Artificial intelligence could “revolutionize” disease diagnosis after researchers managed to develop a tool that requires just a few seconds of a patient’s voice recording to detect type 2 diabetes.
“Our research reveals significant voice differences between people with and without type 2 diabetes and could change the way the medical community conducts diabetes testing,” study lead author Jaycee Kaufman said, according to the New York Post on Wednesday.
To conduct the research, the Klick Labs team asked 267 people with and without diabetes to record themselves six times a day for two weeks to analyze 14 characteristics of different acoustics using a database of more than 18,000 extracts, according to the American newspaper.
Artificial intelligence would thus have been able to recognize certain vocal notes that are imperceptible to the human ear, which could distinguish a patient with type 2 diabetes with an accuracy of 89% in women and 86% in men, it says in a statement.
And this method could potentially be applied to other conditions such as high blood pressure, enabling faster and cheaper diagnosis using an “accessible and affordable” digital screening tool, the researcher said.
“Current detection methods can require significant time, travel and money. Voice technology has the potential to completely remove these barriers [et] could revolutionize healthcare practices,” he added, according to the NY Post.
The lab is already planning to expand its research into prediabetes before type 2 diabetes and high blood pressure develop, he said.