Edited by: Tadhg Eoghan MacIntyre, University of Limerick, Ireland
Reviewed by: Adam Mark Bruton, University of Roehampton, United Kingdom; Heidi Haavik, New Zealand College of Chiropractic, New Zealand
This article was submitted to Movement Science and Sport Psychology, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Motor imagery (MI) is a key technique in motor learning and motor control facilitating brain plasticity (Shumway-Cook and Woollacott,
From research with healthy volunteers and professional athletes we know that MI can improve physical performance and learning (Mulder et al.,
Reviews from the field of neurorehabilitation highlighted the potential neural correlates and the effect of specific brain lesions on MI ability and therefore its consequences in the rehabilitation of patients after stroke or with Parkinson's disease (PD) (McInnes et al.,
Different theories try to explain why muscle activity signals might be detectable during MI (Guillot et al.,
Therefore, in the present study, we aimed to investigate whether EMG activation could be detected during MI of an upper limb task in a chronic stroke population (Dickstein et al.,
The movement investigated by Dickstein et al. was an alternating, bilaterally executed, lower limb task involving closed muscle chains and only few degrees of freedom. In contrast, we were interested in a hand grasping and arm-lifting task, which is a clinically relevant, unilateral executed movement task using open muscle chains. Therefore, the main goal of the present study was to explore the muscle activation of an upper limb task in patients after stroke or with PD during three conditions: MI, PE, and rest. Further, the findings were compared to the EMG activation in healthy volunteers. We hypothesised that an EMG activation during MI and PE of the hand grasping and arm-lifting task would be detectable. Furthermore, based on the findings from Dickstein et al. we expected that EMG activation during MI would differ from the rest condition for all healthy volunteers, patients after stroke, and patients with PD due to their subcortical lesion. All participants underwent a battery of MI ability assessments including mental rotation, mental chronometry, and the Kinaesthetic and Visual Imagery Questionnaire. As a subgoal, we analysed personal data and MI ability assessments of participants, which showed EMG activation during MI.
The study comprised a cross-sectional investigation with two measurement sessions. On the first measurement session participants underwent a cognitive, handedness, and MI ability screening. Additionally, participants were screened regarding their ability to perform a hand grasping and arm-lifting task, and they were introduced to MI practice. On the second measurement session EMG activation was recorded during the hand grasping and arm-lifting task under two conditions: MI and PE. The study was conducted in accordance with the Declaration of Helsinki and was approved by the ethics committee of the Canton Aargovia, Switzerland (Ref. Nr. EK: 2013/034).
Selection criteria for all three participant groups are presented in Table
Participant selection criteria.
All participants |
- Males and females older than 18 years - Able to sit independently with closed eyes on a normal chair - Able to perform the hand grasping and arm lifting task without external help - Provide written informed consent |
- Additional neurological, psychological, or psychiatric disease, severe pulmonary and cardiovascular diseases - Severe pain - Severe deformation of joints of the upper limb with arthritic origin - Present impairments in cognition and communication |
Healthy participants |
- No neurological or psychological disease |
|
Patients after stroke |
- Patients in the subacute or chronic phase after first-ever stroke - Present an arm and hand paresis |
|
Patients with Parkinson's disease (PD) |
- Patients with an idiopathic PD - No treatment with deep brain stimulation |
Data were collected in the rehabilitation centre Reha Rheinfelden in Switzerland between August and October 2013. Participants were tested individually (Figure
Measurement sessions and study procedure. EMG: electromyography; h: hour; MI: motor imagery; min: minute.
To assess cognitive function, the
To assess hand laterality, the
The
Participants had to perform a hand grasping and arm-lifting task with the dominant hand in healthy volunteers and with the more affected hand in patients after stroke and with PD. We were interested if we could detect muscle activation in the paretic limb during MI (Dickstein et al.,
To evaluate participants ability to create a mental image and therefore to be included in the study, participants
Pictures of hands and feet (Moseley,
The KVIQ was developed for patients with sensorimotor impairments. The KVIQ is a valid and reliable questionnaire that can be applied in healthy volunteers, patients after stroke, or with PD (Malouin et al.,
MC is a reliable method to examine the temporal structure of MI in healthy volunteers and post stroke patients (Malouin et al.,
At the end of the first measurement participants were introduced theoretically and practically to the concept of MI. A 30 min session was given based on the MI introduction programme by Wondrusch and Schuster-Amft (
Participants' skin preparation and electrode placement were based on the Surface Electromyography for the Non-Invasive Assessment of Muscles (SENIAM) recommendations (Hermens et al.,
EMG signals were recorded using the wireless device Myon320 (Prophysics AG, Zurich, Switzerland). Data were collected with a sampling frequency of 3 kHz and pre-amplified by a factor of 1,000. During the recordings, signals were displayed using LabView (Service Package 1, 2011, National Instruments, Austin, USA). The experimental protocol was adopted from Dickstein et al. (
Overview on recorded EMG activation. Each participant had to perform 3 blocks of PE followed by 3 blocks of MI. Whereas, one block consist of a 15 s start period, three hand grasping and arm lifting movements and a 15 s end period. Black vertical lines represent metronome beats and grey line displays manual trigger. EMGm electromyography; mV, millivolt, s: seconds.
Our sample size was based on the investigation of Dickstein et al. The authors included nine healthy volunteers and six patients after stroke (Dickstein et al.,
Each participants' average EMG activation was determined in the three task trials of one block for all four muscles. Each recording was time normalised to 1,000 points. Average EMG activation during MI, PE, and rest was calculated for each muscle and each participant. The Wilcoxon signed rank test was applied as
Each trial was analysed individually:
the MI amplitude average in each task block was compared to the average amplitude of the corresponding rest condition; activation during MI was verified by comparing the processed EMG signal against a threshold of 1.5 times of the standard deviation (SD) of the processed EMG signal during the rest condition.
The EMG activation threshold of 1.5 × SD was selected based on Hodges and Bui (
The results of each assessment (MR, MC, KVIQ) were processed individually. For MR, the number of correctly identified hand and feet flash cards was counted. For MC, the ratio (MI/PE) of the time needed to imagine and physically execute the task was calculated. For the KVIQ-10, the average scores of the visual and kinaesthetic subscales were calculated.
The average values over all trials of the visual and kinaesthetic subscales were analysed and are presented as mean and standard deviation.
Statistical analyses were performed using MATLAB (R2012b, MathWorks, Natick, USA,
In total, 22 out of 24 participants (11 females and 11 males) completed both measurement sessions and were included in the data analyses. Two PD patients had to be excluded because they did not meet the previously described inclusion criteria of the MI ability assessments. One of the two excluded PD patients did not achieve 30 points in the KVIQ-10 assessment and achieved a ratio higher than 1.5 in the mental chronometry assessment. The second PD patient only recognised 55% of the hand and feet pictures and was not able to perform the KVIQ-10. Included participants comprised 10 healthy subjects, seven patients after the first-ever stroke and five patients with PD (Tables
Group characteristics.
Age | 65.4 ± 6.0 |
53.7 ± 16.3 |
45.4 ± 15.4 |
52.6 ± 15.8 | 6.54 | |
Time of disease [M] | 60.4 ± 24.5 |
16.3 ± 24.8 |
n.a. | n.a. | n.a. | n.a. |
MMSE | 28.6 ± 1.1 |
28.7 ± 0.8 |
29.2 ± 0.8 |
28.9 ± 0.9 | 1.95 | 0.378 |
MI self-rating (max. 5) | 3.0 ± 1.0 |
3.3 ± 0.7 |
3.3 ± 0.6 |
3.2 ± 0.7 | 0.53 | 0.766 |
MI practice trials | 7.2 ± 4.8 |
6.4 ± 4.9 |
6.2 ± 1.5 |
6.5 ± 3.5 | 0.18 | 0.914 |
Mental rotation (0-64) | 61.0 ± 1.0 |
57.1 ± 4.7 |
63.2 ± 1.1 |
60.8 ± 3.8 | 11.07 | |
Mental chronometry PE [s] | 4.9 ± 2.3 |
3.9 ± 0.8 |
3.5 ± 0.7 |
4.0 ± 1.3 | 3.45 | 0.178 |
Mental chronometry MI [s] | 5.4 ± 4.5 |
3.9 ± 0.8 |
3.3 ± 0.9 |
4.0 ± 2.2 | 1.8 | 0.406 |
Ratio Mental chronometry ( |
1.1 ± 0.28 |
1.03 ± 0.28 |
0.97 ± 0.30 |
0.99 ± 0.28 | 0.19 | 0.910 |
KVIQ visual (5-25) | 17.4 ± 2.9 |
18.0 ± 5.2 |
18.5 ± 4.3 |
18.1 ± 4.1 | 0.32 | 0.852 |
KVIQ kinaesthetic (5–25) | 16.4 ± 2.5 |
14.4 ± 5.0 |
14.2 ± 3.9 |
14.8 ± 4.0 | 1.21 | 0.545 |
KVIQ kinaesthetic + visual (10–50) | 33.8 ± 4.8 |
32.4 ± 7.8 |
32.7 ± 4.7 |
32.9 (±5.6) | 0.32 | 0.854 |
3 | 7 | 31 | R | 1 | 5 | 3 | L | 27 | 60 | 4.27 | 3.02 | 0.71 | 16 | 15 | 31 | 3.0 | 1.3 | |
3 | 10 | 91 | L | 1 | 1 | 4 | L | 28 | 62 | 3.51 | 2.77 | 0.79 | 19 | 18 | 37 | 4.0 | 4.7 | |
3 | 10 | 56 | R | 1 | 5 | 5 | R | 29 | 62 | 4.32 | 3.88 | 0.90 | 20 | 20 | 40 | 3.7 | 3.7 | |
2 | 6 | 2 | L | n.a. | 1 | 4 | R | 29 | 53 | 3.49 | 5.21 | 1.49 | 11 | 14 | 25 | 3.0 | 2.7 | |
2 | 1 | 3 | R | n.a. | 3 | 2 | R | 28 | 57 | 4.17 | 3.13 | 0.75 | 13 | 6 | 19 | 2.0 | 2.0 | |
2 | 8 | 7 | R | n.a. | 1 | 6 | R | 30 | 54 | 3.02 | 3.95 | 1.31 | 21 | 20 | 41 | 3.3 | 4.0 | |
2 | 9 | 23 | L | n.a. | 1 | 4 | R | 29 | 60 | 3.37 | 3.61 | 1.07 | 17 | 17 | 34 | 3.7 | 2.7 | |
1 | 8 | n.a. | n.a. | n.a. | 5 | 4 | R | 30 | 62 | 2.69 | 2.61 | 0.97 | 17 | 13 | 30 | 3.0 | 2.0 | |
1 | 1 | n.a. | n.a. | n.a. | 2 | 4 | R | 30 | 64 | 3.61 | 3.78 | 1.05 | 18 | 21 | 39 | 4.0 | 4.3 | |
1 | 9 | n.a. | n.a. | n.a. | 4 | 5 | R | 29 | 63 | 3.43 | 1.97 | 0.57 | 24 | 16 | 40 | 4.0 | 4.0 | |
1 | 8 | n.a. | n.a. | n.a. | 4 | 4 | R | 28 | 61 | 4.21 | 3.97 | 0.94 | 24 | 9 | 33 | 5.0 | 2.3 | |
1 | 2 | n.a. | n.a. | n.a. | 2 | 5 | R | 29 | 64 | 4.18 | 4.09 | 0.98 | 13 | 14 | 27 | 3.7 | 2.7 | |
1 | 7 | n.a. | n.a. | n.a. | 2 | 6 | R | 29 | 64 | 2.92 | 2.95 | 1.01 | 15 | 14 | 29 | 3.3 | 2.7 | |
3 | 10 | 79 | L | 1.5 | 2 | 1 | R | 30 | 61 | 3.69 | 4.02 | 1.09 | 19 | 14 | 33 | 3.0 | 3.0 | |
3 | 10 | 45 | L | 1 | 2 | 1 | R | 29 | 60 | 8.93 | 13.39 | 1.50 | 13 | 15 | 28 | 2.3 | 1.3 | |
2 | 10 | 70 | L | n.a. | 1 | 6 | R | 29 | 61 | 5.46 | 4.28 | 0.78 | 16 | 19 | 35 | 4.0 | 3.3 | |
2 | 5 | 3 | L | n.a. | 1 | 5 | R | 28 | 64 | 3.36 | 2.79 | 0.83 | 24 | 10 | 34 | 3.3 | 3.0 | |
2 | 11 | 6 | R | n.a. | 2 | 1 | L | 28 | 51 | 4.08 | 4.06 | 1.00 | 24 | 15 | 39 | 4.0 | 4.0 | |
1 | 5 | n.a. | n.a. | n.a. | 2 | 1 | L | 30 | 64 | 2.45 | 3.71 | 1.51 | 21 | 17 | 38 | 4.0 | 3.7 | |
1 | 8 | n.a. | n.a. | n.a. | 4 | 4 | R | 28 | 64 | 3.44 | 4.80 | 1.40 | 20 | 9 | 29 | 2.7 | 3.0 | |
1 | 1 | n.a. | n.a. | n.a. | 1 | 5 | R | 29 | 64 | 4.50 | 2.87 | 0.64 | 21 | 11 | 32 | 4.0 | 3.7 | |
1 | 6 | n.a. | n.a. | n.a. | 3 | 5 | R | 30 | 62 | 3.82 | 2.67 | 0.70 | 12 | 18 | 30 | 2.0 | 4.0 |
Location of brain lesions for the seven included patients after stroke.
6 | Corona radiate, right | hemorrhagic | |
7 | Basal ganglia | ischemic | |
8 | Thalamus and tectum mesencephali | hemorrhagic | |
9 | Area arteria cerebri media and left arteria cerebri anterior | hemorrhagic | |
10 | Frontal cortex, left | hemorrhagic | |
11 | Area of arteria cerebri media, left | hemorrhagic | |
12 | Arteria cerebri media, right | hemorrhagic |
An overview on the EMG activation during MI and PE is provided in Figure
Averaged EMG activation during physical execution and MI within each muscle:
EMG activation during physical execution and MI of four healthy volunteers.
EMG activation during physical execution and MI of three patients after stroke.
EMG activation during physical execution and MI of two patients with Parkinson's disease.
The
Table
The seven participants, who showed an EMG activation during MI had a
For
Five of the seven individuals showed a MI/PE ratio within one standard deviation (±SD = 0.28) of all participants (mean = 1.05) in
The total
Please see Tables For For the All participants scored for the visual subscale of the
No significant correlation was found between the average EMG activation signal during MI and the scores of MI ability assessments (MR
The presented study evaluated EMG activation during kinaesthetic MI and PE of a clinically meaningful upper limb hand grasping and arm lifting task, which is an unilaterally executed movement task using open muscle chains. M. deltoideus pars clavicularis and M. biceps brachii showed significantly higher activation during MI than during the rest condition. We observed EMG activations during PE in all participants and an activation above an EMG signal threshold during MI in seven out of 22 participants: two healthy volunteers, three patients after stroke, and two patients with PD. Furthermore, we provided detailed results of all participants' MI ability assessments. However, MI ability assessments showed no correlation among each other and were not correlated with EMG activation.
The link between EMG activation and MI was reported with inconsistent findings in the literature. Our results are in line with studies of healthy volunteers (Bakker et al.,
An interesting observation of our study is that patients, who showed EMG activation, did so partially with a time lag, i.e., the activation lasted beyond the end trigger and declined 5 to 10 s later.
McAvinue and Robertson postulated that MI vividness is not an one-dimensional ability (McAvinue and Robertson,
Mental chronometry showed comparable MI/PE ratios and a close temporal congruency in patients and healthy volunteers on a moderate to good level of MI ability. These results are in line with findings by Malouin et al. who compared the imagination and execution of stepping movements in patients after stroke and healthy volunteers (Malouin et al.,
Physiological responses during MI were investigated in various brain imaging investigations that included healthy volunteers and patients after stroke. Sharma et al. systematically analysed five functional magnetic resonance imaging (fMRI) studies focusing on imagined and executed hand and arm movements (Sharma et al.,
One limitation of this study is the small number of participants in each group. Therefore, we avoided group comparisons. It must be emphasised that our study population was a pragmatic convenience sample and were not aged or gender matched, and healthy volunteers were younger than participants in both patient groups. However, seven out of 21 participants showed muscle activation pattern during MI above the threshold of a meaningful and clinically relevant unilateral executed hand grasping and arm lifting task using open muscle chains (2 healthy volunteers, 3 patients post stroke, 2 patients with PD). Our ratio of participants with EMG signal patterns during MI is lower than the one of Dickstein et al. (Dickstein et al.,
It could be argued that EMG measurements should have been performed during the MI ability assessments too and correlated with the MI ability assessments scores. However, as all patients were inpatients, we intended to keep the measurement/assessment time and burden for the patients as low as possible.
More distinct EMG signal patterns may have been obtained with a task requiring larger muscle activation, for example using a cup with more weight, since studies have shown that EMG activation during MI is proportional to the imagined weight lifted (Guillot et al.,
To familiarise with the MI technique and the hand grasping and arm-lifting task, participants were asked to practice the task two to three times per day between first and second appointment. It would have been interesting to know how many times participants practiced the task mentally between both appointments. Experience from previous MI investigations with patients suggests that the number of mental trials did not exceed the recommended amount of approximately 30 trials (Schuster et al.,
To summarise, seven out of 22 participants (two healthy volunteers, three patients after stroke and two patients with PD) showed an EMG activation during MI of the hand grasping and arm-lifting task in at least one of the target muscles exceeding the threshold. These findings should be confirmed in future investigations, as with the technological and scientific advances MI-induced EMG may eventually become useful in neurorehabilitation. Results of the present study suggest that subliminal EMG activation might be present in seven out of 22 participants (ratio of 1:3.1) in healthy volunteers, patients after stroke and patients with PD. Inconsistent EMG activations may be explained by individual variations. We provided detailed results of all patients' MI ability and could not find a correlation between EMG activation during MI and results of three MI ability assessments.
MK, BW and CS-A designed the study, helped with the ethics committee application and the interpretation of the statistical analyses and intensively revised the manuscript draft for important intellectual content. Additionally, MK conducted the data. All authors gave final approval of the manuscript to be published and are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The authors would like to thank Dr. Rolf van de Langenberg and Dr. Elke Heremans for their valuable advice regarding EMG measurements and data analysis, and MI ability assessments. We are grateful to Dr. Michael McCaskey for advice in conducting the study, performing the data analysis, and his critical comments on an earlier version of the manuscript, to Matthias Schindelholz for compiling and implementing the EMG recording software, to Prof. Kenneth J. Hunt for provisioning the EMG measurement devices, and all participants for their time and support.