In an era of instant messaging and voice-activated devices, what if our very voices held the key to identifying a common yet often elusive health concern? Diabetes, a growing global concern, typically requires a series of diagnostic tests and doctor’s appointments. But now, an innovative study by software solution pioneers, Klick Labs, reveals the potential for a more straightforward, faster, and remarkably accurate method to detect Type 2 diabetes.
By harnessing the power of voice and artificial intelligence (AI), this groundbreaking approach allows individuals to use their smartphones to uncover whether they may have this more treatable form of diabetes, promising a revolutionary shift in healthcare screening.
Klick Labs has embarked on a transformative journey, fusing voice technology and AI to redefine how we detect diabetes. By just speaking into a smartphone, individuals can potentially unlock valuable insights into their health. This new approach considers basic information like age, gender, height, weight, and a mere six to ten seconds of an individual’s spoken words. Together, these elements form an AI model capable of pinpointing the presence of Type 2 diabetes.
Type 2 diabetes, the more manageable form, can be initially addressed with lifestyle changes, oral medications, or insulin, depending on its severity. As per the study, the accuracy of this novel method reaches 89 percent for women and 86 percent for men, promising a quick, accessible, and cost-effective solution.
The research involved 267 individuals with diabetes who were asked to record a sentence on their smartphones six times a day for two weeks. From the sea of over 18,000 recordings, scientists extracted 14 auditory parameters, meticulously analyzing them to detect variations that distinguished those with Type 2 diabetes from those without.
Jaycee Kaufman, the paper’s first author and a research scientist at Klick Labs, underscores the groundbreaking nature of their work, stating, “Our research illuminates significant vocal disparities between individuals with and without Type 2 diabetes, offering the potential to revolutionize diabetes screening within the medical community. The current methods often entail extensive time, travel, and expenses, whereas voice technology can potentially eliminate these barriers.”
Klick Labs employed AI to scrutinize diverse voice characteristics, such as variations in pitch and intensity, which often escape the human ear’s perception. Using sophisticated signal processing techniques, researchers could discern the subtle vocal changes brought on by Type 2 diabetes.
Yan Fossat, Klick Labs’ vice president and principal investigator of this pioneering study, reaffirms the enormous potential of voice technology within healthcare. He suggests, “Our research underscores the tremendous potential of voice technology not only in detecting Type 2 diabetes but also in identifying other health conditions. It offers an accessible and cost-effective digital screening tool that could revolutionize healthcare practices.”
This non-intrusive, swift approach developed by Klick Labs has the potential to screen a substantial number of individuals, aiding in the early identification of Type 2 diabetes cases that often go unnoticed. Fossat envisions future steps, including replicating the study and expanding research into additional health conditions. Conditions like hypertension, women’s health, and prediabetes may all be diagnosed using voice and AI technology, ushering in a new era of healthcare screening that holds great promise.
