Dr. Maria Dietrich, Assistant Professor and Director of the Vocal Control and Vocal Well-Being Lab in the Department of Speech, Language and Hearing Sciences and her students presented two posters at The Fall Voice conference in Seattle. Graduate student Melinda Pfeiffer and senior Erin Tippit joined Dr. Dietrich in Seattle to present work based on Dietrich’s NIDCD R15 research grant, “Classifying neck surface EMG signals for the early detection of vocal fatigue in student teachers.”
The project classifies surface EMG signals from the anterior neck for the detection of vocal fatigue based on Vocal Fatigue Index scores. A pattern recognition algorithm developed by Co-Investigator Guilherme DeSouza and doctoral student Yixiang Gao from the College of Engineering showed a sensitivity of 90% and specificity of 95% to identify early career teachers with mild or greater vocal fatigue compared with controls. Further, Melinda Pfeiffer with support from Erin Tippit presented data that the acoustic measure voice relative fundamental frequency was sensitive to changes in laryngeal tension during a vocally fatiguing speech production protocol, which was used to collect the surface EMG data. Early career teachers with vocal fatigue had worse values over the course of three time points compared with controls.
The key findings are that a computer algorithm can be trained using signals from neck sensors to determine with high certainty if someone has vocal fatigue or not and that an acoustic marker called relative fundamental frequency shows promise to track vocal fatigue over the course of speech production.