Deep learning models applied to existing clinical data from primary and secondary care settings can detect young-onset type 2 diabetes.

Study: Opportunistic screening of type 2 diabetes with deep metric learning using electronic health records

This study shows that a computer program can look at people’s regular medical records and predict who will get type 2 diabetes years before they’re diagnosed — even when they’re young. It works better than older methods, so yes, it supports the idea that computers can spot this disease early using existing doctor’s notes.

Study: Interpretable type 2 diabetes incidence prediction with AutoScore: A model based on standard clinical parameters

This study shows that computers can use routine blood tests and medical records to predict who might get type 2 diabetes at a young age—even with simple tools. It proves the idea works, even if the exact computer method wasn't deep learning.

Study: Detection of young-onset type 2 diabetes using deep learning across primary and secondary care: a nationwide, retrospective cohort study.

Scientists used computers to analyze people's old medical records and found they could predict who would get type 2 diabetes before they were diagnosed—especially in younger people. The computer was very good at spotting the highest-risk cases.

Study: Predicting and classifying type 2 diabetes using a transparent ensemble model combining random forest, k-nearest neighbor, and neural networks

This study showed that a computer program using common medical data like blood sugar and weight can spot type 2 diabetes with perfect accuracy. That means it’s possible to use routine doctor’s office records to find the disease early.