OSCAR: vOice Screening of CoronA viRus

The OSCAR project aims to develop a screening service for suspected cases of COVID-19 based on the NOS voice technology already on the market. Added to this service is the provision of a platform for epidemiological studies for researchers, so that they can have access to voice data, as well as improving and developing algorithms for investigating respiratory diseases.

The OSCAR project has born from an invitation made by Professor Hugo Gamboa from the Physics Department of NOVAFCT and the Associação Fraunhofer Portugal to participate in a consortium led by NOS Inovação and I accepted this challenge. The consortium needed a clinical partner to build a solid application for a Portugal 2020 call launched by the National Innovation Agency (ANI). The project aimed to “build” algorithms to identify patients with a positive RT-PCR test for SARS CoV-2, through voice and cough sounds. The second goal is to search, through artificial intelligence, for vocalization sounds characteristic of various respiratory diseases such as asthma and Chronic Obstructive Pulmonary Disease. This is how the OSCAR Project was built. This project consists of 2 phases. The first phase is aimed at building and validating COVID-19 recognition algorithms through the voice and cough sounds of people recruited from 4 centers where PCR analyses were carried out for the diagnosis of COVID-19. The second phase, now underway, intends to collect voice and cough sounds from people recruited from the general population who have made a PCR or antigen testing at diagnostic centers. At this stage the test results are obtained by self-report. People are invited to answer questions and record vocalization and coughing sounds through a website.