It’s incredible how Artificial Intelligence (AI) technology has done great things in our modern technology, especially as we can ask questions or give commands, and they can find answers or control an entire home.

Another study discovered that AI can now detect one of the most concerning and yet common medical conditions in the world: diabetes. 

Medical research unveiled that AI can distinguish type 2 diabetes by listening to patients speak for at least six to 10 seconds. The study by Klick Labs and published in Mayo Clinic Proceedings: Digital Health proved the technology’s accuracy at 89% for diagnosing women and 86% among men. 

According to the research, AI highlights important vocal variations between people with or without type 2 diabetes. Further examination into this capability will be significant when screening patients with possible progressing diabetes.

AI-Detect-Diabetes-Health-Conditions

The researchers further investigated by asking 267 participants to record phrases on their mobile devices at least six times a day for a period of two weeks.

Over 18,000 recordings were collected and analyzed for more than 14 various acoustic features, which vary between people with detected diabetes and those who don’t. Participants were also asked to provide basic health information, like age, weight, and height. 

According to one of the authors, Jaycee Kaufman, signal-processing technologies can detect certain notes of vocal pitch that our human ears cannot do. These particular sounds help pinpoint clues regarding a person’s condition. 

It is also true that voice technology could help the medical community detect type 2 diabetes and other medical conditions faster and cost-efficiently than the tools they have now. 

Klick Labs hopes to further its research by replicating this study toward a possible detection of pre-diabetes and other health concerns. 

Do you agree that AI can be involved even with medical practices in the future? Let us know your thoughts.

Leave a comment

Your email address will not be published. Required fields are marked *