Date of Defense

4-27-2016

Date of Graduation

4-2016

Department

Speech Pathology and Audiology

First Advisor

Stephen Tasko

Second Advisor

Gregory Flamme

Abstract

Purpose: The goal of the present study is to develop and evaluate an automated technique for measuring biting and chewing related surface electromyographic (EMG) activity of the masseter muscle.

Methods: Data from 28 neurologically healthy participants between the ages of 18-40 were selected for this study. The procedure for collecting the bite/chew data consisted of the participant biting down and chewing three small breath mints while an EMG sensor recorded the masseter muscle activity. A five-step Matlab-based algorithm was created to accurately identify onsets and offsets of each bite and chew event.

Results: Compared to manual measurements using standard conventions, the algorithm was 77% accurate in identifying the onset and offset bite events, and 95% accurate in identifying the onset and offset of chew events. Error analysis revealed a number of consistencies in the form of the errors. Additionally, bite events were typically longer in duration than the chew events and had a larger integral value (area underneath the curve) than the chew events across participants.

Conclusions: Due to inaccuracies identifying onsets and offsets of masseter activity, variables in the algorithm itself may need to be adjusted for more precise measurement. Most issues that arose were within the bite events themselves, so separate criteria may need to be created for the bite events. Finally, due to the varying morphology of the EMG signal during the bite events, the development of a fully automated routine for identification of bites may be more challenging than for the chewing events.

Access Setting

Honors Thesis-Open Access

Bowles HNRS Thesis Presentation .pdf (4505 kB)
Thesis Presentation

Share

COinS