Top Stories

AHN, local startup to study voice patterns of people with COVID-19

PITTSBURGH — Does a person who has COVID-19 have a different voice pattern than someone who does not?

This is the question doctors and researchers with Allegheny Health Network are asking.

They’ve partnered with a local startup company,, to study the voices of those infected by the coronavirus.

>>RELATED STORY: LIVE UPDATES: Coronavirus in Pennsylvania: What you need to know Thursday

Many patients who have contracted the coronavirus say they have experienced severe respiratory problems.

Experts hope voice analysis will help doctors determine when patients need treatment and when they should stay home.

AHN patients who get a doctor-ordered COVID-19 test can become eligible for voice monitoring.

They record their voice on the website through their smartphone.

Have questions about the spread of the coronavirus? We have an entire section dedicated to coverage of the outbreak. CLICK HERE for more.

If they test positive for COVID-19, they continue to provide voice samples for 30 days. then analyzes the way they breathe and speak.

Through signal processing technology, the app can monitor respiratory infections and chronic lung diseases.

AHN says the voice study started before the coronavirus pandemic but it became even more important.

“This will allow us to learn in this kind of new innovative technology that is very simple. It’s not gonna hurt anybody. I’m not taking blood from anybody. I’m just taking your voice. And the ability of us taking that data analyzing, using and learning from it really has international implications,” Dr. Anil Singh, system director of pulmonology at AHN said.

“If we can rule out the normals. If we can help identify the high-risk people. We can get them to contract their provider we can reduce the overwhelming burden on the health care system,” Satya Venneti, co-founder and chief technical officer said.

If the research is promising, AHN and the doctor will design a clinical trial for FDA testing.

Comments on this article