Methods for Voice and Swallow Assessment through Laryngeal High-Density Surface Electromyography
Skip to main content
eScholarship
Open Access Publications from the University of California

UC San Diego

UC San Diego Electronic Theses and Dissertations bannerUC San Diego

Methods for Voice and Swallow Assessment through Laryngeal High-Density Surface Electromyography

Abstract

Voice and swallow are complex functions made possible through the coordination ofmultiple muscles of the throat. Unfortunately, these tasks are adversely impacted by aging, neurologic disorders, nerve injuries, cancer, and stroke—yet lack the tools for objective and non-invasive assessment. For instance, traditional surface electromyography (sEMG) of the throat suffers from drawbacks of “cross-talk” contamination, skin-electrode impedance, and diminished target-muscle specificity, which lead to performance variability and limited clinical utility. This dissertation explores the use of high-density surface electromyography (HDsEMG) coupled with novel implementation of array signal processing techniques to overcome limitations of traditional sEMG when studying the neck. During phonation in healthy subjects, results yielded power spectrum density energy maps with the capacity to visually distinguish active regions associated with the underlying cricothyroid and anterior strap musculature. Low-pitch and high-pitch differentiation was accomplished using multivariate log likelihood ratio testing with an average Receiver Operating Characteristic area under the curve of 0.97, which exceeds that of traditional sEMG by 0.20. During swallowing in healthy subjects, HDsEMG energy maps confirmed lateral symmetry and dominant activity in the suprahyoid region. Additional studies conducted on human subjects utilizing various swallow textures and complexities demonstrated average EMG duration that increased proportionally with increasing texture complexity. Multivariate analysis improved automated detection of onsets and offsets of swallows and was able to classify one of five distinct textures with an average probability error of 0.16. Preliminary results for validation against the “gold-standard”, needle EMG, demonstrate HDsEMG’s ability to detect specific localized activity similar to the needle electrode underneath. Lastly, we demonstrate the feasibility of using flexible electronic sensor arrays, in lieu of standard needle and clunky electrodes arrays, to provide greater subject comfort, mobility, and adhesion to the curvature of the neck.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View