Scientists at the Radcliffe Department of Medicine, University of Oxford, have developed an Artificial Intelligence (AI) technology capable of predicting the risk of heart problems up to ten years in advance using routine cardiac CT scans. 

Funded by the British Heart Foundation and supported by the National Institute for Health and Care Research Oxford Biomedical Research Centre, this research holds significant promise for the early detection and prevention of heart disease, a leading cause of morbidity and mortality worldwide.

Study Design and Participants

The study analyzed routine cardiac CT scans and clinical data from over 40,000 individuals across eight hospitals in the UK. Participants were followed for up to ten years to monitor the occurrence of major cardiac events, including heart attacks, heart failure, and cardiac death. In addition, this extensive follow-up period allowed researchers to gather robust data on the long-term outcomes of patients based on their initial CT scan results.

Key Findings

Predictive Accuracy of AI: AI technology has demonstrated remarkable accuracy in predicting the risk of future cardiac events. While individuals with significant coronary artery narrowing on their initial CT scan were more likely to experience severe cardiac events, two-thirds of these events occurred in patients without substantial artery narrowing. This finding underscores the potential of AI to identify at-risk individuals who might be overlooked using traditional diagnostic methods.

AI Risk Score: The researchers developed an AI-generated risk score that combines the level of inflammation in the heart arteries, as seen on the CT scan, with the patient's health background, including age, sex, smoking status, diabetes, and high blood pressure. This composite score predicted the likelihood of cardiac events over the next ten years and provided a timeline for when these events were likely to occur. The AI predictions closely matched the actual cardiac events experienced by the patients during the study.

Clinical Impact and Pilot Study

Integrating AI-generated risk scores into clinical practice has already shown promising results. In a pilot study involving four NHS hospitals, doctors received AI risk scores for 744 patients. As a result, treatment plans were altered in up to 45% of cases, highlighting the potential of AI to influence clinical decision-making and improve patient outcomes.

Professor David Adlam from the University of Leicester, a co-investigator in the study, emphasized the importance of providing clinicians and patients with a more precise picture of cardiac risk. By identifying high-risk individuals early, preventive treatments can be administered promptly, potentially reducing the incidence of heart attacks and cardiac deaths.

Expert Opinions and Future Prospects

Bryan Williams, Chief Scientific and Medical Officer at the British Heart Foundation, praised the new AI tool and described it as a valuable asset for clinicians. The ability to foresee the future risk of heart attacks, heart failure, and cardiac death can guide treatment decisions and interventions, ultimately helping patients avoid devastating heart-related conditions.

The researchers hope to see this AI technology widely adopted across the NHS to support at-risk patients. The ongoing pilot in several UK hospitals aims to refine the integration of AI tools with current NHS care practices, ensuring that the technology complements existing diagnostic and treatment protocols.

Conclusion

The development of AI technology that can predict heart problems ten years in advance from routine cardiac CT scans represents a significant advancement in cardiovascular medicine. By accurately identifying individuals at high risk of serious cardiac events, this technology offers the potential for early intervention and personalized treatment strategies. The positive impact observed in pilot studies underscores the importance of harnessing AI to improve patient outcomes and prevent unnecessary heart-related deaths. As the technology continues to be refined and integrated into clinical practice, it promises to transform cardiovascular care and save countless lives.

Source: The Lancet

Image source: COPILOT