An anatomically-based, time-domain acoustic model of the subglottal system for speech production

Julio C Ho, Purdue University

Abstract

Mathematical models of human voice production are widely used in research to study the underlying acoustic mechanisms of phonation. While existing models can recreate the anatomical structures of the supraglottal system (upper respiratory airways) and the vocal folds with realistic detail, they ignore or oversimplify the branching anatomical complexity of the subglottal system (lower respiratory airways) and the acoustic properties associated with this system. Consequently, current modeling techniques fail to incorporate the acoustic contributions of the subglottal system in an anatomically-relevant manner. In this thesis research, a time-domain numerical framework for modeling sound wave propagation in the branching airway generations of the subglottal system was formulated and incorporated into a general scheme of voice production. The model was tested and compared to some of the previous observations reported in the literature. To better understand the acoustic nature of the subglottal system and its impact in the overall phonatory process, two sets of numerical experiments were undertaken. First, an acoustic sensitivity analysis was conducted to assess the importance of the major anatomical features of the subglottal system and, second, the acoustic changes induced by the presence of the subglottal system were examined across three different sustained vowels and four vocal fold configurations. It was found that the most significant acoustic properties of the subglottal system are largely attributable to the anatomical features of the first eight or nine airway generations and, to a lesser extent, to the rate at which the total cross-sectional airway area expands at higher generations. In addition, the simulations of sustained vowels indicate that the subglottal system tends to decrease the amplitude of the vocal folds vibrations, lowers the fundamental frequency, and reduces the mean and maximum glottal volume velocities and the amplitude of the supraglottal pressure. Overall, this thesis research establishes the necessary mathematical framework for a more in-depth study of the acoustic behavior of the subglottal system and its role in speech production.^

Degree

Ph.D.

Advisors

George R. Wodicka, Purdue University.

Subject Area

Engineering, General|Engineering, Biomedical|Physics, Acoustics

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS