A recent medical study conducted by Klick Labs and published in "Mayo Clinic Proceedings: Digital Health" has discovered that artificial intelligence has the ability to detect type 2 diabetes simply by analyzing a patient's voice for as short as six to ten seconds. The research showed high accuracy rates of 89% for women and 86% for men in diagnosing the condition. According to first author Jaycee Kaufman, the study revealed significant vocal differences between individuals with and without type 2 diabetes, potentially revolutionizing how diabetes screenings are conducted within the medical field. Kaufman also emphasized that voice technology has the potential to eliminate the time, travel, and cost constraints associated with current detection methods. In the study, 267 participants, both with and without type 2 diabetes, recorded phrases on their smartphones six times a day for two weeks. Over 18, 000 recordings were analyzed for more than 14 different acoustic features, which were found to be distinct between diabetics and non-diabetics.

Additionally, participants provided basic health information such as age, height, and weight. Signal processing technologies were able to detect specific vocal pitch characteristics that humans cannot perceive, providing important indicators for diabetes diagnosis. Klick VP and principal investigator Yan Fossat emphasized the vast potential of voice technology in identifying various health conditions, including type 2 diabetes. Fossat believes that voice technology has the capability to revolutionize healthcare practices as an accessible and affordable digital screening tool. The next steps for Klick Labs include replicating the study and broadening the applications of vocal analysis to detect pre-diabetes, hypertension, and other health conditions. This groundbreaking research follows a recent breakthrough at MIT involving a bio-implant that can seamlessly adapt to the body, facilitating the administration of medications such as insulin.