A team of researchers from IST (Instituto Superior Técnico – Universidade de Lisboa) and INESC-ID (Instituto de Engenharia de Sistemas e Computadores – Investigação e Desenvolvimento), coordinated by Isabel Trancoso, is developing a new project to explore the possibility of automatically detecting COVID-19 from its effects on cough and speech.

This project would allow developing a cheap, fast and easy to use artificial intelligence-based tool (deployed as a web platform and/or a mobile application) that could provide a preliminary assessment of potential infection by COVID-19. Although not a clinical diagnosis, this is valuable information that would help individuals to adopt preventive measures, and public and private healthcare operators, institutions, companies, etc. to optimize their screening campaigns by allowing them to focus their attention on suspected infected individuals.

For this goal, it is fundamental to collect an extensive dataset with representative examples of speech and simulated coughs and snores from both COVID-19 positive (symptomatic and asymptomatic) and negative individuals (ideally including also participants with respiratory conditions other than COVID-19, such as flu, cold, asthma, etc.). Then, signal processing and machine learning techniques will be used to assess the presence of biomarkers indicative of COVID-19 in coughs and speech, and to develop robust systems for the detection of COVID-19.

“We hope that this work will not end with the current pandemic and will allow us to continue studying clues of diseases that affect the respiratory system through acoustic signals that can be collected in a non-intrusive way”, said Isabel Trancoso, The Project Coordinator.

Your participation in this study is essential and warmly appreciated. To participate, just follow this link (where you can find the informed consent form), or use the following QR code.

More info: https://www.hlt.inesc-id.pt/w/COVID19