Neurophysiological Muscle Activation Scheme for Controlling Vocal Fold Models

Area of Science:

• Biomechanics of speech and voice production
• Computational modeling of the larynx
• Neurophysiology of vocalization

Background:

• Existing lumped-element vocal fold models often simplify muscle activation.
• Neurological fluctuations in laryngeal muscle activity are inherent to voice production.
• Understanding these fluctuations is crucial for explaining voice variability.

Purpose of the Study:

• To develop a physiologically-based scheme for vocal fold modeling that includes inherent neurological fluctuations.
• To investigate the impact of these stochastic muscle activation components on voice production dynamics.
• To assess the model's ability to replicate observed variability in vocal fold behavior.

Main Methods:

• Developed a lumped-element vocal fold model incorporating stochastic components in neural firing rate and motor unit recruitment.
• Utilized a body-cover model for vocal fold dynamics.
• Performed parametric analysis on steady-state sustained vowels to evaluate the effects of muscle activation fluctuations.

Main Results:

• Inherent muscle activation fluctuations introduce low and high-frequency components, varying with firing rate.
• The model demonstrated changes in system dynamics, including fundamental frequency fluctuations and unstable behavior near bifurcations.
• Stochastic components significantly impact vocal fold parameters, fine structure variability, and pitch control accuracy.

Conclusions:

• The proposed physiologically-based scheme provides a novel approach for controlling lumped-element voice production models.
• The model accurately reflects experimental observations of vocal fold variability and muscle activation signals.
• This framework can be extended to study voice production in neuropathological conditions.