ALA046 - 3D Reconstruction of Vocal Fold Medial Surface in Phonation: Effects of Superior and Recurrent Laryngeal Nerve Stimulation on Vibration and Acoustics
Friday, April 29, 2022
11:20 AM – 11:26 AM CT
Location: Landmark D
Neha Reddy, BA; Patrick Schlegel, PhD; Yoonjeong Lee, PhD; Dinesh K. Chhetri, MD
Background: Although the pattern of vocal fold medial surface vibration affects voice quality, most studies have focused on 2D analysis of vocal fold vibration from superior view. We performed 3D reconstruction of vocal fold medial surface from an in vivo canine hemilarynx during phonation. Glottal deformation and resulting acoustics were investigated as a function of superior and recurrent laryngeal nerve (SLN, RLN) stimulation.
Methods: Across graded SLN and RLN activation, the hemilarynx was excited to phonation with airflow. High-speed video and audio were recorded. Medial surface reference points tracked during vibration were mapped into 3D space, and surface shape was restored using interpolation. From reconstructions, glottal surface shape, cycle-based parameters, and intraglottal volume shape during glottal cycle were calculated. For each phonatory condition, F0, cepstral peak prominence (CPP), and harmonic amplitude (H1–H2) were measured.
Results: The glottis was convergent during opening and divergent during closing. Increased SLN activation resulted in decreased closing quotient. Decreased glottal volume was associated with increased SLN and RLN. CPP increased as SLN increased and was high at higher RLN levels. Simultaneously increased SLN and RLN activation exhibited higher F0 and lower H1–H2 (constricted glottis).
Conclusions: 3D reconstruction of in vivo medial surface vibration revealed the effects of laryngeal nerve stimulation on vibratory patterns and acoustic correlates of voice quality. SLN activation results in quicker glottal closure per cycle, decreased glottal volume, and less breathy, higher-pitched, less noisy voice. RLN has a similar effect on acoustic measures but no effect on closure per cycle.