AI model detects type 2 diabetes through voice analysis

Scientists from Thymia, in collaboration with RMIT University, have developed an artificial intelligence model capable of detecting type 2 diabetes through voice analysis. The research, presented at the European Association for the Study of Diabetes congress, focuses on identifying speech changes such as hoarseness or reduced breath control associated with the condition.

The AI was trained on over 63,000 vocal samples from more than 21,000 individuals in the United Kingdom and the United States. In testing involving 7,319 adults in the UK, the model successfully assigned higher risk scores to 80% of those who had previously reported a type 2 diabetes diagnosis. However, researchers noted that the model was less effective for Black individuals due to underrepresentation in the training data.