Differentiating the Physiological Sources of Modulated Voice

Project Summary Essential tremor (ET) is the most prevalent movement disorder affecting approximately 25 million people worldwide. Up to sixty-two percent of patients with ET develop vocal tremor (VT) involving oscillation within the respiratory, laryngeal, pharyngeal-oral, or velopharyngeal-nasal subsystems. These oscillations produce

modulation of the fundamental frequency (pitch) and intensity (loudness) of the voice, resulting in a ‘shaky’ voice and increased effort during speech production. Neurological voice disorders like VT are detrimental to communication, professional productivity, and quality of life. Unfortunately, the current approaches for medical

management of VT have inconsistent effects on voice production and can have adverse effects. Furthermore, current methods for behavioral management of VT do not target specific impairments and employ a wide range of therapeutic techniques leading to inefficient therapy. The challenges in managing VT stem from a lack of

accessible assessment methods to identify the subsystems affected by tremor and guide targeted treatment. Clinical assessment of VT requires the use of nasolaryngoscopy to identify oscillation within the larynx and vocal tract (i.e., pharyngeal-oral and velopharyngeal-nasal subsystems). However, access to

nasolaryngoscopy outside of large voice centers is limited by cost, time, and training demands. Furthermore, nasolaryngoscopy cannot be used to identify oscillation within the respiratory subsystem, necessitating the addition of visual and tactile or respiratory kinematic assessments. Although standard acoustical assessments

can be employed in a variety of settings at low cost, current acoustical analyses of microphone signals cannot differentiate the subsystems affected by tremor because a combination of laryngeal source and vocal tract filter features are represented. Our recent study on vocal vibrato (i.e., a singing technique that involves modulation of voice similar to VT)

used a vibration sensor applied to the neck surface in singers to capture features of the source prior to vocal tract filtering. Analyses of simultaneously recorded vibration sensor and microphone signals revealed distinct patterns in the extent of intensity modulation thought to be related to differences in singers’ involvement of the

respiratory system, larynx, and vocal tract in vibrato. In addition, preliminary studies in computational models and speakers with VT showed distinct patterns of modulation based on the physiological source of tremor. Thus, the proposed study aims to determine if simultaneously recorded neck-skin vibration sensor and head-

mounted microphone signals differentiate the subsystems involved in voice modulation in singers producing vibrato as a model of VT and in speakers with VT. Furthermore, the proposed study aims to determine if vibration sensor and microphone signals capture the effects of laryngeal botulinum toxin injections in speakers

with VT. The findings of this study may address the critical need to advance assessment of VT, facilitate individualized and targeted treatments for speakers with VT, and measure patients’ responses to treatment.