AI could predict type 2 diabetes a decade in advance
New research funded uses AI to analyse ECG readings during routine heart scans and predict risk of developing the disease up to 10 years before people actually do!
Some 5.6m people in the UK currently live with diabetes, with as many as 1.2m thought to have the condition but are not yet diagnosed. Early detection is vital in reducing risks to health from the disease and its associated complications – which can include problems with eyes, heart and feet.
Photo by Gia Willow Alexa Annermarken
Up until now, type 2 diabetes and ‘prediabetes’ have been relatively simple to diagnose with a blood test to assess sugar levels. Now researchers suggest that artificial intelligence (AI) can be used to analyse routine ECG readings and predict a person’s likelihood of developing diabetes.
The research is funded by the British Heart Foundation (BHF) and NIHR Imperial Biomedical Research Centre, a translational research partnership between Imperial College Healthcare NHS Trust and Imperial College London. Findings were presented this week to the American Heart Association’s Scientific Sessions 2024, held in Chicago.
The research is being led by Dr Fu Siong Ng and Dr Arunashis Sau at Imperial College London. They’ve developed the AI-ECG Risk Estimation for Diabetes Mellitus (AIRE-DM) tool, using some 1.2m ECGs from hospital records. With data from the UK Biobank, they then validated the AI’s ability to detect subtle changes in routine ECGs that could signify someone being at higher risk of type 2 diabetes.
In some 70% of cases, AIRE-DM accurately predicted future risk in people of various ages, genders, ethnicities and socioeconomic backgrounds. What’s more, the team achieved greater accuracy when they combined these AI predictions with genetic and clinical information, such as age and blood pressure.
The result is, they say, an even clearer picture of risk which could help to spot people who might otherwise not be identified as likely to develop the condition.
In addition, AIRE-DM could accurately assess data collected from single-lead ECGs, which are often used in wearable devices. That means there is potential to apply this technology in smart devices.
A fuller pilot study will be conducted during 2025 and the researchers are optimistic that it will be available to the NHS in the next few years.
Professor Bryan Williams, Chief Scientific and Medical Officer at the BHF, says: ‘This exciting research uses powerful artificial intelligence to analyse ECGs, revealing how AI can spot things that cannot usually be observed in routinely collected health data. This kind of insight could be a gamechanger in predicting future risk of developing type 2 diabetes, years before the condition begins.
‘Type 2 diabetes is a rapidly growing health challenge that increases the risk of developing heart disease, however with the right support it is possible for people to reduce their risk of developing the condition. We look forward to seeing how this technology could be incorporated into clinical practice, providing an opportunity to intervene early to help reduce risk and even prevent type 2 diabetes and its associated complications, altogether.’
In related news:
Parents of autistic children face blame when seeking support – report
