The quantitative assessment of laryngeal physiology

Current clinical voice assessment uses a range of metrics, many of which lack objectivity and/or specificity. Two populations that are negatively impacted from insufficiencies in current voice assessment are individuals with vocal hyperfunction and transmasculine individuals. The following dissertation details four projects that investigated and validated quantitative measures of laryngeal function in order to improve clinical voice assessment in these two populations. Project 1 observed voice onset time (VOT) during modulations of vocal effort, vocal strain, and fundamental frequency (fo) in individuals with typical voices. VOT mean decreased when fo increased, but not during increases in effort or strain, indicating that the individuals with typical voices likely use different musculature when increasing fo than when increasing vocal effort and vocal strain, despite potential increases in laryngeal tension in all instances. VOT variance did not increase when vocal effort and vocal strain increased, suggesting that previous increases in VOT variance observed in individuals with vocal hyperfunction may be inherent to the voice disorder. Project 2 developed automated algorithms to calculate relative fundamental frequency (RFF) from ambulatory accelerometer signals. Average mean bias errors supported that these algorithms could be used to reliably calculate RFF during ecological momentary assessment. Project 3 developed a novel method for calculating RFF based on the end of vocal fold contact observed from laryngeal imaging during voicing offset. Statistically significant decreases in RFF variability with this novel method suggest that decreases in RFF offset patterns are directly driven by a decrease in vocal fold collision forces during abduction. Project 4 explored the development of a resynthesis algorithm to explore how quantitative features of acoustic signals recorded pre- and post- hormone replacement therapy with exogenous testosterone (HRT) affect speech-based gender perception in transmasculine speakers. Listener ratings suggest that mean fo is the single acoustic feature that drives the greatest changes in speech-based gender perception as a result of HRT. The results of these four projects advance the development of quantitative voice assessment and improve understanding of the underlying laryngeal physiology