Two NHS hospital trusts in London have begun trialing an innovative AI technology to predict the risk of type 2 diabetes in patients up to a decade before the condition develops, BBC reports.
The AI system, named Aire-DM, is being tested at Imperial College Healthcare and Chelsea and Westminster Hospital NHS Foundation Trusts. The system analyzes patients’ ECG heart traces to detect early warning signs of the condition that may be difficult for doctors to identify.
Clinical trials for the AI technology are scheduled for 2025, with early results suggesting the system can detect risk about 70% of the time. The system’s accuracy improves when additional patient information, such as age, sex, and the presence of risk factors like high blood pressure or obesity, is incorporated. Dr. Fu Siong Ng, the lead researcher, explained that while the ECG data alone yields promising results, adding more background details enhances the predictive capability of the AI.
ECGs, or electrocardiograms, record the heart’s electrical activity, including its rate and rhythm. According to Dr. Ng, the subtle changes in the ECG that the AI system identifies are challenging for even experienced doctors to interpret. The AI can detect patterns across various parts of the ECG trace, making it a valuable tool in identifying potential health risks that might otherwise go unnoticed.
The trial will involve up to 1,000 patients at both hospitals, with the AI system reading their ECG scans to see how effectively it predicts the risk of developing type 2 diabetes. Though the technology is not yet being used routinely, researchers are optimistic that it could be integrated into the NHS in the future, potentially within five years or more.
The British Heart Foundation (BHF), which is funding the project, believes that early identification of individuals at risk of type 2 diabetes could ultimately save lives. Uncontrolled type 2 diabetes can lead to serious complications, including heart attacks and strokes. Preventative measures such as maintaining a healthy weight, eating a balanced diet, and exercising are essential in reducing these risks.
Professor Bryan Williams, Chief Scientific and Medical Officer at the BHF, emphasized the transformative potential of AI in healthcare. He highlighted that AI could uncover hidden insights in ECG data, offering a new way to predict the future risk of type 2 diabetes. As type 2 diabetes continues to rise globally, early detection and intervention become even more critical in managing the condition.
Dr. Faye Riley from Diabetes UK also voiced her support for the AI-powered screening method. She noted that type 2 diabetes is often undiagnosed for years, and early identification could help prevent serious complications like heart failure and vision loss. With millions at risk of developing the condition, AI screening could provide a valuable tool for improving early intervention.
Type 2 diabetes is a condition where the body cannot properly regulate blood sugar due to insufficient or ineffective insulin production. It is often linked to being overweight, as excess fat can interfere with the pancreas’ ability to produce insulin. Early detection and management can significantly reduce the risk of complications and improve overall health outcomes for those at risk.
