Development of a vocal biomarker for fatigue monitoring in people with COVID-19

ObjectiveTo develop a vocal biomarker for fatigue monitoring in people with COVID-19. DesignProspective cohort study. SettingPredi-COVID data between May 2020 and May 2021. ParticipantsA total of 1772 voice recordings was used to train an AI-based algorithm to predict fatigue, stratified by gender and smartphones operating system (Android/iOS). The recordings were collected from 296 participants tracked for two weeks following SARS-CoV-2 infection. primary and secondary outcome measuresFour machine learning algorithms (Logistic regression, k-nearest neighbors, support vector machine, and soft voting classifier) were used to train and derive the fatigue vocal biomarker. A t-test was used to evaluate the distribution of the vocal biomarker between the two classes (Fatigue and No fatigue). ResultsThe final study population included 56% of women and had a mean ({+/-}SD) age of 40 ({+/-}13) years. Women were more likely to report fatigue (P<.001). We developed four models for Android female, Android male, iOS female, and iOS male users with a weighted AUC of 79%, 85%, 86%, 82%, and a mean Brier Score of 0.15, 0.12, 0.17, 0.12, respectively. The vocal biomarker derived from the prediction models successfully discriminated COVID-19 participants with and without fatigue (t-test P<.001). ConclusionsThis study demonstrates the feasibility of identifying and remotely monitoring fatigue thanks to voice. Vocal biomarkers, digitally integrated into telemedicine technologies, are expected to improve the monitoring of people with COVID-19 or Long-COVID.

Deep immune profiling reveals early-stage and highly coordinated immune responses in mild COVID-19 patients

While immunopathology has been widely studied in severe COVID-19 patients, immunoprotective factors in non-hospitalized patients have remained largely elusive. We systematically analyzed 484 peripheral immune cell signatures, various serological parameters and TCR repertoire in a longitudinal cohort of 63 mild and 15 hospitalized patients versus 14 asymptomatic and 26 control individuals. Within three days following PCR diagnosis, we observed coordinated responses of CD4 and CD8 T cells, various antigen presenting cells and antibody-secreting cells in mild, but not hospitalized COVID-19 patients. This early-stage SARS-CoV-2-specific response was predominantly characterized by substantially expanded clonotypes of CD4 and less of CD8 T cells. The early-stage responses of T cells and dendritic cells were highly predictive for later seroconversion and protective antibody levels after three weeks in mild non-hospitalized, but not in hospitalized patients. Our systemic analysis provides the first full picture and early-stage trajectory of highly coordinated immune responses in mild COVID-19 patients.