Mini-Camera and AI Accurately Predict Heart Attack Recurrence - European Medical Journal

Mini-Camera and AI Accurately Predict Heart Attack Recurrence

A COMBINATION of miniature camera technology and artificial intelligence can now accurately predict which patients are at the highest risk of suffering another heart attack within two years, according to a Dutch study.  

Heart attacks are a result of blocked coronary arteries, often caused by a build-up of fat and cholesterol forming plaques that can rupture. After surviving their first heart attack, up to 15 percent of patients go on to experience another event within two years, despite receiving standard treatments like angioplasty and stent placement. Identifying which patients remain at risk has traditionally required complex, labour-intensive analysis of high-resolution images taken inside the artery walls, limiting its usefulness beyond highly specialised centres. 

In the study, researchers from Radboud university medical center used a miniature optical coherence tomography (OCT) camera to scan the coronary arteries of 438 heart attack survivors. These microscopic images, collected as the camera travelled through the artery, were then analysed by a newly developed AI algorithm and an independent, expert laboratory. Over two years of follow-up, the AI detected vulnerable regions known as thin-cap fibroatheromas in 34.5 percent of patients, while the core laboratory identified them in 30 percent. AI-based analysis was more strongly associated with predicting the combined outcome of recurrence, unplanned interventions or death, achieving a hazard ratio of 5.50 and a negative predictive value of 97.6 percent for adverse events across the entire artery. This outperformed the traditional focus on single target lesions. 

These advances mean that detailed artery scanning, aided by AI, could become part of routine care, helping physicians better tailor treatment and potentially prevent future heart attacks. In clinical practice, this could allow the identification of hidden high-risk plaques, prompting early intervention with medication or preventive stenting. Although it may take a few years before full implementation, the technology marks a substantial step towards more personalised and proactive heart attack care. 

Reference 

van Royen N et al. Artificial intelligence-based identification of thin-cap fibroatheromas and clinical outcomes: the PECTUS-AI study. European Heart Journal. 2025;DOI:10.1093/eurheartj/ehaf595.  

Author:

Each article is made available under the terms of the Creative Commons Attribution-Non Commercial 4.0 License.

Rate this content's potential impact on patient outcomes

Average rating / 5. Vote count:

No votes so far! Be the first to rate this content.