Recently, an international team of researchers from China and Singapore successfully developed the world's first multi-modal artificial intelligence system for diabetes diagnosis and treatment, DeepDR-LLM. The research findings were published in Nature Medicine on July 19, 2024.

Researchers at Klick in Canada have released an AI large model that can diagnose type 2 diabetes in just 6 to 10 seconds using voice. Validation with 267 volunteer voice samples showed an accuracy rate of 89% for females and 86% for males. The study identified 'pitch' and 'pitch standard deviation' as key diagnostic features. Various models, including logistic regression and naive Bayes, were used for training and testing. The integrated module combines multiple information sources to improve model accuracy and offers potential for early intervention.