Two London NHS trusts are testing artificial intelligence (AI) to determine if it can predict type 2 diabetes up to 10 years before symptoms develop. Imperial College Healthcare and Chelsea and Westminster Hospital NHS Foundation Trusts are training an AI system, Aire-DM, to analyse ECG heart traces for early signs that are often too subtle for doctors to detect.
Clinical trials, scheduled for 2025, will involve up to 1,000 patients. Preliminary findings suggest Aire-DM can accurately identify diabetes risk in about 70% of cases. The system’s predictive accuracy improves further when additional data, such as age, sex, blood pressure, and weight, are factored in, says lead researcher Dr Fu Siong Ng. The AI analyses complex ECG changes that are undetectable to the human eye, identifying patterns that could indicate future risk.
If successful, the technology could eventually be adopted across the NHS, though widespread use may take five years or more. The British Heart Foundation, funding the research, believes this innovation could save lives by identifying those at risk and preventing complications like heart disease and strokes.
Type 2 diabetes, often linked to being overweight, is a growing health concern. Experts say early detection could transform care by allowing individuals to adopt lifestyle changes, such as healthier eating and regular exercise, to reduce their risk and improve long-term outcomes.
