Researchers at Stanford are using artificial intelligence (AI) to get deeper research into diabetes diagnosis and the results could mean better, more accessible care.
AI has helped to Identify patient’s subtype which can significantly influence the effectiveness of treatment. This new move enables doctors to create personalised treatment plans and allocate resources more effectively, ultimately reducing costs.
The study uses Artificial intelligence to analyse data that is already collected by glucose monitors, meaning the algorithm does not require a more extensive or complex clinical setting to function.
Diabetes is typically categorised as either Type 1 or Type 2. However, in recent years, scientists have identified significant variations, or subtypes, within Type 2 diabetes which accounts for 95% of diagnoses; shedding light on associated risks such as kidney, heart, or liver conditions.
A professor of endocrinology at Stanford, Tracey McLaughlin, shed light on the challenges of understanding the physiology behind diabetes. McLaughlin said the tests are cumbersome, expensive and not practical for clinical uses.
Using data gathered from glucose monitors, researchers developed an algorithm capable of identifying three of the four most common subtypes of Type 2 diabetes.
When compared to clinical data, the algorithm was shown to predict metabolic subtypes, such as insulin resistance and beta-cell deficiency, with greater accuracy than traditional metabolic tests – achieving roughly 90% accuracy.
Researchers are optimistic that the algorithm will make health information more accessible at home, particularly for individuals who lack access to healthcare services due to geographical, financial, or other barriers.
Nearly 13% of the US population has been diagnosed with diabetes, these detailed insights could have a profound impact on treatment options and outcomes – especially since AI can derive valuable information from wearable devices that patients often already own and rely upon.
Following CES 2025, where two over-the-counter glucose monitors were recognised as honourees in Digital Health, this study represents another significant step towards accessible healthcare technology.
