The science behind Virufy

Artificial intelligence (AI) is capable of detecting minute differences in vocal patterns to detect disease presence.

The science behind Virufy

Artificial intelligence (AI) is capable of detecting minute differences in vocal patterns to detect disease presence.

Our Technology

Disease analysis with AI

Our AI

The clinical presentation of COVID-19 can be highly variable, however, a dry cough is a distinctive feature of most cases. Because COVID-19 creates a unique respiratory signature in the throat and lungs, distinct from respiratory infections that produce a wet cough, it has been suggested that cough sounds can be analyzed to detect COVID-19. Though the differences may be subtle to the human ear, AI algorithms have shown promise in classifying cough sounds to identify various respiratory diseases including pertussis, asthma, and pneumonia by using various features extracted from audio waveforms, including MFCC coefficient, pitch, spectral energy, and zero crossing. Although the cough samples from various diseases sound similar, they each have a distinguishable pattern due to small differences in audio features, which AI algorithms are able to pick up on. Beyond COVID, it has been demonstrated that AI can classify certain diagnostic breath sounds, such as wheezing and rhonchi, from smartphone audio recordings.

Our papers

Read The Latest Research

Want to learn more?