biostudies-literatureElbeji AAndré Losch FoundationLuxembourg National Research FundLuxembourg Institute of Healthe062463https://www.ebi.ac.uk/biostudies/studies/S-EPMC9684280biostudies-literatureUnknown12(11)<h4>Objective</h4>To develop a vocal biomarker for fatigue monitoring in people with COVID-19.<h4>Design</h4>Prospective cohort study.<h4>Setting</h4>Predi-COVID data between May 2020 and May 2021.<h4>Participants</h4>A total of 1772 voice recordings were used to train an AI-based algorithm to predict fatigue, stratified by gender and smartphone's operating system (Android/iOS). The recordings were collected from 296 participants tracked for 2 weeks following SARS-CoV-2 infection.<h4>Primary and secondary outcome measures</h4>Four machine learning algorithms (logistic regression, k-nearest neighbours, support vector machine and soft voting classifier) were used to train and derive the fatigue vocal biomarker. The models were evaluated based on the following metrics: area under the curve (ABMJ openVocal biomarker predicts fatigue in people with COVID-19: results from the prospective Predi-COVID cohort study.PMC9684280N/A14716273Despotovic VNazarov PVElbeji AHiga EFischer AZhang LAguayo GFagherazzi GfalseVocal biomarker predicts fatigue in people with COVID-19: results from the prospective Predi-COVID cohort study.<h4>Objective</h4>To develop a vocal biomarker for fatigue monitoring in people with COVID-19.<h4>Design</h4>Prospective cohort study.<h4>Setting</h4>Predi-COVID data between May 2020 and May 2021.<h4>Participants</h4>A total of 1772 voice recordings were used to train an AI-based algorithm to predict fatigue, stratified by gender and smartphone's operating system (Android/iOS). The recordings were collected from 296 participants tracked for 2 weeks following SARS-CoV-2 infection.<h4>Primary and secondary outcome measures</h4>Four machine learning algorithms (logistic regression, k-nearest neighbours, support vector machine and soft voting classifier) were used to train and derive the fatigue vocal biomarker. The models were evaluated based on the following metrics: area under the curve (A2022-01-01T00:00:00Z2022 Nov2026-05-27T19:11:11.878Z2025-02-19T03:23:28.901ZS-EPMC96842803641429410.1136/bmjopen-2022-062463