A study conducted by researchers from the Radcliffe Department of Medicine at the University of Oxford has investigated the potential use of artificial intelligence in diagnosing heart disease.
Funded by the British Heart Foundation and supported by the National Institute for Health and Care Research Oxford Biomedical Research Centre, the study analyzed routine cardiac computed tomography (CT) scans and medical data from over 40,000 individuals at eight hospitals in the UK. Patients were monitored for up to 10 years to track any major cardiac events such as heart attacks, heart failure, or cardiac death.
The research found that while individuals with significant coronary artery narrowing on CT scans were at a higher risk of serious cardiac events, two-thirds of these events occurred in patients without such narrowings.
Kenneth Chan, a clinical research fellow at the University of Oxford, explained that AI technology was able to detect inflammation levels in heart arteries by identifying changes in fat tissue surrounding the arteries that are not visible to the human eye. This allowed for the identification of high-risk patients.
Based on the level of inflammation in heart arteries and the patient’s health history, including factors like age, sex, smoking habits, diabetes, and high blood pressure, an AI risk-score was given to predict the likelihood of future cardiac events. This risk-score accurately predicted the occurrence of such events over a 10-year period.
The integration of AI technology in four NHS hospitals resulted in doctors modifying treatment plans for approximately 45% of patients based on the AI-generated risk scores. This approach could lead to more timely preventive treatments and reduce the number of heart attacks and cardiac deaths.
The use of AI in predicting cardiac risks and guiding treatment decisions was emphasized as an important tool in preventing unnecessary heart-related deaths. The hope is to implement this technology on a larger scale within the NHS to aid individuals at risk of heart disease.
Current pilot programs are testing the integration of AI tools with existing NHS care practices. The findings of the study, titled “Inflammatory risk and cardiovascular events in patients without obstructive coronary artery disease: The Orfan multicentre, longitudinal cohort study,” were published in The Lancet last week.