Cardiac MRI AI Tool Improves Mortality Risk Prediction

These measurements can influence heart failure therapy, implanted device placement, revascularization procedures, surgical risk assessment, and monitoring during chemotherapy. New findings now show that artificial intelligence-based measurements from cardiac MRI may more accurately identify patients at increased risk of death.

Researchers at City St George’s, University of London and University College London evaluated an artificial intelligence tool that automatically measures ejection fraction from cardiac MRI scans. The tool generates heart-function measurements from imaging data that clinicians would normally assess manually, allowing its performance to be compared directly with measurements made by experienced clinicians. The study was published in The Lancet Digital Health on November 2, 2026.

The retrospective, multicenter prognostic study included 27,382 patients in the U.K. and United States who were undergoing investigation for known or suspected heart conditions. Patients were followed for an average of more than five years, during which 3,458 deaths occurred. AI-derived ejection fraction measurements were slightly but significantly better than clinician-derived measurements at identifying patients at increased risk of death.

The AI tool also completed the analysis substantially faster than manual assessment, processing scans in approximately two to three seconds per patient. Previous work by the same team found that expert clinician analysis typically takes around 13 minutes. Errors from the AI system were uncommon, occurring in approximately 1% of patients and affecting 0.08% of images.

The findings support the potential use of AI as a first-line approach for cardiac MRI analysis, with expert review retained when needed. The researchers emphasized that clinicians remain essential for verifying results and interpreting imaging findings within the broader clinical context. The team now plans to combine AI-derived heart scan measurements with other diagnostic data, particularly electrocardiograms, to help identify heart muscle abnormalities earlier.

“Ejection fraction is one of the most important measurements we make in cardiology because it helps determine how patients are treated and how closely they should be monitored. Our findings suggest AI can measure this important marker with remarkable consistency and precision, resulting in a small but meaningful improvement in identifying patients at higher risk,” said Dr. George Joy, NIHR clinical lecturer from the Cardiovascular Clinical Academic Group at City St George’s, University of London and lead author.

“Even modest improvements in accuracy can have a significant impact when applied across thousands of patients. AI is not replacing clinicians, but it can provide faster, more consistent measurements at low cost and support more streamlined clinical decision-making that ultimately benefits patients,” said Dr. Rhodri Davies, associate professor from University College London, senior author and inventor of the AI tool.

Related Links
City St George’s, University of London 
University College London