Response of synthetic vocal fold models with geometry based on visible human project data.

Abstract

Numerous synthetic and computational models have been and are currently being used in research studies of human vocal fold vibration. Model geometry plays an integral role in governing dynamic response. However, the various model geometry definitions have typically been idealized and often exhibit wide variability with each other. The present research compares the response of synthetic vocal fold models of different geometries. One model is based on idealized geometry, while the other is based on geometry obtained from the National Library of Medicine’s visible human project (VHP). The process for image extraction, model definition, and model fabrication is described, including: (1) image conversion from 2-D VHP image sequences to 3-D stereolithography (STL) format, (2) conversion to 3-D computer model format in which model geometric manipulations can be performed, and (3) fabrication of synthetic models using rapid prototyping. Results of measurements to characterize the dynamic response of self-oscillating synthetic vocal fold models, including onset and offset pressure, instantaneous glottal width using high-speed imaging, and glottal jet velocity profiles using particle image velocimetry (PIV), are presented for models based on both VHP data and idealized geometries. The sensitivity of the models to geometry changes is also reported.