A Deep Neural Network Based Glottal Flow Model for Predicting Fluid-Structure Interactions during Voice Production.

Yang Zhang,Qian Xue,Xudong Zheng

doi:10.3390/app10020705

Abstract

This paper proposes a machine-learning based reduced-order model that can provide fast and accurate prediction of the glottal flow during voice production. The model is based on the Bernoulli equation with a viscous loss term predicted by a deep neural network (DNN) model. The training data of the DNN model is a Navier-Stokes (N-S) equation-based three-dimensional simulation of glottal flows in various glottal shapes generated by a synthetic shape function, which can be obtained by superimposing the instantaneous modal displacements during vibration on the prephonatory geometry of the glottal shape. The input parameters of the DNN model are the geometric and flow parameters extracted from discretized cross sections of the glottal shapes and the output target is the corresponding flow resistance coefficient. With this trained DNN-Bernoulli model, the flow resistance coefficient as well as the flow rate and pressure distribution in any given glottal shape generated by the synthetic shape function can be predicted. The model is further coupled with a finite-element method based solid dynamics solver for simulating fluid-structure interactions (FSI). The prediction performance of the model for both static shape and FSI simulations is evaluated by comparing the solutions to those obtained by the Bernoulli and N-S model. The model shows a good prediction performance in accuracy and efficiency, suggesting a promise for future clinical use.

Highlights

Voiced sound production in the human larynx is a complex fluid-structure interaction process in which the forced air from the lungs interacts with vocal fold tissues to initiate sustained vibrations that modulate the glottal airflow
A continuum-mechanics based vocal fold model is coupled with the Bernoulli model to obtain various shapes extracted from one vibration cycle; based on these shapes, the deep neural network (DNN) model is trained in the same way as in the synthetic shape case
The training data for the DNN model is collected by generating various glottal shapes using the synthetic shape function

Summary

Introduction

Voiced sound production in the human larynx is a complex fluid-structure interaction process in which the forced air from the lungs interacts with vocal fold tissues to initiate sustained vibrations that modulate the glottal airflow. One of the important research goals in voice production is to understand the interaction mechanism between glottal aerodynamics and vocal fold tissue mechanics. The glottis, which refers to the space between the two vocal folds, forms a convergent shape during vocal fold opening and a divergent shape during vocal fold closing This alternative convergent-divergent glottal shape generates a temporal pressure asymmetry inside the glottis, which ensures sustained energy transfer from airflow to vocal fold tissues to sustained vibrations. Voice disorders are often associated with vocal fold pathologies, such as nodule, cyst, scar, paralysis, and so forth. These pathologies alter the geometry and material properties of vocal fold tissues, resulting in irregular mucosal waves

Objectives

Methods

Results

Conclusion