Quantitative evaluation of motor functional recovery process in chronic stroke patients during robot-assisted wrist training

https://doi.org/10.1016/j.jelekin.2008.04.002Get rights and content

Abstract

This study was to investigate the motor functional recovery process in chronic stroke during robot-assisted wrist training. Fifteen subjects with chronic upper extremity paresis after stroke attended a 20-session wrist tracking training using an interactive rehabilitation robot. Electromyographic (EMG) parameters, i.e., EMG activation levels of four muscles: biceps brachii (BIC), triceps brachii (TRI, lateral head), flexor carpiradialis (FCR), and extensor carpiradialis (ECR) and their co-contraction indexes (CI) were used to monitor the neuromuscular changes during the training course. The EMG activation levels of the FCR (11.1% of decrease from the initial), BIC (17.1% of decrease from the initial), and ECR (29.4% of decrease from the initial) muscles decreased significantly during the training (P < 0.05). Such decrease was associated with decreased Modified Ashworth Scores for both the wrist and elbow joints (P < 0.05). Significant decrease (P < 0.05) was also found in CIs of muscle pairs, BIC&TRI (21% of decrease from the initial), FCR&BIC (11.3% of decrease from the initial), ECR&BIC (49.3% of decrease from the initial). The decreased CIs related to the BIC muscle were mainly caused by the reduction in the BIC EMG activation level, suggesting a better isolation of the wrist movements from the elbow motions. The decreased CI of ECR& FCR in the later training sessions (P < 0.05) was due to the reduced co-contraction phase of the antagonist muscle pair in the tracking tasks. Significant improvements (P < 0.05) were also found in motor outcomes related to the shoulder/elbow and wrist/hand scores assessed by the Fugl–Meyer assessment before and after the training. According to the evolution of the EMG parameters along the training course, further motor improvements could be obtained by providing more training sessions, since the decreases of the EMG parameters did not reach a steady state before the end of the training. The results in this study provided an objective and quantitative EMG measure to describe the motor recovery process during poststroke robot-assisted wrist for the further understanding on the neuromuscular mechanism associated with the recovery.

Introduction

Stroke is a leading cause of permanent disability in adults. Approximately 70–80% of people after stroke have upper-extremity impairment (Nakayama et al., 1994, Parker et al., 1986) and require continuous long-term medical care for reducing the physical impairments. Rehabilitation is important for releasing the uncomfortable symptoms and partially restoring motor functions. The traditional view on poststroke rehabilitation is that significant improvements in motor recovery only occur within the first year after stroke, associated greatly with the spontaneous recovery of the injured brain (Duncan et al., 1992). However, recent studies suggested that intensive therapeutic interventions, such as constraint-induced movement therapy and task-relevant repetitive practice of the affected limb could even contribute to significantly reduced motor impairment and improved functional use of the affected arm for chronic stroke (Lum et al., 2002, Milner et al., 1995, Miltner et al., 1999). In the rehabilitation of the upper limb, however, many stroke survivors experienced reasonable motor recovery of their proximal upper limb (shoulder and elbow), but limited recovery at the distal (wrist) (Chae et al., 2002, Chae and Hart, 2003). It is important to explore the effects of a poststroke treatment and the related recovery process in patients for the design and improvement of a rehabilitation program. The possible reasons for the relatively poorer recovery achieved in poststroke wrist rehabilitation are the limited understanding of the recovery process related to this joint and lack of quantitative measure to monitor the progressive changes during poststroke wrist training.

Many methods for evaluating poststroke training effects on upper-limb motor functions have been widely used for pre- and post- training assessments, including the Fugl–Meyer assessment (FMA) (Fugl-Meyer et al., 1975), the FIM instrument (keith et al., 1987), the Motor Status Score (MSS) test (Ferraro et al., 2002), and the Modified Ashworth Scale (MAS) (Ashworth, 1964) for grading spasticity during passive limb movements and motor functional performance. However, these clinical evaluations are accomplished through observation of unconstrained commonplace movements, and by the subjective judgment of the evaluator. Furthermore, it is difficult to conduct clinical assessments during each session in a rehabilitation training program, due to the time-consuming assessing procedure and the limited manpower support. However, clinical assessments before and after the training could not reveal the motor recovery process during the training. Therefore, more efforts are needed on the quantitative description of the poststroke motor recovery process related to a training program. Changes of movement smoothness during a robot-assisted stroke recovery for the upper limb have been described by Rohrer’s group (Rohrer et al., 2002), however, the kinematic parameters used in their study for the evaluation of the movement smoothness did not directly reveal the evolution of the poststroke motor system during rehabilitation. Hu et al. quantitatively illustrated the motor recovery process during robot-assisted elbow training for chronic stroke by electromyographic (EMG) parameters (Hu et al., 2007a). However, the above studies were not related to the wrist joint rehabilitation.

Poststroke rehabilitation programs are usually time consuming and labor demanding to both therapist and patient with one-to-one manual interaction. Recent technologies have made it possible to use robotic devices as the assistance to the therapist, providing safe and intensive rehabilitation with repeated motions to persons after stroke (Colombo et al., 2005, Lum et al., 2002). One of the primary benefits of robotic technology is that the robot can assist the human therapists to conduct rehabilitation programs with more quantitative and reproducible training motions, providing a platform for the quantitative measurement of motor recovery with well-controlled training intensities. The most commonly reported motion types provided by the developed rehabilitation robots are: (1) continuous passive motion, (2) active-assisted movement, and (3) active-resisted movement (Lum et al., 2002). During a treatment with continuous passive motion, the movements of patient’s limb(s) are guided by the robot system as the patient remained in relaxed condition. This type of intervention was found to be effective in temporarily reducing hypertonia in chronic strokes (Schmit et al., 2000), and in maintaining joint flexibility and stability for persons after stroke in early stage (within 3 weeks after the onset) (Volpe et al., 2000). However, passive movement did not significantly benefit motor improvement (Volpe et al., 2000). In an active-assisted robotic treatment (or interactive robotic treatment), the rehabilitation robot would provide external assisting forces, when the patient cannot complete a desired movement independently. It has been found that with voluntary attempts from stroke patients, the interactive robotic treatments were more effective on motor functional improvement than those with continuous passive motions on both the upper limb (Volpe et al., 2000) and lower limb stroke rehabilitation (Tong et al., 2006). The robotic treatment with active-resisted motion involved voluntarily completing movements against programmed resistance (Fasoli et al., 2003, Hu et al., 2007a, Hu et al., 2007b, Song et al., 2006). This type of training was found to be effective on the increase of muscle force (Fasoli et al., 2003). However, most of the developed rehabilitation robots for poststroke training on the upper limb are mainly for the movements associated with large joints, i.e., the shoulder and elbow. Poststroke training program by rehabilitation robot with the focus on the wrist function recovery has not been well studied yet. In our previous work, an interactive rehabilitation robot driven by electromyography (EMG) for the elbow training on chronic stroke was developed (Hu et al., 2007a, Song et al., 2006); and the motor recovery process associated with a 20-session robot-assisted training was evaluated by quantitative EMG parameters (Hu et al., 2007a). Reduced muscle spasticity and improved muscle co-activation patterns were observed during the training process. Motor improvement in the upper extremity was also obtained after the robot-assisted elbow training assessed by the clinical scales of FMA, MSS, and MAS (Hu et al., 2007a). In this work, the motor recovery process during a robot-assisted wrist training for chronic stroke was quantitatively studied, with the attempt to provide further understanding on the neuromuscular mechanisms associated with the wrist functional recovery.

Section snippets

Methodology

After obtaining approval from the Human Subjects Ethics Subcommittee of the Hong Kong Polytechnic University, we recruited 15 hemiplegic stroke survivors with single cerebral unilateral lesion for the study (12 males and 3 females). All of the subjects were in the chronic stage (at least 1 year after the onset of stroke, age = 53.7 ± 11.2). All subjects received a robot-assisted wrist training, consisting of 20 sessions, at least 3 sessions/week and at most 5 sessions/week, and finished within 7

Clinical score and tracking performance

Fig. 3 shows the results of the clinical assessments before and after the training. There was no statistical difference found in all clinical scales (P > 0.05) among the three repeated assessments before the training. After the training, the MAS for both the elbow and wrist parts decreased significantly compared to the scores of the pre-training (P = 0.0037 for the elbow MAS, and P = 0.013 for the wrist MAS, and P < 0.05 in post hoc tests). The FMA scores for the shoulder/elbow and wrist/hand after the

Motor recovery assessed by clinical scores and tracking performance

The motor recovery associated with the robot-assisted training was first assessed by the traditional clinical scores before and after the training as many other studies (Fasoli et al., 2003, Volpe et al., 2004). Before the training, the motor functions of the recruited subjects were relatively stable, represented by the non-significant variations in the three pre-training clinical scores. The decreases in the MAS and FMA scores after the training suggested the reduced muscle spasticity, or

Conclusion

In this work, the motor functional recovery during the poststroke robot-assisted wrist training was quantitatively assessed by the EMG parameters, i.e., the EMG activation level and CI values in relation to the muscles of FCR, ECR, BIC, and TRI. The evolution of the improvements in the motor outcome, e.g., the decreased muscle hypertonia and better muscle coordinating patterns, could be assessed by the EMG parameters proposed in this study. These results were also consistent with the findings

Acknowledgment

The work described in this paper was fully supported by a grant from the Hong Kong Polytechnic University (1-BB50).

X.L. Hu received her PhD degree in Biomedical Engineering from the Department of Electronic Engineering, The Chinese University of Hong Kong, in 2002. After that, she joined the Department of Health Technology and Informatics, The Hong Kong Polytechnic University, for postdoctoral training. She is currently an Instructor in the Department of Health Technology and Informatics with research interests in the neuromuscular modeling, bio-signal processing, and design of stroke rehabilitation

References (36)

  • J. Chae et al.

    Muscle weakness and cocontraction in upper limb hemiparesis: relationship to motor impairment and physical disability

    Neurorehabil Neural Repair

    (2002)
  • H.S. Cheng et al.

    Improving elbow torque output of stroke petients with assistive torque controlled by EMG signals

    J Biomed Eng

    (2003)
  • R. Colombo et al.

    Robotic techniques for upper limb evaluation and rehabilitation of stroke patients

    IEEE T Neural Syst Rehabil

    (2005)
  • J.R. Cram

    Introduction to surface electromyography

    (1998)
  • P.W. Duncan et al.

    Measurement of motor recovery after stroke

    Stroke

    (1992)
  • S.E. Fasoli et al.

    Effects of robotic therapy on motor impairment and recovery in chronic stroke

    Arch Phys Med Rehabil

    (2003)
  • M. Ferraro et al.

    Assessing the motor status score: a scale for the evaluation of upper limb motor outcomes in patients after stroke

    Neurorehabil Neural Repair

    (2002)
  • D. Flament et al.

    Time course and temporal order of changes in movement kinematics during learning of fast and accurate elbow flexions

    Exp Brain Res

    (1999)
  • Cited by (0)

    X.L. Hu received her PhD degree in Biomedical Engineering from the Department of Electronic Engineering, The Chinese University of Hong Kong, in 2002. After that, she joined the Department of Health Technology and Informatics, The Hong Kong Polytechnic University, for postdoctoral training. She is currently an Instructor in the Department of Health Technology and Informatics with research interests in the neuromuscular modeling, bio-signal processing, and design of stroke rehabilitation programs.

    Kai-yu Tong received his Ph.D. in Bioengineering from the University of Strathclyde, Glasgow, UK in 1998. He spent four months as a research fellow at Strathclyde University and participated in a joint project with the Spinal Cord Injury Unit, Southern General Hospital, Scotland, UK. He joined the Hong Kong Polytechnic University in 1999 and as an assistant professor in the Department of Health Technology and Informatics in 2001. His research interests include rehabilitation robot, the control of functional electrical stimulation for upper and lower extremity functions, sensor development, stroke rat model and gait training rehabilitation on persons after stroke.

    Rong Song received the B.Sc. degree in electrical engineering from Tsinghua University in 1999 and M.Sc. degree in electrical and computer engineering from Shantou University in 2002. Then, he obtained the Ph.D. degree in biomedical engineering in 2006 from the Department of Health Technology and Informatics, The Hong Kong Polytechnic University. His research interests include musculoskeletal modeling, biomedical processing, human motion analysis and robot assisted stroke rehabilitation.

    X.J. Zheng received her B.Eng. degree in Electronic Information Engineering from Northwestern Polytechincal University in 2004. In 2006, she received her M.Eng. degree in Biomedical Engineering from Zhejiang University. And then she served as the research assistant for the Department of Health Technology and Informatics in The Hong Kong Polytechnic University. Her research interests include the biomedical signal processing and modeling.

    K.H. Lui is currently an undergraduate of Biomedical Engineering in Hong Kong Polytechnic University. He worked as a part-time research assistant since 2006. His current research interest is related to Brain Computer Interface and Function Electrical Stimulation for stroke patients.

    Wallace Woon-Fong Leung received his B.Sc. in Mech. Eng. from Cornell in 1977 graduating at top of the entire Engineering Class of 530 students. He also received MSME (1978) and ScD (1981) in Mech. Eng. from MIT. He has worked for Gulf Oil Corp. and Schlumberger in his earlier oil-patch career. He was the R&D Director of a Bird-Baker Hughes for 18 years specializing in separation equipment and later President of Advantech Eng. specializing in biotech separation. He is currently Chair Professor of Mech. Eng. and Director of the Research Institute of Innovative Products & Technologies of The Hong Kong Polytechnic University. His research interests are in healthcare products and technologies, nanofiber filtration, stenosis in arteries and centrifugation.

    Shamay Ng is a registered physical therapist. She received her Master of Applied Sciences in 1999 from University of Sydney, Australia; and her Ph.D. degree from The Hong Kong Polytechnic University, Hong Kong in 2005. She is currently an assistant professor in the Department of Rehabilitation Sciences, The Hong Kong Polytechnic University. Her research interests are development of innovative treatment for motor recovery in people with stroke.

    Stephanie S.Y. Au-Yeung received her M.Phil. and Ph.D. in 2006 from The Hong Kong Polytechnic University. Having a professional background in physiotherapy, her research has been focussed on physical therapeutics that enhances neurological and functional recovery in patients after stroke. She is currently an assistant professor in The Hong Kong Polytechnic University.

    View full text