the AI-based ECG platform for STEMI triage at UC Davis Medical Center and the emergency departments at two other hospitals in the United States.
Researchers reviewed records from more than 1,000 patients suspected of having a STEMI heart attack in cases that activated emergency treatment protocols. The data came from the three geographically diverse hospitals collected between January 2020 and May 2024. Each patient's initial ECG underwent analysis by the AI ECG Model designed to detect blocked coronary arteries and distinguish them from false positives.
Follow-up tests confirmed that 58% of patients had STEMI, while 42% were false alarms. The AI model outperformed standard triage, correctly identifying 553 of the confirmed cases on the initial ECG compared to 427 detected by traditional methods. It also dramatically reduced false positives, about 8% versus nearly 42% with standard triage.
"These results indicate that using the AI-based model at the first medical contact has the potential to shorten time to treatment and reduce false activations," shared Mumma.
The AI model used in the study, she noted, should be interpreted with caution. While AI offers powerful capabilities, it must serve as a support tool — not a substitute for clinical judgment.
"We're dedicated to driving innovation in emergency care at UC Davis Health," explained Mumma. "Taking part in this study reflects our commitment to equipping physicians with advanced tools to protect patient health. The best outcomes happen when technology and clinicians work hand in hand, combining powerful tools with expert medical judgment."
