Routine Heart Scan with AI Can Predict Heart Failure Five Years in Advance

Experts develop simple scan that can predict deadly heart failure - five years before it strikes A routine heart scan could soon reveal your risk of heart failure up to five years before it strikes, researchers say. This breakthrough marks a potential turning point in cardiovascular medicine, offering a glimpse into a future where heart failure—a condition that affects nearly one million people in Britain and kills around 170,000 annually—can be intercepted long before symptoms manifest. The innovation hinges on artificial intelligence, which analyzes cardiac CT scans to detect subtle changes in the fat surrounding the heart. These changes, invisible to conventional tests, signal early inflammation in the heart muscle, a key driver of the disease.

The condition, which occurs when the heart can no longer pump blood effectively, is expected to double in prevalence by 2040. This surge underscores the urgency of early detection, as late diagnosis often means patients arrive at hospitals with severe, irreversible damage. The new method, developed by scientists at the University of Oxford, could change that. By identifying high-risk individuals years in advance, it offers a window for intervention—whether through lifestyle changes, medication, or closer monitoring.

The AI system was trained on data from 72,000 patients in England who underwent cardiac CT scans between 2007 and 2022. It found that those flagged as high risk were 20 times more likely to develop heart failure than those at the lowest risk. Remarkably, high-risk patients had a one-in-four chance of developing the condition within five years, with the method predicting outcomes with 86% accuracy. This level of precision, achieved without invasive procedures or additional testing, could revolutionize how heart failure is managed.

Dr. Sonya Babu-Narayan, clinical director at the British Heart Foundation, emphasized the transformative potential of the tool. "Late diagnosis may mean patients already have severe damage to their heart muscle which might have been avoided," she said. "This approach could help doctors spot heart failure earlier, by monitoring more closely those at highest risk." The British Heart Foundation, which funded the research, noted that no reliable method existed previously to identify individuals who would develop heart failure.

The system's ability to generate an absolute risk score for each patient without human input is a major leap forward. Professor Charalambos Antoniades, who led the study, envisions the tool being integrated into routine chest scans, regardless of the scan's original purpose. "This will allow doctors to make more informed decisions about treatment," he said. "We hope that, if this programme is rolled out nationwide, it could reduce hospital pressures by helping patients live well for longer."

The NHS currently lists symptoms such as breathlessness, fatigue, dizziness, and swollen ankles or legs as indicators of heart failure. However, these often develop gradually, making early detection challenging. The new scan could shift the paradigm, enabling proactive care rather than reactive treatment. As the technology advances, questions about data privacy and equitable access to AI-driven diagnostics will arise. Yet, for now, the focus remains on saving lives—by catching a silent killer before it strikes.