Decoding of Ankle Joint Movements in Stroke Patients Using Surface Electromyography

1. Introduction

Stroke remains one of the leading causes of social isolation, disability, and death [1]. In children, the incidence of stroke is rare [2] and it has been estimated that in both men and women, the risk of stroke increases with age [3] while women have more stroke events than men [4].As the average age of population is increasing across the world due to multiple reasons such as advances in medical technology, health care system, and provision of cheap and readily available medicines, it is expected that the number of stroke patients will rise [5,6]. Consequently, more patients will need physical rehabilitation in the future and governments will require induction of an increased number of healthcare professionals than usual to provide physical rehabilitation to these individuals. It is also more likely that the economic burden of stroke will increase and pose challenges to those health systems with limited resources [7].A stroke survivor faces long-term effects after the acute phase of stroke. These effects include the development of impairment, limitations of activities (disability), and reduced participation (handicap) [8]. Although stroke results in a variety of physical and cognitive abnormalities, the most widely recognized is motor impairment, which affects 80% of stroke patients [9]. Commonly, stroke results in loss of movement control of one side of the body, impacting locomotion.In neurological disorders including stroke, the restoration of physical function heavily depends on the onset, the type of injury, and the paradigm being followed for motor function recovery [10]. For the rehabilitation of upper limb motor function in stroke patients, constraint-induced movement therapy (CIMT), robotics, brain–computer interfaces (BCIs), electromyographic biofeedback, and mental practice (MP) combined with motor imagery have shown improvements in motor function [11,12,13,14]. Additionally, repetitive task training, high intensity physiotherapy/physical therapy (PT) and PT in combination with MP [15] have resulted in improved functional outcomes for lower limb mobility [16].Assessing the outcomes of PT over time is very important in evaluating the functional performance of patients as well as the administered intervention. There are many scales used to access motor performance after stroke, but the most commonly used scale for assessing motor impairments in clinical practice is the Fugl-Meyer Assessment (FMA) scale [17].The use of the upper limb is more frequent in performing activities of daily life (ADLs) and the upper limb has been targeted vastly in the areas of physical rehabilitation. Currently, there are various commercially available and widely used rehabilitative systems for upper limb rehabilitation after stroke, such as exoskeletons [18], rehabilitation robots [19], gaming devices [20], and virtual reality (VR) based systems [21]. Many of these rehabilitative devices are electromyography (EMG)-based. However, there is no device commercially available and clinically proven for stroke patients with lower limb motor impairments. Some of the main challenges hindering the commercial availability of the many proposed lower limb rehabilitative devices include design limitations and do not account for the physical requirements of stroke patients [22]. Additionally, more research focus on upper limbs is also an important factor that has resulted in researchers’ comparatively less technical inclination toward the development of lower limb rehabilitative devices.The first step toward the development of an EMG-controlled and home-based lower limb motor rehabilitation device is to investigate the movements of the lower limb in stroke patients. In the normal functionality of the lower limbs, movements that occur at the ankle joint complex have major significance in gait and balance. The available literature on the decoding of ankle joint motions from the movement intention of a user using surface electromyography (sEMG) in healthy subjects as well as in stroke patients is limited. However, Al-Quraishi et al. [23] successfully decoded ankle joint movements in healthy subjects while investigating the impact of different feature extraction and dimensionality reduction techniques on classification accuracies using autoregressive (AR) features and the following classifiers: K-nearest neighbor (k-NN), multilayer perceptron (MLP), and linear discriminant analysis (LDA). Their findings suggested that k-NN along with fuzzy neighborhood preserving analysis with QR (FNPA-QR) decomposition, as a dimensionality reduction technique, provides superior results with an average accuracy of 96.20% ± 4.1%. In another study exploring the biomechanical strategies used by healthy individuals during walking over uneven terrain, Gregory et al. [24] utilized time domain (TD) features (second-order AR coefficients, integrated EMG (IEMG), variance (VAR), waveform length (WL), moving average, and root mean square (RMS)) to predict user intent of performing ankle joint motions using LDA and the classification tree (CART) from sEMG signals. They reported the highest classification accuracy of 77.2% using LDA. Furthermore, Waris et al. [25] evaluated six different classifiers (LDA, ANN, K-NN, SVM, TREE, and naïve Bayes) in a multiday evaluation to identify the most suitable algorithm for sEMG classification of hand motions. In their study, ANN performed better of all classifiers.The purpose of this study was to investigate the potential use of sEMG for a home-based ankle joint rehabilitative device using PR approaches and to evaluate the performance of two classifiers (ANN and LDA). In this study, the intent of performing different ankle joint movements was decoded from the recorded sEMG of stroke patients and the relationship between motor impairment and functional movements was explored. Previously, it has been reported that the classification accuracy for upper limbs is affected by impairment level in stroke patients [26,27]. Additionally, patients were observed by the data collection team throughout the experimental protocol and notes were taken. A total of four movements take place at the ankle joint complex: dorsiflexion and plantar flexion in the sagittal plane, and eversion and inversion in the frontal plane [27].

— Update: 08-01-2023 — cohaitungchi.com found an additional article The Association between Mental Motor Imagery and Real Movement in Stroke from the website www.ncbi.nlm.nih.gov for the keyword deciphering real or not movements of stroke extremeties.

1. Introduction

A stroke occurs when the blood supply to part of the brain is interrupted or reduced, preventing brain tissue from receiving oxygen and nutrients. Brain cells begin to die in minutes. There are two main causes of stroke: ischemic, which occurs when a vessel supplying blood to the brain is obstructed, and hemorrhagic, in which a weakened blood vessel supplying the brain bursts. The consequence of stroke in both cases is a neurological deficit derived from the fact that a part or area of the brain stops working properly. The neurological deficit is one of the main causes of disability worldwide. [1].

Read more  Sharing Mayo Clinic: Heart health for myself and my family

There are over 13.7 million new strokes of all types each year worldwide. Every year, over 116 million years of healthy life are lost due to stroke [2]. The predominant long-term disability, which usually features the worst prognosis, is upper limb (UL) movement [3,4]. Approximately only one-third of people affected by stroke achieve fully functional UL recovery [3,5].

Thus far, a wide range of strategies and devices have been developed for the purpose of promoting upper limb motor recovery after stroke by taking advantage of the brain’s ability to reorganize its neural networks after injury. The functional organization of the motor system is modified by use, and it has been suggested that use-dependent plasticity may play a major role in the recovery of function after stroke [6,7,8].

This approach includes neuromodulation techniques, such as transcranial magnetic stimulation or transcranial direct current stimulation, or sensory transformation techniques, such as mental practice/mental imagery or mirror therapy, which enhance use-dependent plasticity. More precisely, it should be noted that motor imagery (MI) is the mental representation of an action without a physical movement or muscular activation.

Research supports the idea that motor imagery should, to some extent at least, involve the same neuronal substrate as an executed movement [9]. Indeed, several studies reported that greater activation is achieved in the supplementary motor area (SMA), the premotor cortex (PMC), and the primary motor cortex (M1) in subjects during both executed and imagined movement [10,11,12,13,14]. Some brain areas that are activated during motor imagery belong to the neural network known to be involved in the early stage of motor control (i.e., motor programming) [10].

Read more  How to Execute Different Strokes in Table Tennis

This finding has opened many interesting lines of research, which can be described as follows: (i) knowledge of MI for the application of therapy [15,16,17,18,19]; (ii) evidence regarding the neurophysiological bases of MI [20,21,22,23]; (iii) knowledge regarding the relationship between MI and physical movements [24,25]; (iv) analysis and validation of assessment tools to measure the ability to visualize movements [26,27]. Thus, this research aimed to examine the relationship between MI and real UL movements in individuals affected by stroke. Considering that the visualization of a movement can activate neurons in the same motor areas, the purpose of this study was to address the following research question: what happens if we cannot create the mental MI, as in the case of stroke? Would the ability to move be affected? In other words, this work aimed to deepen the knowledge about the relationship between mental imagery and motor deficits after stroke.

Many studies support the idea of functional equivalence between motor images and motor execution [28,29,30]. Most of us can imagine moving our fingers typing on a computer, but apparently, we cannot imagine typing faster than we can actually move our fingers [31,32]. A study that analyses the functional equivalence between images and action concludes that if an action cannot be physically performed, it cannot be imagined with a high functional equivalence [33]. Most of these studies have been carried out with healthy participants, analyzing functional equivalence by checking the temporal coupling between motor visualization and their subsequent motor performance of a task [32,34,35]. The innovative approach of the current research is focused on investigating the lack of ability to visualize motor patterns, and how this is related to an actual loss of UL movement and functionality. To this end, it is expected that people who suffered a stroke and are not capable of visualizing movement experience a greater UL handicap, with poor fine motor skills, muscle weakness or spasticity, and loss of functionality and independence.

Recommended For You

About the Author: Tung Chi