A reduced-order flow model for vocal fold vibration: from idealized to subject-specific models

Area of Science:

• Computational biomechanics
• Biofluid dynamics
• Acoustic phonetics

Background:

• Accurate simulation of vocal fold vibration is crucial for understanding phonation.
• Existing full 3D fluid-structure interaction (FSI) models are computationally expensive.
• Reduced-order models offer a computationally efficient alternative for FSI simulations.

Purpose of the Study:

• To develop and validate a reduced-order model for simulating vocal fold vibration during phonation.
• To couple 3D tissue mechanics with a 1D glottal airflow model.
• To assess the model's accuracy using both idealized and subject-specific vocal fold geometries.

Main Methods:

• A hybrid 1D/3D FSI model was developed, integrating 3D vocal fold tissue mechanics with a 1D airflow model.
• The airflow model incorporated glottal entrance effects and pressure losses.
• Simulations were performed on idealized and rabbit-specific vocal fold geometries (from MRI data).
• Results were compared against full 3D FSI simulations and in vivo high-speed imaging experiments.

Main Results:

• The reduced-order 1D/3D FSI model demonstrated good agreement with full 3D simulations for idealized geometries.
• Simulations using subject-specific geometries closely matched experimental data for vibration frequency, amplitude, and phase delay.
• The model accurately predicted vocal fold deformation patterns.

Conclusions:

• The developed reduced-order FSI model provides an accurate and efficient method for simulating vocal fold vibration.
• This approach shows significant potential for patient-specific modeling of phonation and voice disorders.
• The validated model can facilitate clinical applications and further research in voice production.