Artificial Intelligence Detects Diabetes in 10 Seconds

MetaversePlanet

October 20, 2023

A recent groundbreaking study has shown that artificial intelligence (AI) can detect type 2 diabetes in just a few seconds of conversation.

Artificial intelligence stands as a pivotal innovation of our era. Its application in medical research holds the potential for transformative breakthroughs.

This innovative study reveals that AI can determine the existence of diabetes from just 10 seconds of voice interaction, demonstrating an impressive accuracy rate of 89 percent for women and 86 percent for men.

Jaycee Kaufman, the lead author of the study, stated, “Our study highlights unique vocal signatures in individuals with and without Type 2 diabetes, potentially transforming the way we conduct diabetes screening.” Conventional methods of detection can be laborious and costly. Kaufman suggests that voice technology could revolutionize this process by providing a quicker and more economical solution.

In this study, 267 participants, some with type 2 diabetes and others without, were asked to record voice notes on their mobile devices several times a day for two weeks. From the more than 18,000 recordings gathered, the research identified 14 distinct acoustic features that differed between the two groups.

Thanks to sophisticated signal processing techniques, the technology was able to identify specific pitches that are imperceptible to the human ear. These subtle vocal tones were key indicators, as explained by Kaufman.

Emphasizing the significance and potential impact of these findings, lead researcher Yan Fossat remarked, “This research highlights the enormous potential of voice technology in detecting Type 2 diabetes and possibly other medical conditions. Voice-driven technology could soon become a pivotal tool in healthcare, providing an accessible and cost-effective digital diagnostic method.”

You may also like this content