In a groundbreaking study, researchers from Klick Labs in the US have developed an AI model that can identify Type 2 diabetes in individuals with 89% accuracy for women and 86% accuracy for men. The study combines voice technology with artificial intelligence, making it possible to determine if someone has diabetes by analyzing just six to ten seconds of their voice and basic health data, including age, sex, height, and weight.
For the study, 267 participants, both non-diabetic and Type 2 diabetic, recorded a specific phrase into their smartphones six times daily for two weeks. Analysis of over 18,000 recordings revealed 14 acoustic features that distinguished between non-diabetic and Type 2 diabetic individuals.
Jaycee Kaufman, a research scientist at Klick Labs, emphasized the significance of vocal variations in diabetes detection. Current detection methods can be time-consuming and costly, but voice technology offers a non-intrusive and accessible solution that could potentially transform diabetes screening.
The study explored various vocal features, including subtle changes in pitch and intensity that may not be discernible to the human ear. Interestingly, vocal changes manifested differently for males and females.
With nearly 90% of diabetic cases being Type 2 diabetes, and a large percentage of individuals with diabetes remaining undiagnosed, this innovative approach using voice technology offers the potential to screen vast numbers of people and identify those with Type 2 diabetes.
Yan Fossat, Vice President of Klick Labs, emphasized the transformative potential of voice technology in healthcare and its accessibility as a digital screening tool.
The most commonly used diagnostic tests for prediabetes and Type 2 diabetes, such as the glycated haemoglobin (A1C), fasting blood glucose (FBG), and oral glucose tolerance test (OGTT), could see significant improvements with the adoption of this new approach.
