Read more

November 17, 2020
2 min read
Save

AI program uses vocal biomarkers to diagnose COVID-19

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

An artificial intelligence voice analysis tool can help diagnose COVID-19 in asymptomatic patients, according to its manufacturer, Vocalis Health.

The technology — called VocalisCheck — works by comparing a person’s voice sample to a COVID-19-positive voice composite. VocalisCheck assesses their risk level of testing positive for COVID-19 and whether they require further testing.

The quote is: “The cell phone has expanded exponentially in its capabilities over the last decade." The source of the quote is: Peter Katona, MD, FACP, FIDSA

According to the company, early study results show that VocalisCheck had a sensitivity of 87% and specificity of of 53%, when used alone, adding “even better” results were achieved when combined with a symptom questionnaire.

“Over time, we will collect more and more data, which can strengthen the AI and make the vocal biomarker even more accurate,” Shady Hassan, MD, co‐founder, chief medical and chief operating officer of Vocalis Health, told Healio Primary Care.

Shady Hassan

Hassan stressed that VocalisCheck is intended to complement, not replace, physician-patient interactions.

“Whether it is monitoring patient symptoms between visits, supporting physicians’ decisions while diagnosing health conditions remotely, providing initial screening to a very large population, or detecting subtle changes not discernable to the human ear, we seek to help physicians and their patients and enhance the quality of care,” he said.

Meanwhile, scientists at the Massachusetts Institute of Technology (MIT) developed an AI algorithm to differentiate patients with possible COVID-19 using forced cough recordings. The technology had been tested in 1,064 patients, demonstrating a sensitivity of 98.5% and a specificity of 94.2%.

The MIT researchers wrote in IEE Open Journal of Engineering in Medicine and Biology that AI technology can be used to “augment current approaches in containing the spread of COVID-19” — for example, daily screening of students, workers and the public.

Peter Katona, MD, FACP, FIDSA, clinical professor of medicine at the David Geffen School of Medicine at the University of California, Los Angeles, told Healio Primary Care that with improved sensitivity and specificity, tools like VocalisCheck are on the cusp of becoming mainstream in medicine someday.

“The cell phone has expanded exponentially in its capabilities over the last decade,” he said, adding that VocalisCheck “is a brilliant way to advance our ability to test people efficiently.”

Katona, who is uninvolved with Vocalis Health, also said the AI program may make it easier to for people to get tested for COVID-19.

“Here at [UCLA], some have objected to forced COVID-19 testing,” he said. “But if you can do the test with a cell phone, the legal objections may not be sustainable. I'm very optimistic about advancing this kind of technology.”

References: