New AI device detects high blood pressure — a 'silent killer' — through voice recordings

On a black screen, a white sound wave depicts a vocal recording.

A disease dubbed “the silent killer” by the World Health Organization may now be detected using the sound of a patient’s voice.

Researchers at Klick Labs have announced their latest work using artificial intelligence to predict chronic high blood pressure — or hypertension — simply by using voice recordings.

A white woman with shoulder-length blond hair wears a black shirt and a lab coat, while she types on a laptop next to a microscope. Beside her is a white man with short brown hair, also wearing a black shirt and a white lab coat, holding a smartphone.
Photo courtesy of Klick Labs

This technology could not only revolutionize blood pressure screenings (like using an arm cuff at your doctor’s office) but could detect hypertension sooner, helping people get diagnosed and treated more quickly, and therefore reducing risks of heart attack and stroke.

But… how does it even work?

Klick researchers created an app that uses machine learning to analyze hundreds of vocal biomarkers, or unique characteristics in one’s voice that are “indiscernible to the human ear.” 

These biomarkers included things like variability in pitch, patterns in speech and energy distribution, and the sharpness of sound changes. 

In a study recently published in the peer-reviewed journal IEEE Access, 245 participants were asked to record their voices up to six times daily for two weeks, allowing the app to analyze a variety of data samples.

Three scientists stand smiling, in white lab coats. On the left, a man with a fedora. In the center, a woman with blond hair pulled back, and on the right, a man with dark hair and a dark beard.
The study's researchers. From left to right: Yan Fossat, Jaycee Kaufman, and Behrad Taghibeyglou. Photo courtesy of Klick Labs

The app detected high blood pressure with accuracies up to 84% for female participants and 77% for males.

“By leveraging various classifiers and establishing gender-based predictive models, we discovered a more accessible way to detect hypertension, which we hope will lead to earlier intervention for this widespread global health issue,” Yan Fossat — senior vice president of Klick Labs and principal investigator of the study — said in a statement

“Hypertension can lead to a number of complications, from heart attacks and kidney problems to dementia,” he added.

Klick Labs has also previously researched the use of vocal biomarkers to screen for Type 2 diabetes, and another recent study from the same team confirms a relationship between blood glucose levels and voice pitch.

So, while the idea of speaking into an app and getting back valuable health insights might sound like a plot from a science fiction drama, scientists are clear: This technology is not only possible — it’s here. 

“Voice technology has the potential to exponentially transform healthcare, making it more accessible and affordable,” co-author and Klick Labs researcher Jaycee Kaufman said, “especially for large, underserved populations.”

It’s true; hypertension is especially prevalent in both Black and rural communities, both of which see glaring healthcare disparities, compared to white, urban-dwelling demographics

An easy-to-use mobile app presents a promising solution to help bridge those gaps, offering a way to screen for hypertension that is more accessible than traditional means.

“Conventional methods of measuring blood pressure (and, accordingly, identifying hypertension) include using an arm cuff or an automatic blood pressure measurement device,” a press release from Klick Labs explained.

A close-up of a blood pressure cuff on a patient's arm, with a nurse watching the results.
Photo by Photo by Julia Taubitz on Unsplash

“However, these methods may require technical expertise, specialized equipment, and may not be readily accessible to people in underserved areas.”

Though this study marks an exciting step into using voice technology in medicine, scientists are still working on refining and improving  their technology before it becomes widely available. 

This latest breakthrough marks Klick Labs’ first venture into using AI to identify conditions beyond diabetes. The company said it will continue to expand its research “to assess its AI algorithms’ effectiveness in detecting and managing a broader range of health conditions.”

Building on collaborations with hospitals, academic institutions, and public health authorities worldwide, the Klick Labs scientists are eager to see — or hear — what comes next.

Kaufman said, “Our ongoing research increasingly demonstrates the significant promise of vocal biomarkers in detecting hypertension, diabetes, and a growing list of other health conditions.”

Header image courtesy of Egor Komarov/Pexels

Article Details

September 11, 2024 10:37 AM
A cut-out of the state of Florida sits atop a grayscale image of a hurricane. On the lower, southern corner of the state sits a red heart.

Hurricane Milton relief: How to help, donate & volunteer

In the aftermath of Hurricane Milton, here’s how you can help survivors in Florida.
Mental Health Day

21 Ways To Take Action & Celebrate World Mental Health Day

World Mental Health Day is a catalyst for us to learn more, seek help, offer support, and open our hearts and minds to vulnerable conversations.
No items found.

Want to stay up-to-date on positive news?

The best email in your inbox.
Filled with the day’s best good news.