An artificial intelligence system developed by Oxford University researchers could identify people at risk of heart attacks up to a decade in advance, potentially saving thousands of lives annually. The technology is currently being assessed by the National Institute for Health and Care Excellence (Nice), with a decision on NHS use expected by the end of the year.
The AI analyses standard CT scans to detect hidden signs of arterial inflammation, which are often missed by conventional methods. According to Professor Charalambos Antoniades, lead of the Orfan study, fewer than 20% of patients scanned for chest pain show obvious blockages, but two-thirds of the remaining 'safe' group later suffer major cardiac events. 'We have been missing signals from our scans,' he said.
The technique, published in the Lancet, uses enhanced imaging to reveal damage from inflammation and changes in fat around arteries. It has been validated using data from 40,000 UK patients. In 45% of cases, clinicians altered treatment—such as prescribing statins or colchicine—based on the AI's risk assessment.
The British Heart Foundation-funded study is also being evaluated in the US by the Food and Drug Administration and is already approved for clinical use in Europe. Researchers are developing similar AI systems to predict stroke and diabetes risk.
