Scientists at the University of Oxford have created a new artificial intelligence tool that is capable of predicting a person’s risk of developing heart failure up to five years in advance, The Guardian has reported.

Heart failure occurs when the heart is unable to pump blood effectively around the body and experts say identifying those at risk earlier could transform care, allowing doctors to intervene sooner or even prevent the condition from developing.

The newly developed AI system analyses routine cardiac CT scans, focusing on subtle changes in fat surrounding the heart that may signal inflammation and poor health - markers that are not detectable by the human eye. Researchers say this approach provides, for the first time, an accurate way to assess heart failure risk using standard imaging.

The tool generates an individual risk score to support clinical decision making, helping doctors determine how closely a patient should be monitored or treated. In trials, patients identified as high risk were up to 20 times more likely to develop heart failure than those at low risk, with around a 25% chance of developing the condition within five years.

The model was trained and tested on data from 72,000 patients across nine NHS trusts in England and followed over a 10-year period. It achieved an accuracy rate of 86% in predicting heart failure within five years. The findings were published in the Journal of the American College of Cardiology.

The research team is now seeking regulatory approval to introduce the technology into clinical practice.