Estimation of Subglottal Pressure, Vocal Fold Collision Pressure, and Intrinsic Laryngeal Muscle Activation From Neck-Surface Vibration Using a Neural Network Framework and a Voice Production Model

Abstract

The ambulatory assessment of vocal function can be significantly enhanced by having access to physiologically based features that describe underlying pathophysiological mechanisms in individuals with voice disorders. This type of enhancement can improve methods for the prevention, diagnosis, and treatment of behaviorally based voice disorders. Unfortunately, the direct measurement of important vocal features such as subglottal pressure, vocal fold collision pressure, and laryngeal muscle activation is impractical in laboratory and ambulatory settings. In this study, we introduce a method to estimate these features during phonation from a neck-surface vibration signal through a framework that integrates a physiologically relevant model of voice production and machine learning tools. The signal from a neck-surface accelerometer is first processed using subglottal impedance-based inverse filtering to yield an estimate of the unsteady glottal airflow. Seven aerodynamic and acoustic features are extracted from the neck surface accelerometer and an optional microphone signal. A neural network architecture is selected to provide a mapping between the seven input features and subglottal pressure, vocal fold collision pressure, and cricothyroid and thyroarytenoid muscle activation. This non-linear mapping is trained solely with 13,000 Monte Carlo simulations of a voice production model that utilizes a symmetric triangular body-cover model of the vocal folds. The performance of the method was compared against laboratory data from synchronous recordings of oral airflow, intraoral pressure, microphone, and neck-surface vibration in 79 vocally healthy female participants uttering consecutive /pæ/ syllable strings at comfortable, loud, and soft levels. The mean absolute error and root-mean-square error for estimating the mean subglottal pressure were 191 Pa (1.95 cm H2O) and 243 Pa (2.48 cm H2O), respectively, which are comparable with previous studies but with the key advantage of not requiring subject-specific training and yielding more output measures. The validation of vocal fold collision pressure and laryngeal muscle activation was performed with synthetic values as reference. These initial results provide valuable insight for further vocal fold model refinement and constitute a proof of concept that the proposed machine learning method is a feasible option for providing physiologically relevant measures for laboratory and ambulatory assessment of vocal function.

Kalman Filter Implementation of Subglottal Impedance-Based Inverse Filtering to Estimate Glottal Airflow during Phonation

“…This supports the use of the original FIR version of the IBIF scheme for such classification tasks, which indicates that factors affecting the classification performance in [16] were not degraded by the airflow estimates. However, other applications more sensitive to signal quality (for instance, the estimation of glottal biomechanics and assessment of tissue-flow-acoustic interaction [65]) can further benefit from the enhancement offered by the proposed Kalman implementation to estimate more accurate glottal airflow in running speech and/or ambulatory scenarios.…”

Kalman Filter Implementation of Subglottal Impedance-Based Inverse Filtering to Estimate Glottal Airflow during Phonation

“…This supports the use of the original FIR version of the IBIF scheme for such classification tasks, which indicates that factors affecting the classification performance in [27] were not degraded by the airflow estimates. However, other applications more sensitive to signal quality (for instance, the estimation of glottal biomechanics and assessment of tissueflow-acoustic interaction [66]) can further benefit from the enhancement offered by the proposed Kalman implementation to estimate more accurate glottal airflow in running speech and/or ambulatory scenarios.…”

LaDIVA: A neurocomputational model providing laryngeal motor control for speech acquisition and production

“…Neurocomputational models of speech are designed to systematically unify available knowledge on speech motor control, to simulate behavior, and test hypotheses related to speech motor control architecture [9][10][11][12]. Likewise, computational models of voice production have improved our understanding of typical and pathological phonation by providing access to relevant features that are difficult, if not impossible, to directly measure [13][14][15][16]. However, current models do not yet integrate physiologically relevant models of phonation into computational models of neural motor control, which would allow for investigation of auditory and somatosensory processing in voice production.…”