Using Surface Electromyography to Characterize Swallows In Stroke Survivors

• Author / Creator

  Kuffel Kupferschmidt, Kristina L

• Background: 9.44 million adults in the United States experience swallowing difficulties (i.e., dysphagia), with the most common etiology being stroke (Bhattacharyya, 2014). Upon diagnosis, patients are typically referred to clinicians who prescribe swallowing exercises, however access to this therapy is limited (Nund et al., 2014). Researchers have suggested that providing surface electromyography (sEMG) biofeedback during therapy may improve functional outcomes (Langmore & Pisegna, 2015). sEMG sensors are an inexpensive and simple tool used to evaluate muscle activity. Limited studies to-date have looked at characterizing these signals in patients with dysphagia after stroke. Furthermore, these studies have only included time-domain measures such as duration and amplitude (Crary & Baldwin, 1997; Kim et al., 2015). Objectives: sEMG swallowing signals were characterized in patients with dysphagia after stroke in a novel manner using both frequency and time domain analyses. These signals were then used to test if an existing algorithm developed for swallow detection in head and neck cancer patients, can be used to provide accurate feedback in the stroke population. The primary objective of this research was to understand how swallow-detection can be optimized for stroke specific sEMG characteristics. Methods: Two groups of participants were recruited: a post-stroke group with oropharyngeal dysphagia (n=10) and an age- and sex-matched healthy control group (n=10). All participants were outfitted with a wireless sEMG data acquisition system on their submental area. They completed a baseline measurement and 20 regular saliva swallows. Test-retest was evaluated by removing the device and repeating the study procedure. Two studies were completed. The first study used five independent Mann-Whitney U tests to identify if significant differences existed between the sEMG swallow signals in the stroke and control groups. The signal parameters compared were: duration, normalized peak amplitude, median and mean frequency, and signal to noise ratio. Additionally, test-retest of all parameters was completed for the healthy and stroke groups using intraclass correlation coefficients. In the second study, the performance of a swallow-detection algorithm, developed for home-based dysphagia therapy for head and neck cancer patients, was tested using swallows collected in stroke patients. Recall was used as the measure of algorithm performance. If the recall for the first algorithm presented unsatisfactory results, a modified stroke specific algorithm would be generated. In this case a one-tailed pairwise t-test would be completed to understand if the modified algorithm performed better than the original. Results: The first study found that SNR was significantly higher in the healthy (Mdn=13.7, SD=5.4) group than in the stroke (Mdn=8.1, SD=4.6) group, U=325, p<0.001. None of the other tested parameters suggested that differences exist between the two groups. Additionally, test-retest reliability of normalized peak amplitude and duration were found to be poor in the stroke group. All other parameters suggested moderate to very good reliability. In the second study, the original algorithm performed with a recall of 74.55%, which was deemed to be outside of the acceptable range. A modified algorithm was generated and tested. This modified algorithm (M=84.24, SD=11.26) performed significantly better than the original algorithm (M=74.55, SD=16.55), t(10)= -2.667, p=0.024. Significance: These results suggest that signal quality is lower in individuals who have dysphagia after stroke when compared with healthy individuals. Additionally, the findings of the second study suggest that the modified version of the algorithm created using stroke data can perform within the acceptable range but may be improved by taking into consideration more characteristics of stroke specific sEMG signals.

• Subjects / Keywords

  • Dysphagia
  • Swallow detection
  • Surface electromyography

• Graduation date

  Spring 2018

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/R3251G138

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.