AI Joins the Fight Against Heart Attacks

Sep 05, 2025
 

In this new edition of our newsletter, co-authored by Amir Aryani and Gavin Lambert , we focus on the application of "AI in Cardiology" as we are briefing on a recently published article in the European Heart Journal by Rick Volleberg, Niels Van Royen and their colleagues.

It seems it's no longer a question of if an AI will become a cardiologist's most trusted partner, but simply a question of when.

Imagine a hidden threat growing quietly inside your arteries, so small and hard to detect that it often isn’t noticed until it’s too late. This hidden danger is called a “thin-cap fibroatheroma”, a super high risk type of plaque that, if its fragile cap ruptures, can trigger a massive blood clot and a sudden heart attack or stroke. For years, cardiologists have been up against an enormous challenge trying to find these tiny "time bombs."

But what if technology could step in as a "plaque detective," offering a new level of precision and insight? A fascinating new study reveals that AI is doing just that, potentially changing how we predict and prevent heart attacks.

The Human Challenge: A Needle in a Haystack

Manually analysing artery scans is a difficult,  time-consuming task that demands specialized expertise. Even more perplexing, top experts often disagree on what they're seeing in the same scan, making a clear, confident diagnosis difficult at a time when it's most needed, perhaps during a procedure. This interobserver variability is a significant hurdle in consistent risk assessment.

Enter AI: The Plaque Detective

This is where advanced technology, specifically an AI algorithm called OCT-AID, steps in. The algorithm works in conjunction with Optical Coherence Tomography (OCT), which acts like a high resolution camera, providing incredibly detailed, high-definition views from inside the artery to detect coronary thin-cap fibroatheromas (TCFA). TCFAs have been clearly shown to be associated with adverse outcomes in patients.

This technology doesn't just glance at these images; it performs a pixel-by-pixel deep dive on every single frame of an entire scan, meticulously hunting for the telltale signs of a vulnerable plaque.

The Showdown: Humans vs. AI on Target Lesions

Researchers set up a head-to-head competition, using a dataset from 414 high-risk patients (with 488 lesions) who had already experienced a heart attack. The primary endpoint was a combination of death from any cause, subsequent non-fatal myocardial infarction or unplanned revascularisation within 2 years from the initial event.

When both human experts (a laboratory team focused on "target lesions"—the most obvious spots of concern) and the AI analysed the target regions, the results were telling.

AI TCFA detection was significantly associated with the primary endpoint whilst analysis by the core laboratory was not significant. The study demonstrated that AI augmented image analysis was able to identify patients at high risk for a future cardiac event.

Detection improved when the algorithm was applied to the entire scanned artery (a "complete pullback"). This is a task that is time-consuming for human teams to perform manually.

The results of this comprehensive analysis showed when the AI found a plaque anywhere in the entire artery, the patient's risk of a major cardiac event (including death, another heart attack, or emergency surgery) shot up by 5.5 times.

OCT-AID’s performance allowing clinicians to quickly and reliably see this complete picture was impressive.

Beyond Risk: The Power of Reassurance

The AI isn't just good at finding risk; it's also incredibly good at confirming safety. The study found a negative predictive value of 97.6%. What does this mean for a patient? 

If the AI scans your entire artery and gives you the "all-clear," you have a nearly 98% chance of not experiencing a major cardiac event for the next two years; thus providing a level of reassurance for patients and doctors.

The Future of Cardiology

The implications for cardiology are significant:

  • Standardisation: AI offers a way to improve and standardize how plaque is analysed, improving accuracy and the problem of experts disagreeing. It never gets tired and can consistently check the entire artery for risk.
  • Confident Decisions: This technology could give doctors the confidence needed to make critical, on-the-spot decisions, bridging the gap between research findings and daily clinical practice.
  • Comprehensive Insight: As the study authors conclude, looking at the complete picture with AI provides significantly more predictive power than focusing on just a single trouble spot.

This study demonstrates that AI isn't just matching human expertise; by enabling a bigger, more comprehensive picture, it's actually exceeding it in some ways. It seems the question is no longer if AI will help cardiologists by augmenting their expertise, but when it will become a mainstream capability.

Reference

Volleberg, R. H. J. A., Luttikholt, T. J., van der Waerden, R. G. A., Cancian, P., van der Zande, J. L., Gu, X., … van Royen, N. (2025). Artificial intelligence-based identification of thin-cap fibroatheromas and clinical outcomes: the PECTUS-AI study. European Heart Journal. https://doi.org/10.1093/eurheartj/ehaf595


We’d love to hear your thoughts, do you see AI becoming a trusted tool in clinical practice, or are there hurdles we still need to overcome? Share your perspective in the comments.

Best wishes, 

Amir Aryani & Gavin Lambert

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Cras sed sapien quam. Sed dapibus est id enim facilisis, at posuere turpis adipiscing. Quisque sit amet dui dui.
Call To Action

Stay connected with news and updates!

Join our mailing list to receive the latest news and updates from our team.
Don't worry, your information will not be shared.

We hate SPAM. We will never sell your information, for any reason.