A Framework for Biomarkers of COVID-19 Based on Coordination of Speech-Production Subsystems

Goal: We propose a speech modeling and signal-processing framework to detect and track COVID-19 through asymptomatic and symptomatic stages. Methods: The approach is based on complexity of neuromotor coordination across speech subsystems involved in respiration, phonation and articulation, motivated by the distinct nature of COVID-19 involving lower (i.e., bronchial, diaphragm, lower tracheal) versus upper (i.e., laryngeal, pharyngeal, oral and nasal) respiratory tract inflammation, as well as by the growing evidence of the virus' neurological manifestations. Preliminary results: An exploratory study with audio interviews of five subjects provides Cohen's d effect sizes between pre-COVID-19 (pre-exposure) and post-COVID-19 (after positive diagnosis but presumed asymptomatic) using: coordination of respiration (as measured through acoustic waveform amplitude) and laryngeal motion (fundamental frequency and cepstral peak prominence), and coordination of laryngeal and articulatory (formant center frequencies) motion. Conclusions: While there is a strong subject-dependence, the group-level morphology of effect sizes indicates a reduced complexity of subsystem coordination. Validation is needed with larger more controlled datasets and to address confounding influences such as different recording conditions, unbalanced data quantities, and changes in underlying vocal status from pre-to-post time recordings.