Experts from Canadian technology company Klick Labs have developed a tool based on artificial intelligence (AI) that can detect the presence of type 2 diabetes based on a person’s voice. The study was published in the scientific journal magazine Proceedings of the Mayo Clinic: Digital Health (MCPDH).
According to the creators of the artificial intelligence, to make a diagnosis, the program only needs a recording of a conversation lasting six to 10 seconds, as well as basic data on health status, such as gender, age, height and weight. The accuracy of the model is 89% for women and 86% for men.
As part of the study, experts selected 267 volunteers with both type 2 diabetes and healthy volunteers. They were asked to record sentences on their smartphones six times a day for two weeks. Scientists then analyzed more than 18 thousand audio recordings and identified 14 acoustic features that indicate the presence of the disease.
“Our study reveals significant differences in voice between people with and without type 2 diabetes and may change the way the medical community diagnoses the disease,” said Klick Labs researcher and lead author Jacy Kaufman.
According to the International Diabetes Federation, almost one in two of the 480 million people with diabetes worldwide do not know they have diabetes. However, approximately 90% of diabetes cases are type 2 diabetes.
Previously AI taught Diagnosis of polycystic ovary syndrome.