In an astounding revelation, a recent study by Klick Labs suggests that screening for Type 2 diabetes may soon become as simple as speaking into a smartphone for 10 seconds. This innovative approach harnesses the power of artificial intelligence (AI) and voice technology to detect subtle vocal variations associated with the disease. The study, featured in the October issue of Mayo Clinic Proceedings: Digital Health, could mark a significant step forward in diabetes detection.
Researchers at KlickLabs embarked on this pioneering endeavor by enlisting 267 participants, both with and without a Type 2 diabetes diagnosis. Over two weeks, these individuals were requested to record a specific phrase using their smartphones six times daily. Klick made this announcement live on their company’s LinkedIn profile, surprising its followers with the ground-breaking nature of this study.
The recorded voices, coupled with primary health data such as age, sex, height, and weight, were then fed into an AI model. This model demonstrated an impressive 89 percent accuracy rate in identifying women with Type 2 diabetes and 86 percent for men.
The analysis delved into a range of vocal features, including imperceptible variations in pitch and intensity. These subtle changes could be detected through signal processing, indicating the presence of Type 2 diabetes. Interestingly, the vocal variations manifested differently in males and females. This marks yet another impressive use case of AI in healthcare development.
According to the International Diabetes Federation, nearly 240 million adults worldwide remain unaware of their diabetes status, with approximately 9 percent of cases being Type 2 diabetes. Current diagnostic tests typically require visits to healthcare providers, such as the glycated hemoglobin (A1C) test, fasting blood glucose (FBG) test, and oral glucose tolerance test (OGTT).
Yan Fossat, Vice President of Klick Labs and principal investigator of the study, emphasized the potential of this non-intrusive and accessible approach to screen many people and potentially identify undiagnosed cases of Type 2 diabetes. Fossat believes that voice technology has the potential to revolutionize healthcare practices as an accessible and cost-effective digital screening tool.
The researchers at Klick Labs plan to replicate their study and expand their exploration of voice as a diagnostic tool in various health areas, including prediabetes, women’s health, and hypertension.
This groundbreaking discovery is a testament to Klick Labs’ dedication and expertise in machine learning, data science, and artificial intelligence, spanning over a decade of research in various therapeutic areas, including diabetes.
