SEMG-Based Detection of Compensation Caused by Fatigue During Rehabilitation Therapy: A Pilot Study

Shuangyuan Huang,Longhan Xie,Yan Chen,Guofeng Li,Siqi Cai,Ke Ma

doi:10.1109/access.2019.2933287

Shuangyuan Huang, Longhan Xie + Show 4 more

Open Access

https://doi.org/10.1109/access.2019.2933287

Copy DOI

Abstract

To stimulate neuroplasticity, the stroke patient often receives repetitive and high load robotic rehabilitation therapy. Given their impaired motor function, the patients are prone to muscle fatigue. Muscle fatigue lead patients to compensate for upper limb motion by recruiting trunk and shoulder motions, resulting in undesirable rehabilitation motion and a subsequent risk of injury. However, fatigue compensation is not detected during upper-limb robotic rehabilitation training in the existing rehabilitation robot. The aim of this study was to detect the compensation caused by fatigue based on surface electromyography (sEMG). Eight healthy subjects performed three basic repetitive resistance rehabilitation training tasks, to elicit three types of common stroke fatigue compensatory synergies. The subjective fatigue score and sEMGs of the main muscles used were acquired to determine the fatigue state. The compensatory motion was recorded by a motion capture apparatus. The sEMG median frequency (MDF) of the main muscles used and the overall fatigue compensation were calculated. With the development of fatigue, the subjects exhibited more signs of fatigue compensation. The motion types slightly increased the degree of corresponding basic compensation. However, the subjects exhibited similar ranges and trends in the overall compensation. A strong correlation was found between sEMG MDF and overall fatigue compensation. Thus, fatigue compensation can be detected based on the body's status regardless of the type of motion. The sEMG-based detection of fatigue compensation proposed in this study is a reliable way to detect fatigue compensation and improve the quality of therapeutic exercise during upper-limb robotic rehabilitation training.

Highlights

Stroke is the leading cause of acquired disability for adults [1]
EXPERIMENTAL SETUP AND MEASUREMENT PROCEDURE The data acquisition and rehabilitation training experiments were conducted on the ReROBOT platform, as shown in Figure 1, which was developed in our laboratory
Based on the above analysis, we proposed detecting fatigue compensation based on the surface electromyography (sEMG) median frequency (MDF) regardless of the motion type, which is different from detecting pathological compensation

Summary

Introduction

Stroke is the leading cause of acquired disability for adults [1]. Depending on the lesion locations, strokes may damage the motor and sensory neural system and frequently lead to permanent neurological impairment that is associated with serious physical and cognitive dysfunction [2], [3]. Most stroke survivors lose the ability to perform daily activities [4], [5]. The associate editor coordinating the review of this manuscript and approving it for publication was Rajeeb Dey. According to neuroplasticity, which describes how the human central nervous system can structurally and functionally adapt to acquire new skills, highly repetitive rehabilitation exercise can help recover physical and cognitive abilities [6], [7]. Robotic stroke rehabilitation therapy, which has the potential to improve the recovery process for stroke survivors, can deliver highly intense repetitive training [8]

Objectives

Methods

Results