University of Oxford-led research will examine whether artificial intelligence can identify atrial myopathy and predict cardioembolic stroke from routine cardiac CT at ESC 2026.

AI and cardiac CT

Researchers from the University of Oxford will present new studies exploring the use of artificial intelligence to assess cardiovascular risk from cardiac computed tomography (CT) scans at the European Society of Cardiology (ESC) Congress in Munich from August 28-31.

The research includes late-breaking data from the ORFAN-MAESTRIA study examining whether AI analysis of routine cardiac CT can identify atrial myopathy and predict cardioembolic stroke.

The study will be presented by Charalambos Antoniades, British Heart Foundation Chair of Cardiovascular Medicine at the University of Oxford, during the Late-Breaking Clinical Science session on advanced cardiovascular imaging on August 31.

The research programme also includes an investigation into whether coronary inflammation can be quantified from both contrast and non-contrast cardiac CT and identify people at increased cardiovascular risk independently of coronary artery calcium.

The findings are part of a wider programme examining how AI-based analysis of routine medical imaging could provide additional information about cardiovascular disease and risk.

Non-contrast CT under investigation

One study due to be presented at ESC will assess coronary inflammation on contrast and non-contrast cardiac CT in people with different levels of coronary calcification, including those with a zero calcium score.

The work will be presented by Kenneth Chan from the University of Oxford during a session focused on imaging, artificial intelligence and the future of coronary artery disease diagnosis.

Another ORFAN study will examine whether coronary inflammation can predict cardiovascular risk in people without standard modifiable risk factors. The research will be presented by Leanne Eveson on August 30.

The programme also includes a study described as a “virtual biopsy” of arterial cytokine activation. The research involves more than 80,000 patients from the UK and USA and will examine whether measures of arterial inflammation can predict long-term cardiac mortality.

Christos Kotanidis will present the research during an ESC session focused on prevention scores, risk prediction and polygenic risk on August 29.

The studies use artificial intelligence technology developed by Caristo Diagnostics, but the research presentations are being led by University of Oxford investigators.

Expanding AI-based cardiovascular risk assessment

Caristo CEO Steve Deitsch said: “With CaRi-Heart now FDA-authorized and in clinical use in the U.S., the research to be presented at ESC will explore the broader potential of our platform, including early-stage data applying FAI-Score to non-contrast cardiac CT that we’re continuing to develop.”

He added: “New stroke prediction research will also be presented, examining the potential of our technology to help address cardiovascular diseases beyond the coronary artery disease in the future.”

The ESC programme reflects growing interest in using AI to extract additional information from routinely acquired medical images rather than relying solely on conventional measures such as coronary artery calcium or visible coronary disease.

The research presented at the meeting will explore whether markers of inflammation and structural cardiac changes can provide additional information about future cardiovascular risk.

The stroke prediction study is particularly relevant because cardioembolic stroke can be associated with underlying atrial abnormalities that may not be identified through conventional assessment alone.

However, the ESC presentations represent research being presented for further scientific evaluation. The findings should therefore be considered in the context of the individual studies and their published results rather than as evidence that AI-based cardiac CT can currently replace established clinical assessment.

Pic suggestion: Realistic cardiac CT scan being reviewed on a medical imaging workstation by a radiologist, detailed heart and coronary artery imaging visible on screen, modern cardiovascular imaging department, clinical hospital environment, artificial intelligence and medical imaging concept, no futuristic CGI.

SEO tags: University of Oxford, ESC 2026, artificial intelligence, AI, cardiac CT, computed tomography, stroke prediction, cardioembolic stroke, atrial myopathy, cardiovascular disease, coronary inflammation, non-contrast CT, coronary artery disease, cardiovascular risk, medical imaging, AI in healthcare, Caristo Diagnostics, ORFAN-MAESTRIA, FAI-Score, cardiovascular imaging