Based on sounds heard through a stethoscope during an examination, doctors in Wuhan, Hubei province say they can tell, with the help of a computer, which patients are infected with the coronavirus.

Doctors at Tongji Hospital affiliated with Tongji Medical College of Huazhong University of Science and Technology in Wuhan teamed up with professors of information engineering to create an artificial intelligence model that can identify infected patients, the hospital said.

The model distinguishes COVID-19 patients from others with more than 95 percent accuracy, it said.

The medical team led by Zeng Hesong of the hospital captured sounds from the heart and lungs of 172 COVID-19 patients from April 1 to 5 last year. They uploaded the data into a database, along with similar information collected from 50 other patients who had other diseases. Then comparisons were made.

The innovation will be helpful, especially given the infectious environment because it's easy to collect audio data and then upload it into the AI model. The process goes fast and it doesn't require large medical devices. No samples are needed, Zeng said.