OTTAWA, April 24, 2025 — In a major advance for heart attack treatment, researchers from the Ottawa Heart Institute, Montreal Heart Institute, and Centre hospitalier de l'Université de Montréal (CHUM) have successfully used artificial intelligence (AI) to assess heart function in real time during emergency procedures. Their findings, published in New England Journal of Medicine Artificial Intelligence (NEJM AI), show how CathEF, a new AI tool, can quickly and accurately measure how well the heart is pumping during a life-saving medical procedure.
“This technology allows us to monitor heart function instantly and accurately during critical procedures,” said Pascal Thériault-Lauzier, MD, a clinician-scientist at the Ottawa Heart Institute and lead author of the study published in NEJM AI. “With AI integrated into the catheterization lab, doctors can make faster, better-informed decisions, improving patient care in life-or-death situations.”
Currently, when a patient suffers sudden reduced blood flow to the heart, they undergo a procedure called coronary angiography in which doctors use X-rays to look at the heart's blood vessels to assess how well the heart is pumping. This is called ejection fraction. Traditionally, additional tests, such as echocardiograms (heart ultrasounds) or the more invasive imaging procedure ventriculography, are required to guide clinical decision-making. These tests however take time and carry risks.
CathEF solves this problem by analyzing routine X-ray images taken during angiography. In a study at two academic tertiary care centres (the Ottawa Heart Institute and Montreal Heart Institute), the tool was shown to provide heart function measurements that are just as accurate as traditional tests, but without the extra procedures, contrast dye, or complications. The result leads to better, faster decisions and improved outcomes for patients suffering acute coronary syndromes and heart attacks.
“We demonstrated that CathEF can reliably estimate ejection fraction in patients experiencing a heart attack, paving the way for early identification of those with normal or reduced cardiac function,” said Robert Avram, MD (in French), an interventional cardiologist and AI researcher at the Montreal Heart Institute and corresponding author of the study. “This has the potential to guide more targeted management strategies and will need to be confirmed in future studies.”
The researchers also hope future studies will explore whether AI can be used in the treatment of other heart conditions and whether using CathEF as a screening tool could be cost-effective.
PACS-AI: A platform powering AI in patient care
This breakthrough was made possible by PACS-AI, a platform co-developed by the Ottawa Heart Institute, Montreal Heart Institute, and CHUM. PACS-AI is an open-source system designed to integrate AI tools into hospitals' existing imaging systems.
In the past, AI tools have been difficult to use in clinical settings because of data privacy concerns and compatibility challenges. PACS-AI addresses this problem by acting as a bridge between AI and hospital systems that store and manage medical images (known as PACS or Picture Archiving and Communication System).
Supported by Canadian Institute for Advanced Research (CIFAR) Solution Network on Integrated AI for Health Imaging and Institute for Data Valorization (IVADO), PACS-AI represents a major step toward making AI a regular part of medical care.
A new era of AI in medicine
“By integrating AI tools like CathEF into PACS-AI, this research demonstrates that AI can be used ethically and effectively in hospitals to enhance human decision-making, improve care, and deliver better outcomes for patients,” said Dr. Thériault-Lauzier. This technology is designed to complement human intervention, ensuring that innovation works in tandem with the compassion and intuition of health professionals.
The researchers believe that this technology could eventually be a standard part of heart attack care worldwide, helping doctors make faster decisions and improving outcomes for patients everywhere.
For more information
- Read the paper: Clinical deployment of real-time left ventricular ejection fraction estimation from coronary angiography
Media contact
Leigh B. Morris
Communications Officer
University of Ottawa Heart Institute
613-316-6409 (cell)
lmorris@ottawaheart.ca
Geneviève Bettez
Montreal Heart Institute
Head of Public Relations and Digital Communications
Deputy Director of Communications and Media Relations
514-318-2938
genevieve.bettez@icm-mhi.org
