Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors

Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors

+ MODEL Hong Kong Physiotherapy Journal (2014) xx, 1e11 Available online at www.sciencedirect.com ScienceDirect journal homepage: www.hkpj-online...

2MB Sizes 5 Downloads 30 Views

+

MODEL

Hong Kong Physiotherapy Journal (2014) xx, 1e11

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.hkpj-online.com

RESEARCH REPORT

Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors C.V. Shendkar, MTech a, P.K. Lenka, PhD b, Abhishek Biswas, MD b, Ratnesh Kumar, D Ortho, MS, DNB c, M. Mahadevappa, PhD a,* a

Medical Instrumentation Laboratory, School of Medical Science & Technology, Indian Institute of technology (IIT) Kharagpur, West Bengal 721302, India b National Institute for the Orthopaedically Handicapped e Kolkata, West Bengal, 700 090, India c Department of Disability Affairs, Ministry of Social Justice and Empowerment, New Delhi, 110001, India

KEYWORDS foot drop; functional electrical stimulation (FES); gait; physiotherapy; surface electromyography (sEMG)

Abstract The purpose of this parallel group controlled clinical trial was to investigate the therapeutic effect of functional electrical stimulation (FES) on gait, motor recovery, and motor cortex activity. Adults experiencing foot drop <6 months poststroke were allocated to the FES group (physiotherapy and FES stimulation, n Z 14) or the control group (physiotherapy, n Z 14). Each group received their respective therapy 5 days/week for 12 weeks. Gait, surface electromyography (sEMG) of the tibialis anterior muscle in the affected leg, and electroencephalogram (EEG) signals from the foot motor area were assessed at baseline and again after the 12-week intervention. The results showed that the FES intervention induced significantly more changes in various gait swing parameters such as foot pulling acceleration (measured in unit of gravitational constant G; net between-group difference: 0.11  0.02 G, p Z 0.021), swing power (0.11  0.03 G, p Z 0.027) and ground impact (0.12  0.04 G, p Z 0.046) than the control group. EEG analysis revealed that the FES group had significantly altered beta-3 mean (0.50  0.09, p Z 0.021), beta-4 mean (0.60  0.05, p Z 0.024) and alpha peak frequency (0.15  0.02, p Z 0.035). Finally, analysis of sEMG data showed a significantly greater increase in amplitude (in root mean square; 13.2  2.11 mV, p Z 0.033), mean power frequency (5.5  0.80 Hz, p Z 0.024) and median power frequency (6.5  0.90 Hz, p Z 0.021) of the tibialis anterior muscle on the affected side in the FES group. We concluded that the FES combined with physiotherapy induced better outcomes in the swing phase of the gait cycle,

Conflicts of interest: The authors declare that there are no financial or nonfinancial conflicts of interest in this study. * Corresponding author. School of Medical Science and Technology, Indian Institute of Technology, Kharagpur, West Bengal, 721302, India. E-mail addresses: [email protected], [email protected] (M. Mahadevappa). http://dx.doi.org/10.1016/j.hkpj.2014.10.003 1013-7025/Copyright ª 2014, Hong Kong Physiotherapy Association Ltd. Published by Elsevier (Singapore) Pte Ltd. All rights reserved.

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003

+

MODEL

2

C.V. Shendkar et al. activation of the affected ankle dorsiflexor muscles, and cortical function, when compared with conventional physiotherapy alone. Copyright ª 2014, Hong Kong Physiotherapy Association Ltd. Published by Elsevier (Singapore) Pte Ltd. All rights reserved.

Introduction Foot drop often results from the inability to dorsiflex the foot during the swing phase of the gait cycle, and may be observed in neurological conditions such as stroke. Functional electrical stimulation (FES) has been used clinically for correcting foot drop. It involves the application of a sequence of stimulus pulses to the peroneal nerve of the affected limb in patients with stroke. In 2009, National Health Science (NHS)-UK issued an interventional procedure guidance (National Institute for Health and Clinical Excellence guideline Number IPG278) on the use of FES. NHS-UK recommended the use of FES to address foot drop problems of neurological origins [1]. An extensive review of the literature on FES clinical trials revealed that the majority of research groups used walking speed as the outcome measure to evaluate the efficacy of FES [2]. Additional outcomes used were the Modified Ashworth Scale, FugleMeyer Assessment score, and range of motion [3,4]. Presumably, if the use of the peroneal nerve stimulator has a therapeutic effect on functional performance, physiological changes should also be observed. Yet, research on investigating the physiological changes and the associated functional changes is scarce [5]. Further clinical trials are thus necessary to investigate the physiological changes occurring in the motor cortex, muscle function, and improvement in gait. We advocate evaluating not only the gait cycle, but also the surface electromyograms (sEMGs) of stimulated muscles and the electroencephalogram (EEG) signals from the stroke survivor’s motor cortex, to obtain a more complete picture of post-FES changes. Conventional gait parameters such as walking speed, cadence, and step length do not adequately explain effects of FES. When analysing hemiplegic gait, the swing phase dynamics are more important than stance phase parameters [6], but so far are unexplored by researchers [7]. In this study, we used novel parameters that could be beneficial in exploring issues such as quality of walking and foot clearance while walking. Post-FES sEMG changes should be analysed by looking at their association with gait changes. FES is supposed to increase muscle performance and consequently improve the walking in people with hemiplegia. In addition to performing an extended gait analysis for assessing functional changes, we performed sEMG analysis to allow a better understanding of how FES therapy affects muscle condition. The available literature revealed that most of the EEGrelated work on persons with stroke had focused on analysis of short duration EEG signals for identifying the motorrelated cortical potential or event-related desynchronisation [8e10]. The aim of these studies, however, was to analyse EEG for use in brain-computer interface applications, rather than evaluating the therapeutic effects of FES. Until 2006, the majority of clinical trials on FES efficacy used

gait or changes in muscle function to evaluate post-FES improvement. A frequently cited hypothesis by Rushton [11] provided a new outlook on how FES improved motor function. He highlighted that FES not only restored functional abilities, but may also affect neural remodelling or brain plasticity. Following this new outlook, researchers began new experiments to explore how FES improved gait by analysing EEG signals and movement-related cortical potentials. One important weakness of many FES-induced EEG modulation studies is that only a single session (30 minutes) of FES was used and healthy individuals were evaluated rather than clinically relevant populations [8e10]. To evaluate FES efficacy, a longer treatment may be required as well. This study was also designed to evaluate the effects of FES on the motor cortex by analysing EEGs obtained after 3 months of FES application.

Methods Participant enrolment and trial registration Individuals with the following criteria were included in the study: history of stroke with hemiparesis at least 3 months prior, medically stable, able to walk at least 10 metres independently, having no established contraindication for electric stimulation, and ability to give informed consent. Exclusion criteria were: motor neuron injury affecting the lower limbs with poor response to stimulation, history of frequent fall (>1/week), severe cardiac disease such as myocardial infarction, congestive heart failure, use of demand pacemaker, fixed ankle contraction of 5 in plantar flexion direction, or inability to operate the safety button of the device on their own, if required. Please refer to our detailed protocol of the registered trial available at www. ctri.gov.in bearing the registration number CTRI/2012/09/ 003019 for details. Our trial details are also available at the World Health Organization International Clinical Trials Registry Platform. The study was approved by the ethical committees of the Indian Institute of Technology, Kharagpur, India and the National Institute for the Orthopaedically Handicapped, Kolkata, India. Written, informed consent was obtained from the participants before actual data collection and interventions took place. All procedures were conducted in accordance with the Declaration of Helsinki.

Group assignment A randomized controlled study design is considered the best design for establishing cause and effect [12]. However, such design has several limitations, including substantial financial costs, intensive use of resources, compromised generalisability of research findings, and the necessity for external monitoring [13,14]. A controlled clinical trial study design

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003

+

MODEL

3 was used instead of the randomized controlled design after consideration of practical feasibility and available resources. Participants were assigned sequentially to the FES or control group (Fig. 1). Stroke patients were assigned to groups on a first-in, first-out basis. The study involved 28 stroke patients equally divided into two equal groups e FES and control. There were no statistically significant differences in the characteristics between the two groups of patients at baseline. The details of enrolment are shown in Fig. 1 and the characteristics of individuals in the FES and control groups are shown in Table 1.

Intervention The first arm (FES group) underwent FES therapy for 30 minutes and 30 minutes of conventional physiotherapy. A single channel stimulator (CEFAR Step II of Cefar Compex, now part of DGO global, Vista, California, USA) was used to apply the

Figure 1.

FES (Fig. 2A). The placement of the stimulator unit, electrodes, and foot switch is shown in Fig. 2A. Patients walked with the stimulator turned on for 30 minutes daily. Biphasic charge balanced stimulation waveform was used for applying the stimulation. FES was delivered with 0.18-millisecond pulses at 40 Hz; stimulation intensity level was adjusted between 20 mA and 50 mA to cause dorsiflexion in the affected foot. FES was provided by electrotherapists, with analysis performed by the authors. Physiotherapy was provided by physiotherapists recognized by the government of India. Conventional physiotherapy was provided for the second arm (control) as it would have been unethical to withdraw all treatment. Patients in the control group underwent exercise with a trained physiotherapist for 60 minutes, involving stretching, a range of motion enhancement exercises, muscle strengthening, and gait and balance training. Interventions for both groups were applied 5 days/week, for up to 12 weeks.

Flow diagram of the study. FES Z functional electrical stimulation.

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003

+

MODEL

4

C.V. Shendkar et al. Table 1

Characteristics of participants

Parameter

FES group

p (between group comparisons for All recruited Participants who All recruited Participants who participants completed follow-up participants completed follow-up those who completed (n Z 17) assessment (n Z 14) (n Z 17) assessment (n Z 14) follow-up assessment)

Male/female, n Age (y) Body weight (kg) Body height (cm) Left/right side affected, n Time since stroke (mo) Stroke type (ischemic/haemorrhagic), n

11/6 50.2  9.3 59.3  7.3 164.4  5.8 6/11 5.2  0.9 15/2

9/5 51.0  11.0 60.4  8.0 164.2  6.0 5/9 5.5  0.7 12/2

Control group

11/6 49.8  10.7 62.7  9.5 165.6  4.7 5/12 5.3  0.7 15/2

9/5 48.7  12.7 62.2  12.7 165.2  4.0 4/10 5.2  0.8 12/2

0.993 0.454 0.665 0.569 0.877 0.825 0.992

Data are presented as mean  standard deviation, unless otherwise indicated. FES Z functional electrical stimulation.

Measurements Gait, EEG, and sEMG measurements were performed at two time points: baseline and after 12 weeks of FES intervention. Gait, EEG, and sEMG signals were obtained sequentially. The definitions of various outcome variables derived from these measurements are shown in Appendix 1. The persons who performed the outcome measurements were blinded to group allocation. A wireless portable gait analyser (MiniSun LLC, Fresno, CA, USA) was used to acquire gait data. Fig. 2B shows the gait acquisition device with placement of sensors. The patients were asked to walk along a hospital passage over a 50 metre course. All patients walked this course without stimulation. Parameters including walking speed, cadence, step length, single limb support (SLS), double limb support (DLS), pulling power, swing power, and ground impact were acquired using the portable gait analyser. The details are given in Table 2. EEG signals were also measured. The feasibility of using one or two EEG channels (by placing the electrodes on Fz,

Cz locations as per 10e20 electrode placement system) for acquisition and analysis of foot movement-related changes in the motor cortex area has been well established in stroke rehabilitation in the past decade [8e10]. We used the guidelines of the International Federation of Clinical Neurophysiology (IFCN) for digital recording of the EEG data [13]. We used a single EEG electrode on the Cz location, which is in proximity to the foot motor area. Placement of an electrode on the Cz location was performed as shown in Fig. 2C. Somatotopic mapping of the human primary sensorimotor cortex during foot motor imagery and motor execution shows that a large part of the motor cortex is associated with foot movement [15]. This area is represented as the Cz point in the 10e20 EEG electrode placement system. The chances of incorrect Cz electrode placement was small, due to the comparatively large size of the motor area. Standard processes were applied while positioning the electrode to avoid variability and misplacement of electrodes. The EEG signal was recorded with a linked ear reference from the Cz location of the scalp. To ensure reproducibility of the signal, EEG signal

Figure 2. Protocol for functional electrical stimulation (FES) delivery and method used for acquisition of GAIT, surface electromyography (sEMG) and electroencephalogram (EEG) signals. (A) Method used for application of FES; (B) gait measurement using wireless portable gait analyser; (C) EEG acquisition from Cz point according to standard 10e20 International EEG system; and (D) sEMG recording from tibialis anterior (TA) muscles of affected limb.

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003

+

MODEL

0.200 0.200 0.200 0.021*** 0.027*** 0.046*** 0.100 0.119 0.100 0.082 4.2 3.2 4.2 0.02 0.03 0.04 1.3 3.4 0.05 0.22           22.8 17.7 19.0 0.11 0.11 0.12 5.3 10.7 0.06 0.09 0.319 0.164 0.182 0.009** 0.024** 0.049** 0.082 0.082 0.073 0.066 24.8 24.9 24.8 0.22 0.12 0.23 3.2 7.2 0.10 0.16           394.1 143.6 274.4 0.78 0.69 0.94 55.3 108.5 0.58 1.04 0.012* 0.015 0.010* 0.008* 0.010* 0.020* 0.041 0.042 0.036 0.060

435.5 171.1 254.4 0.72 0.64 0.89 50.9 99.3 0.54 0.95

         

29.1 34.2 32.6 0.21 0.11 0.21 4.1 6.9 0.05 0.03

Post Baseline

35.5 33.8 31.2 0.22 0.12 0.23 4.1 8.2 0.04 0.10           385.0 131.0 293.9 0.82 0.72 1.08 62.0 111.3 0.62 1.14 Single limb support (ms) Double limb support (ms) SLS/DLS (%) Pulling acceleration (G) Swing power (G) Ground impact (G) Speed (m/min) Cadence (steps/min) Step length (m) Stride length (m)

449.1 176.2 254.9 0.72 0.63 0.91 52.3 91.5 0.52 0.96

         

31.2 53.1 21.8 0.21 0.11 0.21 5.3 8.1 0.02 0.04

Post Baseline

Data are presented as mean  standard deviation, unless indicated otherwise. * Statistically significant difference over time (paired t test, p < 0.025). ** Statistically significant group  time interaction (two-way repeated measures analysis of variance, p < 0.05). *** Statistically significant between-group difference in change score (independent t test, p < 0.05).

0.020* 0.025* 0.020* 0.015* 0.020* 0.020* 0.048 0.048 0.050 0.080

1.139 2.248 2.020 8.911 6.020 4.793 3.684 3.684 3.802 4.020

p Mean difference  SD p F

Difference in change score Group  time interaction effect p Control group p FES Group Parameter

Table 2

Change in gait parameters in both functional electrical stimulation (FES) and control group within the group and comparison of change scores

5 electrodes were placed close to the foot motor area, which was represented by the Cz point according to the international 10e20 system. EEG data were recorded using an acquisition device (Biograph Infiniti, Thought Technology, Montreal, QC, Canada). Participants were asked to sit comfortably on a chair with the knee flexed at 90 and the ankle in a neutral position as shown in Fig. 2C. Data were recorded for a total period of 5 minutes while performing dorsiflexion of the ankle of the affected limb at intervals of 5 seconds. A sampling frequency of 256 Hz was used for data acquisition. Data were later processed using Matlab (Math Works Inc., Natick, MA, USA). European recommendations (the Surface Electromyography for the Non-Invasive Assessment of Muscles or SENIAM project) and the International Society of Electrophysiology and Kinesiology (ISEK) standards for acquisition and processing of sEMG data were followed [16]. Electrode position, sampling frequency, and filter parameters were set according to these international standards. Data acquisition was performed using a bio-potential data acquisition system (PowerLab, AD Instruments, Bella Vista, NSW, Australia). sEMG of the tibialis anterior muscle of the affected limb was recorded during voluntary contraction. The FES device was turned off while the sEMG signal was recorded. Electrode placement and the sEMG acquisition process are shown in Fig. 2D. Isometric maximum voluntary contraction was used as an sEMG normalization strategy [17]. The participants were asked to sit on a chair with the knee flexed at 90 and the ankle in a neutral position. Each performed active ankle dorsiflexion of the affected lower limb at a uniform contraction rate and force approximately 15 times during the 150 seconds of data recording.

Data processing The gait, sEMG, and EEG signals were acquired at pre- and post-intervention for each individual. Typical gait, sEMG, and EEG signals obtained are displayed in Fig. 3A, B, and C, respectively. The gait cycle measured by the sensor from both hemiplegic and contralateral lower limbs, EEG signal acquired from motor cortex by positioning electrodes at the Cz location according to standard 10e20 electrode placement system, and sEMG signal acquired from tibialis anterior muscle of the affected limb are shown. EEG signals were processed and separated into delta, theta, alpha, SMR, beta, beta-1, -2, -3, gamma, and wide band. IFCN guidelines were used to select filter type, filter order, ripple setting, cut-off frequencies, and dampening settings [14]. Information about the time and frequency domain was extracted from all the bands for analysis. Analysis was performed using MATLAB R2010b (Mathwork Inc. Natick, MA, USA) and the statistical analysis software SPSS Inc. (version 17.0, Chicago, IL, USA). SENIAM, ISEK standards, and general good practice guidelines were followed to remove artefacts while processing the sEMG data. All data were sampled at 1000 Hz with a 12-bit A/D converter. Low-pass filtering (4th-order Bessel filter,  3% accuracy, frequencies 500 Hz at 3 dB), high-pass filtering (1st-order filter,  0.25% accuracy, frequencies 20 Hz at

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003

+

MODEL

6

C.V. Shendkar et al.

Figure 3. Gait, surface electromyography (sEMG) and electroencephalogram (EEG) signals. (A) Shows gait cycle measured by the sensor from both, hemiplegic and counter lower limb; (B) EEG signal acquired from the motor cortex by positioning electrode at Cz location according to standard 10e20 electrode placement system; and (C) sEMG signal acquired from tibialis anterior muscle of the affected limb.

3 dB), and notch filtering (2nd-order filter, frequencies 50 Hz at 32 dB) were used for data processing. We followed the standards for reporting sEMG data [15]. Segments of sEMG during dorsiflexion of the lower limb were used for analysis. Both temporal and spectral parameter analyses were performed.

Statistical analysis Statistical analyses were performed with SPSS 17.0 (SPSS Inc.). First, differences in the characteristics of the

participants between the FES group and the control group were compared using the independent t test (for interval or ratio data), the Mann-Whitney U test (for ordinal data), and the Chi-square test (for nominal data). The KolmogorovSmirnov test was used to check the normality of data. As aforementioned, no statistically significant differences in the characteristics between the two groups of patients were observed (Table 1). To investigate the therapeutic effects of FES on the outcome measures, two-way analysis of variance (ANOVA) with repeated measures was used. In each of the ANOVA models, a mixed design was used, with time as the within-

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003

+

MODEL

0.248 0.021*** 0.024*** 0.073 0.035*** 0.9 0.1 0.1 0.1 0.0     

Mean Difference  SD p

2.7 0.5 0.6 0.7 0.2 0.291 0.018** 0.022** 0.091 0.025**

p F

1.248 7.139 6.575 3.684 6.466 0.082 0.045 0.046 0.100 0.046

Post Baseline Baseline

Post

    

4.9 1.9 1.5 0.2 1.2

0.050 16.8  6.1 20.1  6.1 0.033 3.4  0.6 4.1  0. 9 0.044 2.7  0.4 3.2  0.9 0.047 3.4  0.7 3.9  0.2 0.048 9.1  0.7 9.8  1.2

Difference in change score Group  time interaction effect p*

Data are presented as mean  standard deviation, unless indicated otherwise. *p for within-group comparison over time (paired t test, level of significance Z 0.025). ** Statistically significant difference in group  time interaction (two-way repeated measures analysis of variance, p < 0.05). *** Statistically significant between-group difference in change score (independent t test, p < 0.05).

Significant group  time interaction effects were detected for Root mean square (RMS; also known as quadtratic mean) value (F Z 6.575, p Z 0.028), mean power frequency

EEG-spectral

sEMG analysis

Wide band amplitude 17.3  4.7 23.3 Beta 3 mean 3.5  0.7 4.7 Beta 4 mean 2.8  0.5 3.9 Beta 5 mean 3.3  0.7 4.5 Alpha peak frequency 9.3  1.0 10.1

There was a significant group  time interaction effect for beta-3 mean (F Z 7.139, p Z 0.018), beta-4 mean (F Z 6.575, p Z 0.022), and alpha peak frequency (F Z 6.466, p Z 0.025) (Table 3). Analysis of the change scores revealed that the FES group had significantly more changes in beta-3 mean (mean between-group difference Z 0.50  0.09, p Z 0.021), beta4 mean (0.60  0.05, p Z 0.024) and alpha peak frequency (0.15  0.02, p Z 0.035) compared with the controls.

EEG-temporal

EEG analysis

Control group

In order to gain a comprehensive overview of the influence of FES on gait, both conventional parameters, such as walking speed, cadence, step length, SLS, and DLS, and new parameters, such as pulling power, swing power, and ground impact, were used [18,19]. Among the gait variables (Table 2), the pulling acceleration (F Z 8.911, p Z 0.009), swing power (F Z 6.020, p Z 0.024), and ground impact (F Z 4.793, p Z 0.049) demonstrated significant group  time interactions. In the FES group, pulling acceleration (initial swing) improved by 13.8% (p Z 0.008), swing power (terminal swing) by 14.2% (p Z 0.010), and ground impact by 18.6% (p Z 0.020; Table 2). In the control group, the pulling power improved by only 8.3% (p Z 0.015), swing power by 7.8% (p Z 0.020), and ground impact by 5.6% (p Z 0.020). The change score results showed that the FES intervention induced significantly more changes in various gait swing parameters such as foot pulling acceleration (mean between-group difference Z 0.11  0.02, p Z 0.021), swing power (0.11  0.03, p Z 0.027), and ground impact (0.12  0.04, p Z 0.046) than the control group. Other gait variables demonstrated no significant group  time interaction effects (p > 0.05), indicating that FES had no therapeutic effects on these outcomes compared with the control treatment.

p*

Gait analysis

FES group

Results

Signal-domain Parameter

participant factor (baseline measurement vs. postintervention measurement) and group as the betweenparticipant factor (FES vs. control). For those variables that showed significant group  time interaction in the ANOVA, independent t tests were used to compare the change score between the two groups. In addition, paired t tests were used to examine changes over time in each group. A p value <0.05 was considered statistically significant in change score and group  time interaction analysis. The significance level was adjusted to 0.025 for the post hoc comparison of the pre- and post-test scores for each group.

Table 3 Changes in electroencephalogram (EEG) temporal and spectral parameters observed within the functional electrical stimulation (FES) and control groups and comparison of change scores

7

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003

MODEL

0.033*** 0.703 0.307 0.024*** 0.021*** 0.307 2.1 19.2 10434 0.8 0.9 2.4       13.2 40.4 30944 5.5 6.5 1.9 0.028** 0.057 0.291 0.010** 0.014** 0.310 Data are presented as mean  standard deviation, unless indicated otherwise. *p for within-group comparison over time (paired t test, level of significance Z 0.025). ** Statistically significant group  time interaction (two-way repeated measures analysis of variance, p < 0.05). *** Statistically significant between-group difference in change score (independent t test, p < 0.05).

Post Baseline Post Baseline

0.028 108.1  14.3 121.5  16.1 0.050 371.6  79.1 440.1  60.8 0.046 431558  114224 496472  108157 0.028 81.1  11.2 91.6  7.1 0.028 73.2  9.1 80.7  10.2 0.046 170.6  32.0 200.4  38.9

0.028 0.050 0.082 0.046 0.037 0.073

6.575 4.802 1.248 8.802 6.910 1.139

Mean p difference  SD p F

Difference in change score Group  time interaction effect p* Control group p* FES group

This work showed the functional and physiological changes that occurred following 3 months of FES therapy. This study is unique for two main reasons. First, extensive gait analysis was performed to improve our understanding of post-FES stance and swing phase dynamics and second, the muscular (motor recovery pattern) and cortical changes were evaluated in addition to the gait outcomes. The gait assessment used in this study was different from existing studies in two main ways. First, instead of using a conventional walking test such as the 6 minute walking test, figure-of-8 walking test, or the 10 metre walkway, we conducted a longer-distance walking test (50 metres). This improved gait data acquisition and minimized the effect of individual walking style, which can be present when shorter distances are examined. We analysed parameters including pulling power, swing power, ground impact, and ratio of SLS to DLS (SLS/DLS) [23,24], to provide both stance phase and swing phase dynamics. The results showed that the FES therapy had significant treatment effects on pulling power, swing power, and ground impact, as revealed by the significant group  time interaction. These changes reflect an improvement in ankle dorsiflexion ability and thus foot clearance. It is important to determine whether the gait changes are accompanied by changes in the neural system. However, the literature concerning cortical changes associated with long-term use of FES is limited [8,25]. Studies evaluating EEGs of healthy individuals have already established that limb movement (flexion) modulates the cortical potential. Nascimento and co-workers [26] concluded that information concerning movement-related parameters for plantar-flexion tasks are encoded in the primary motor cortex (M1), and a separate group of researchers found that beta oscillations induced by movement or stimulation of the lower limb were best recorded over the corresponding cortical representation area (Cz) [27]. We therefore focused on finding the FES-related changes in the motor cortex.

Signal-domain Parameter

Discussion

Table 4 Changes in surface electromyography (sEMG) temporal and spectral parameters observed within the functional electrical stimulation (FES) and control groups and comparison of change scores

(F Z 8.802, p Z 0.010), and median power frequency (F Z 6.910, p Z 0.014) (Table 4). Post hoc analysis of the change scores showed that the FES group had significantly more improvements in root mean square value (mean between-group difference Z 13.2  2.11, p Z 0.033), mean power frequency (5.5  0.80, p Z 0.024), and median power frequency (6.5  0.90, p Z 0.021) of the tibialis anterior muscle on the affected side compared with the controls. No other significant results were observed in other sEMG parameters. Temporal parameter improvement showed the enhancement of muscle force produced by the tibialis anterior muscle [20,21]. Analysis of spectral parameters of sEMG signals showed a significant shift in spectral values post-FES, and shifts in the mean and median power frequencies towards the higher frequencies. Spectral edge frequencies also shifted towards higher values post-FES. These variations in sEMG spectral frequency post-FES therapy indicated a concomitant increase in muscle fibre conduction velocity [22].

C.V. Shendkar et al.

sEMG-temporal sEMG RMS value (mV) 112.5  14.3 139.1  15.7 sEMG peak value (mV) 381.1  77.8 490.0  91.6 Average slope (mV/s) 421558  124438 517415  143792 sEMG-spectral Mean power frequency (Hz) 82.3  12.7 98.3  11.5 Median power frequency (Hz) 75.1  8.1 89.1  11.1 Spectral edge frequency (Hz) 191.2  25.1 220.1  27.8

+ 8

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003

+

MODEL

9 Human EEG studies have shown that changes in the alpha (8e13 Hz) and beta (15e25 Hz) bands can be used to detect functional activation of sensorimotor cortex. Research carried out by Fabrizio De Vico Fallani and coworkers, or you could keep this as De Vico Fallani and coworkers [28] showed how stroke significantly changed the structural properties of the brain networks associated with preparation and execution of motor movements, and concluded that poststroke changes were particularly prominent in the beta band (13e29 Hz), known to be involved in motor tasks. Giaquinto and co-workers [29] (1994) attempted to quantify values of EEG following stroke and their possible correlation with stroke recovery, and found that alpha mean frequency was reduced in the injured hemisphere following stroke. These findings thus suggested that alterations in beta band and alpha frequency may be indicative of functional improvement following neurostimulation or physiotherapy. In this study, EEG was used as an outcome measure to evaluate the effects of FES on the motor cortex. EEG temporal parameters (alpha, beta, gamma, delta, theta, sensorimotor rhythm-mu, and wide band) and spectral parameters (alpha peak frequency) were analysed to determine the possible effects of FES on the motor cortex. Our analysis showed that our FES protocol had significant effects on beta-3 and -4 band amplitudes and alpha peak modulation, suggesting that FES induced changes in the sensory-motor cortex area (e.g., strengthened corticospinal connections) in persons with foot drop [30]. Similar results were observed by Neuper and Pfurtscheller [27] in healthy individuals. They found that different cortical processes, such as recovery from movement and stimulation, resulted in beta oscillations in the frequency range of 14e35 Hz in healthy individuals [27]. We performed tibialis anterior muscle sEMG analysis in both temporal and spectral domains and evaluated the significance of improvement in muscle condition. In the kinesiology research community, using only spectral analysis for understanding muscle conditions after therapy is controversial [31]. We therefore focused our work on investigating motor changes by including both the temporal and spectral features of a

GAIT parameters

Single limb support (ms) Double limb support (ms) SLS/DLS (%) Pulling acceleration (initial swing) (G) Swing power (terminal swing) (G) Ground impact (G) Speed (m/min) Cadence (steps/min) Step length (m)

post-FES signal. These changes were associated with muscle fibre conduction velocity. The FES program induced significant improvements in EMG amplitude (in RMS value), mean frequency, and median frequency compared with the control group (Table 4). These changes are indicative of improved motor strength and muscle fibre conduction velocity following FES therapy [32].

Limitations A major limitation of this study was the lack of random allocation of individuals to the two treatment groups. However, no significant differences in important characteristics were identified between the two groups.

Conclusion Our study showed that 3 months of FES intervention induced significant therapeutic effects in hemiplegic patients with foot drop. Comprehensive gait analysis revealed that FES improved quality of walking and foot clearance. These improvements are accompanied by changes in ankle muscle activation and cortical activity.

Acknowledgements The authors would like to thank the National Institute of Orthopedically Handicapped, Kolkata (under the Department of Disability Affairs, Ministry of Social Justice and Empowerment, Government of India, New Delhi) for providing financial support for this research work. We express our sincere gratitude to the participants in this study and their caregivers.

Appendix 1. Definitions of analysed Gait, electroencephalogram (EEG) and surface electromyography (sEMG) parameters

Single-limb support is the duration within the gait cycle for which the body mass is carried by a single limb Double-limb support is the duration of the gait cycle during which both feet are in ground contact It is the ratio of single-limb support duration to the double-limb support duration The pulling acceleration/power is defined as the maximum forward acceleration of the foot during the initial swing phase The swing power is defined as the maximum deceleration during the mid and terminal swing phases Ground impact is defined as the maximum deceleration in the vertical direction during the weight acceptance Speed is computed using the actual stride length/actual stride time Cadence is the rate at which a person walks Step length is the distance between the proximal end position (continued on next page)

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003

+

MODEL

10

C.V. Shendkar et al.

(continued )

Stride length (m)

EEG parameters

Wide band amplitude (mV) Beta-3 mean (mV) Beta-4 mean (mV) Beta-5 mean (mV) Alpha peak frequency (Hz)

sEMG parameters

sEMG RMS value (mV) sEMG peak value (mV) Average slope (mV/s) Mean power frequency (Hz) Median power frequency (Hz) Spectral edge frequency (Hz)

of the foot at ipsilateral heel strike to the proximal end position of the foot at the next contralateral heel strike Stride length is the distance between proximal end position of the foot at ipsilateral heel strike to the proximal end position of the foot at the next ipsilateral heel strike It is periodic averages of EEG signal epoch for 1e38 Hz frequency band It is periodic averages of EEG signal epoch for 22e26 Hz frequency band It is periodic averages of EEG signal epoch for 26e30 Hz frequency band It is periodic averages of EEG signal epoch for 30e38 Hz frequency band From the raw EEG, a fast Fourier transform calculation (FFT) is calculated and the frequency value of the highest frequency in the alpha range (8e12 Hz) of the spectrum is measured as alpha peak frequency sEMG RMS value obtained by computing the root mean square value in a particular time window (here dorsiflexion duration) of the raw EMG It is a value where sEMG amplitude reaches to its peak in particular time window (here dorsiflexion duration) It is a time domain feature that measures the relative average slope of the EMG signal Mean power frequency is a frequency at which the average power within the signal is reached Median power frequency is a frequency at which the total power within the signal reaches 50% of its maximum value Spectral edge frequency indicates the highest frequency below which 95% of the total power is located

References [1] National Institute for Health and Clinical Excellence (NICE, UK) interventional procedure guidance 278. Functional electrical stimulation for drop foot of central neurological origin. 2009. p. 1e9 [accessed 12.01.12], http://guidance.nice.org. uk/IPG278. [2] Santaguida P, Oremus M, Walker K, Wishart LR, Siegel KL, Raina P. Systematic reviews identify important methodological flaws in stroke rehabilitation therapy primary studies: review of reviews. J Clin Epidemiol 2012;65:358e67. [3] Subramanya K, Pinto A, Kumar M, Arya BK, Mahadevappa M. Surface electrical stimulation technology for stroke rehabilitation: a review of 50 years of research. J Med Imaging Health Inform 2012;2:1e14. [4] Pomeroy VM, Chambers SH, Giakas G, Bland M. Reliability of measurement of tempo-spatial parameters of gait after stroke using GaitMat II. Clin Rehabil 2004;18:222e7. [5] Rabadi MH. Review of the randomized clinical stroke rehabilitation trials in 2009. Med Sci Monit 2011;17:RA25e43. [6] Simonsen EB, Alkjær T. The variability problem of normal human walking. Med Eng Phys 2012;34:219e24. [7] Wilkie KM, Shiels JE, Bulley C, Salisbury LG. Functional electrical stimulation (FES) impacted on important aspects of my life-A qualitative exploration of chronic stroke patients’ and carers’ perceptions of FES in the management of dropped foot. Physiother Theory Pract 2012;28:1e9. [8] Takahashi M, Gouko M, Ito K. Functional Electrical Stimulation (FES) effects for Event Related Desynchronization (ERD) on foot motor area. ICME International Conference on Complex Medical Engineering 2009;1:1e6. [9] Knash ME, Kido A, Gorassini M, Chan KM, Stein RB. Electrical stimulation of the human common peroneal nerve elicits lasting facilitation of cortical motor-evoked potentials. Exp Brain Res 2003;153:366e77.

[10] Mu ¨ller-Putz GR, Kaiser V, Solis-Escalante T, Pfurtscheller G. Fast set-up asynchronous brain-switch based on detection of foot motor imagery in 1-channel EEG. Med Biol Eng Comput 2010;48:229e33. [11] Rushton DN. Functional electrical stimulation and rehabilitationdan hypothesis. Med Eng Phys 2003;25:75e8. [12] http://ccg.cochrane.org/non-randomised-controlled-studynrs-designs [accessed 01.04.14]. Non-randomised controlled study (NRS) designs. Review group of Cochrane collaboration. Page last updated 8th December 2009. [13] Bush T. Beyond HERS: some (not so) random thoughts on randomized clinical trials. Int J Fertil Women’s Med 2000;2: 55e9. [14] Nuwer MR, Comi G, Emerson R, Fuglsang-Frederiksen A, Guerit JM, Hinrichs H, et al. IFCN standards for digital recording of clinical EEG. The International Federation of Clinical Neurophysiology. Electroencephalogr Clin Neurophysiol Suppl 1999;52:11e4. [15] Stippich C, Henrik O, Klaus S. Somatotopic mapping of the human primary sensorimotor cortex during motor imagery and motor execution by functional magnetic resonance imaging. Neurosci Lett 2002;331:50e4. [16] Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol 2000;10: 361e74. [17] Burden A. How should we normalize electromyograms obtained from healthy participants? What we have learned from over 25 years of research. J Electromyogr Kinesiol 2010;20: 1023e35. [18] Merletti R, Di Torino P. Standards for reporting EMG data. J Electromyogr Kinesiol 1999;9:3e4. [19] Daly JJ, Zimbelman J, Roenigk KL, McCabe JP, Rogers JM, Butler K, et al. Recovery of coordinated gait: randomized controlled stroke trial of functional electrical stimulation (FES)

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003

+

MODEL

11

[20]

[21]

[22]

[23]

[24]

[25]

versus no FES, with weight-supported treadmill and overground training. Neurorehabil Neural Repair 2011;25:588e96. Sabut SK, Sikdar C, Mondal R, Kumar R, Mahadevappa M. Restoration of gait and motor recovery by functional electrical stimulation therapy in persons with stroke. Disabil Rehabil 2010;32:1594e603. Staudenmann D, Roeleveld K, Stegeman DF, van Diee ¨n JH. Methodological aspects of SEMG recordings for force estimationea tutorial and review. J Electromyogr Kinesiol 2010;20:375e87. Masuda T, Kizuka T, Zhe JY, Yamada H, Saitou K, Sadoyama T, et al. Influence of contraction force and speed on muscle fiber conduction velocity during dynamic voluntary exercise. J Electromyogr Kinesiol 2001;11:85e94. Robbins SM, Houghton PE, Woodbury MG, Brown JL. The therapeutic effect of functional and transcutaneous electric stimulation on improving gait speed in stroke patients: a meta-analysis. Arch Phys Med Rehabil 2006;87:853e9. Kottink AI, Oostendorp LJ, Buurke JH, Nene AV, Hermens HJ, IJzerman MJ. The orthotic effect of functional electrical stimulation on the improvement of walking in stroke patients with a dropped foot: a systematic review. Artif Organs 2004; 28:577e86. Weingarden H, Ring H. Functional electrical stimulationinduced neural changes and recovery after stroke. Eura medicophys 2007;42:87e90.

[26] Nascimento OF, Nielsen KD, Voigt M. Movement-related parameters modulate cortical activity during imaginary isometric plantar-flexions. Exp Brain Res 2006;171:78e90. [27] Neuper C, Pfurtscheller G. Evidence for distinct beta resonance frequencies in human EEG related to specific sensorimotor cortical areas. Clin Neurophysiol 2001;112: 2084e97. [28] de Vico Fallani F, Astolfi L, Cincotti F, Mattia D, la Rocca D, Maksuti E, et al. Evaluation of the brain network organization from EEG signals: a preliminary evidence in stroke patient. Anat Rec 2009;292:2023e31. [29] Giaquinto S, Cobianchi A, Macera F, Nolfe G. EEG recordings in the course of recovery from stroke. Stroke 1994;25:2204e9. [30] Everaert DG, Thompson AK, Chong SL, Stein RB. Does functional electrical stimulation for foot drop strengthen corticospinal connections? Neurorehabil Neural Repair 2010;24: 168e77. [31] Farina D. Counterpoint: spectral properties of the surface EMG do not provide information about motor unit recruitment and muscle fiber type. J Appl Physiol 2008;105:1673e4. [32] Farina D, Pozzo M, Merlo E, Bottin A, Merletti R. Assessment of average muscle fiber conduction velocity from surface EMG signals during fatiguing dynamic contractions. IEEE Trans Biomed Eng 2004;51:1383e93.

Please cite this article in press as: Shendkar CV, et al., Therapeutic effects of functional electrical stimulation on gait, motor recovery, and motor cortex in stroke survivors, Hong Kong Physiotherapy Journal (2014), http://dx.doi.org/10.1016/j.hkpj.2014.10.003