Journal of Electromyography and Kinesiology 23 (2013) 430–437
Contents lists available at SciVerse ScienceDirect
Journal of Electromyography and Kinesiology journal homepage: www.elsevier.com/locate/jelekin
Synergistic co-activation in multi-directional postural control in humans Hiroaki Imagawa a, Shota Hagio b, Motoki Kouzaki a,b,⇑ a b
Faculty of Integrated Human Studies, Kyoto University, Kyoto, Japan Laboratory of Neurophysiology, Graduate School of Human and Environmental Studies, Kyoto University, Kyoto, Japan
a r t i c l e
i n f o
Article history: Received 19 August 2012 Received in revised form 28 September 2012 Accepted 5 November 2012
Keywords: Bipedal standing Center of pressure Electromyogram
a b s t r a c t To examine the muscle synergies of multi-directional postural control, we calculated the target-directed variance fraction (g) and net action direction of each muscle using the electromyogram-weighted averaging (EWA) method. Subjects stood barefoot on a force platform and maintained their posture by producing a center of pressure (COP) in twelve target directions. Surface electromyograms were recorded from 6 right-sided muscles: tibialis anterior (TA), soleus (SOL), lateral gastrocnemius (LG), medial gastrocnemius (MG), fibularis longus (FL), and gluteus medius (GM). g was calculated from COP with duration of 20-s, during which the COP was relatively constant. The EWA method was applied to the EMG and the two COP components to estimate the net action direction of each muscle. The results showed that g values in all directions did not cross the 0.8 threshold. This suggests that human postural control is achieved by synergistic co-activation. The EWA revealed that the net action directions of TA, SOL, LG, MG, and GM were 277.6°, 71.1°, 87.7°, 94.0°, and 2.2°, respectively. This suggests that postural maintenance by muscle synergy can be attributed to the relevant muscles having various action directions. These results demonstrate that muscle synergies can be investigated using COP fluctuations. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Human bipedal standing is inherently unstable because, in standing, a large body mass with a high elevation center is kept in an erect posture over a relatively small base of support. Therefore, we easily lose balance given a very little external force or disturbance in everyday life and face challenges of equilibrium not only from the anterior–posterior directions but also the lateral directions. To avoid falling down, the body’s center of mass needs to be positioned over the base of support. Although there are various approaches to revealing how postural control can be achieved, for example kinematic (Kluzik et al., 2007), simulated (Peterka and Loughlin, 2004), and physiological (Collins and De Luca, 1993) approaches, the mechanisms of postural control are still the subject of discussion. Humans can accomplish many different motor tasks by transmitting signals from the central nervous system (CNS) to the appropriate muscles involved in a particular motor task. Performing these motor tasks would be very complex if the CNS had to control each muscle individually to generate the desired movements. The concept of ‘muscle synergy’, a form of modular organization, has therefore been proposed to simplify these motor controls. ⇑ Corresponding author at: Laboratory of Neurophysiology, Graduate School of Human and Environmental Studies, Kyoto University, Yoshida-nihonmatsu-cho, Sakyo-ku, Kyoto 606-8501, Japan. Tel./fax: +81 75 753 2927. E-mail address:
[email protected] (M. Kouzaki). 1050-6411/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jelekin.2012.11.003
Muscle synergies are patterns of activation among multiple muscles involved in controlling movements, acting about the relevant degree of freedom. This decreases the amount of information the CNS has to process. Muscle synergies have been proposed in the implementation of many types of motor tasks, for example, postural tasks in cats (Ting and Macpherson, 2005), isometric endpoint force by the index finger (Kutch et al., 2008), and multi-directional reaching tasks (Muceli et al., 2010). Some earlier studies have investigated muscle synergies in postural control using surface translations (Henry et al., 1998; Torres-Oviedo and Ting, 2007; Chvatal et al., 2011). In such studies, muscle synergies in postural control were investigated using mainly electromyogram (EMG) data. However, the variability of postural control should be considered for examining muscle synergy during bipedal standing because center of pressure (COP) fluctuations reflect postural control dynamics (Collins and De Luca, 1993). Kutch et al. (2008) found that muscle coordination of the index finger could be examined from force fluctuations in voluntary isometric endpoint force production. This implies that muscle synergies in human bipedal standing could also be investigated by COP fluctuations during postural control because the COP variability reflect the fluctuations in torque around the ankle joint during quiet standing (Masani et al., 2003; Morasso and Schieppati, 1999). To examine fluctuations occurring during postural control using two analysis methods described below, the present study used tasks in which subjects voluntarily lean and maintain the position of the COP in multiple directions in the horizontal plane (in the
H. Imagawa et al. / Journal of Electromyography and Kinesiology 23 (2013) 430–437
left–right and anterior–posterior directions) by modulating endpoint COP. To evaluate muscle coordination in postural control, the present study used the variability of COP against target directions as an indicator of underlying muscle coordination. Kutch et al. (2008) examined muscle coordination by mapping isometric endpoint force variability for an array of targets distributed uniformly across the endpoint force space. This endpoint force variability in continuous isometric force production has been suggested to depend on muscle action direction and on how these muscles are coordinated. Thus, the variability of the endpoint force would be a linear superposition of individual contributions of each muscle, if the endpoint force is controlled by multiple muscles through synergistic coactivation; whereas, it would fluctuate along the target direction if controlled by a single muscle: a ‘prime mover’. In this study, we used COP variability in the target direction instead of endpoint force variability. Using this method, we were able to quantify the relevant amount of COP variability that occurs in the target direction and estimate whether postural control in each direction was achieved by synergistic co-activation. The relevant amount of variability was evaluated using ‘‘g’’. If g has a low value, we estimate that postural control in a given direction is achieved through synergistic co-activation. In addition to COP variability, we examined which muscles are involved in postural control in each direction by the EMG-weighted averaging (EWA) method (Kutch et al., 2010). The EWA trajectory should theoretically yield information about the net action direction of each muscle in multi-directional postural control. In the previous ‘index finger study’, the net action direction of the first dorsal interosseous and extensor indicis muscles was estimated using the EWA method (Kutch et al., 2010). This suggests that the EWA method is capable of correctly evaluating the net action direction of a muscle during multi-directional postural control. The aims of this study were (1) to investigate muscle synergies in multi-directional bipedal standing from endpoint COP fluctuations occurring during postural control, and (2) to identify muscles contributing to postural control in each direction. To examine these aims, we employed postural task that subjects lean in a given direction and maintain the end-point position because of the characteristics of analysis methods, target-directed variance fraction and EWA, used in this study.
431
5KNSA13B, Kyowa, Tokyo, Japan) with their arms comfortably lying to the side and as little distance between their feet as possible (Fig. 1). During multi-directional postural control, the foot center of pressure (COP) position was measured. Each subject was instructed to gradually shift their position of COP from the original position to the target point by leaning his body around the ankle joint and then to hold the target COP position as precisely as possible for approximately 25 s. They were also instructed to change the angles of hip and knee joints as little as possible during trials. The target directions include 12 different directions distributed equally over the horizontal plane at 30° increments, and the distance between the target COP position and the original position was 30 mm since, in this study, we focused on not low frequency COP fluctuations seen during quiet standing but high frequency component of them (>5 Hz). For each trial, they viewed the desired COP position as a target on a visual display in front of them. The order of the twelve trials was randomized, and a sufficient resting period was allowed between trials to prevent fatigue. The trial was repeated if investigators made the decision that the hip and knee joint angles had changed, and the COP displacement was not constant at the target position. A rest trial was collected for each subject to establish an EMG baseline when they were seated on a chair. Surface EMGs were recorded from the following six right-sided muscles: tibialis anterior (TA), soleus (SOL), lateral gastrocnemius (LG), medial gastrocnemius (MG), fibularis longus (FL), and gluteus medius (GM). The surface EMG of GM was recorded to check the contribution of hip strategy during postural control since biomechanical study has indicated that postural maintenance in medial–lateral direction is controlled under hip strategy (Winter et al., 1996). The EMGs were recorded using Ag–AgCl electrodes with a diameter of 5 mm and an inter-electrode distance of 10 mm. We used a small inter-electrode distance to prevent cross talk among neighboring muscles. A reference electrode was placed on the external malleolus. The EMG signals were amplified (MEG6116M, Nihon-kohden, Tokyo, Japan) with band-pass filtering between 5 and 1000 Hz. All electrical signals were stored with a sampling frequency of 2000 Hz on the hard disk of a personal computer using a 16-bit analog-to-digital converter (PowerLab/16SP; AD Instruments, Sydney, Australia). 2.3. Data analysis
2. Methods 2.1. Subjects Nine healthy males participated in this study. Their mean (±SD) age, height, and body mass were 22.1 ± 1.5 yrs, 170.9 ± 6.0 cm, and 67.1 ± 10.7 kg, respectively. All subjects were healthy, had no history of any neurological disorder, and had vision that was corrected to normal levels. They gave their written informed consent for this study after receiving a detailed explanation of the purpose, potential benefits and risks concerned with participating in the study. The experimental procedures used in this study were in accordance with the Declaration of Helsinki and were approved by the Committee for Human Experimentation at the Graduate School of Human and Environmental Studies, Kyoto University. 2.2. Experimental protocol and measurement The basic procedure for the setup and measurement of postural sway during quiet standing has been described in our previous studies (Kouzaki and Masani, 2008, 2012; Kouzaki and Shinohara, 2010). The subjects stood barefoot on a force platform (EFP-S-1,
For all recorded signals, data for a 20 s period were selected for analysis of individual trials where the COP trajectory was relatively constant. The COP data were filtered with a zero-phase-lag, fourthorder Butterworth band-pass filter at 5–30 Hz to remove the effect of postural sway induced by voluntary contribution and non-physiological noise (Kutch et al., 2008). Raw EMG data were filtered with a zero-phase-lag, fourth-order Butterworth high-pass filter at 20 Hz to reduce the baseline noise and movement artifacts (De Luca et al., 2010). EMG traces were then rectified and averaged across the extracted period and the rest period, with differences between the two serving as the net EMG. To examine how the CNS controls muscles accomplish postural control in multiple directions, we quantified the COP variability that occurred in the target direction. The degree to which the COP variability for each trial was aligned with each trial’s target direction was quantified as the fraction of the total variance (g) of COP that occurs in the target direction, according to the previous study by Kutch et al. (2008). g was calculated from following equation:
h i bt e b D target cov D D target n h io g¼ ~ Trace cov D
432
H. Imagawa et al. / Journal of Electromyography and Kinesiology 23 (2013) 430–437
90º
0º
180º
270º Fig. 1. Experimental arrangement. Subjects stood barefoot on the force platform with their arms lying by their sides. Twelve target directions were uniformly distributed over the horizontal plane at 30° increments. The center of pressure (COP) in the anterior–posterior and left–right directions and target positions for visual feedback were displayed on a 16-in. monitor in front of the subject. The subject exerted his COP to move the cursor (indicated by a dot) towards the target position (open circle).
where the denominator is the sum of the amount of total variance as a scalar and the numerator quantifies the amount of variance b target was defined as the unit vecoccurring in the target direction. D tor of the empirical target force vector Dtarget, defined to be equal to the average COP vector for each trial. If COP variability from the target direction is large, which suggests that postural control in that direction is controlled by synergistic co-activation, g would be small. Conversely, if COP variability is almost aligned with the target direction, suggesting that postural control in that direction was controlled by a ‘prime mover’, g would be large. Using g, we investigated whether human postural maintenance was controlled by synergistic co-activation. We also investigated which muscles were involved in postural control in each direction by estimating the net action direction of each muscle according to the EMG-weighted averaging (EWA) method (Kutch et al., 2010). The EWA is based on a cross-correlation of EMGs and COP signals (Fig. 2). Such analysis was performed over an approximately steady period of COP in both directions lasting 18 s out of the time course used in prior analysis. We used surface EMG recordings from a muscle EMGm (t) (t denotes discrete time point), along with two COP components (COPL–R: left–right direction, COPA–P: anterior– posterior direction) for cross-correlation. The EWA trajectory was calculated using following equations:
EWAmL—R ðtÞ ¼
X
COPL—R ði þ tÞEMGm ðtÞ
i
EWAmA—P ðtÞ ¼
X
COPA—P ði þ tÞEMGm ðtÞ
i
where summation was carried out over the extracted time intervals. The EWA trajectory in the horizontal plane was shifted so that it started from the origin at the 0 time lag. Each EWA was quantified temporally and spatially based on a time lag from 0 to 100 ms, dur-
ing which the EWA trajectory reached its first peak magnitude. We used this time lag to define the EWA time-to-peak and used the corresponding spatial direction to define the EWA direction of each trial.
3. Results Subjects maintained approximately constant COP around the target position in all directions. The COP variability in the target direction for one representative subject is shown in Fig. 3. It appears that COP variability was roughly dependent on the target direction. The values of g for individual subjects and the average value in each direction are shown in Fig. 4. For almost all target directions, especially the lateral directions, the COP variability was relatively broad and little aligned with the target direction. The higher g was found in directions of 90°, 120°, 270° and 300° with average g values of 0.68 ± 0.08, 0.67 ± 0.08, 0.67 ± 0.17, and 0.62 ± 0.09, respectively (±SD). On the other hand, g was lower in directions of 0°, 180°, 210°, and 330° with average values of 0.36 ± 0.17, 0.38 ± 0.14, 0.31 ± 0.09, and 0.43 ± 0.12, respectively. To compare the g value from each direction, we performed oneway ANOVA, which revealed that there were significant differences between g in the pure anterior–posterior direction and in the pure lateral direction (p < 0.05). Polar plots of normalized EMG data and polar histograms of the EWA direction of all trials of TA, SOL, LG, MG, FL, and GM are shown in Figs. 5 and 6, respectively. The range where EWA directions were distributed in the task plane was focused near a certain direction, although muscles were active in relatively wide range (Fig. 5). TA was active mainly in the backward half, while the plantar flexor muscle group, i.e., SOL, LG, and MG was mainly active in rightanterior quarter of the task plane. The GM was active in the right
433
H. Imagawa et al. / Journal of Electromyography and Kinesiology 23 (2013) 430–437
forward 90º
(A) Filtered data COP
L-R
0.05 mm
COP
A-P
0.05 mm
EMG
180º
0º
0.05 mV 0
20
(B) EWA waveforms
40 (s)
(C) EWAdirection anterior
left-right
270º backward anterior -posterior left
right posterior
Fig. 2. A representative example of the electromyogram-weighted averaging (EWA) method. (A) Filtered data for COP (left–right), COP (anterior–posterior), and rectified electromyogram (EMG) recorded from TA. (B) EWA waveforms peaking at a time lag of 45 ms obtained from a cross-correlation of rectified EMG and COP components. (C) The EWA direction was represented by plotting each correlation value at the EWA time-to-peak.
forward 90º
Fig. 4. Target-directed variance fraction (g). Each point represents a target, with direction given by the target direction and magnitude given by g. The solid red line represents g averaged across all subjects in each direction.
measured counterclockwise from the pure right direction with a 95% confidence interval for the average. In contrast, the polar histogram of EWA directions for FL had bipolar directions, showing peaks at approximately 112.5° and 247.5°. The EWA time-to-peak histograms for all muscles examined in this study exhibited a first peak with a time lag from 0 to 100 ms for most trials, exemplified by TA, SOL, LG, MG, FL, and GM in Fig. 7. The average time lags of the first peak were 42.2 ± 10.7 ms, 43.5 ± 14.1 ms, 45.6 ± 13.0 ms, 41.2 ± 12.0 ms, 46.4 ± 14.8 ms, and 47.2 ± 13.7 ms, respectively (±SD). However, in certain muscles, especially lower limb muscles, the EWA time-to-peak histogram exhibited a second peak at approximately 90 ms (e.g., TA: 85.6 ± 9.87 ms, SOL: 89.0 ± 7.04 ms, MG: 87.2 ± 11.3 ms). 4. Discussion
180º
0º
270º backward Fig. 3. COP variability in all target directions from one representative subject. The variability is magnified by a factor of 15 for visualization to show that the subject performed tasks as instructed by the experimenter. h is the angle between the target direction and the right axis.
direction. In contrast, the activities of FL widely spread in all directions. The polar histogram of EWA directions of almost all muscles showed a monopolar direction, whereas that of FL showed bipolar directions (Fig. 6). The average EWA directions for TA, SOL, MG, LG, and GM, which showed a monopolar direction, were 277.6°, 71.1°, 87.7°, 94.0°, and 2.2°, respectively. These average directions were
We have demonstrated COP variability in target directions during bipedal standing and its relation to the activities of relevant muscles according to the EWA method (Kutch et al., 2010). The main results were that (1) the average value of g was lower than the 0.8 threshold in all directions, (2) the net action direction of lower limb muscles could be estimated, and (3) two EWA timeto-peaks were found at 50 and 90 ms. Our novel findings were that synergistic activation of multi-directional postural control could be revealed by endpoint COP fluctuations along with the net action direction of relevant muscles for bipedal standing. 4.1. Implication for muscle synergies in human bipedal standing We investigated muscle synergies from endpoint COP fluctuations occurring during postural control with various directions. Target-directed variance fraction showed that g was different depending on target directions (Fig. 4). Moreover, although a relatively high value of g was found in anterior–posterior directions (90°, 120°, and 270°), there was no direction where g was above 0.8, which was set as a threshold in the previous study (Kutch et al., 2008). These results suggest that human postural control in the horizontal plane was achieved by synergistic co-activation of multiple muscles. As shown in Fig. 5, most of the lower limb muscles were at peak activity in the diagonal directions. If we investigate
434
H. Imagawa et al. / Journal of Electromyography and Kinesiology 23 (2013) 430–437
TA
1
SOL
1
LG
1
MG
1
TA
SOL
LG
MG
FL
GM
90º
GM
FL
90º 1
1
0º
180º 180º
0º
270º
270º Fig. 5. Polar plots of EMG activity for tibialis anterior (TA), soleus (SOL), lateral gastrocnemius (LG), medial gastrocnemius (MG), fibularis longus (FL), and gluteus medius (GM) for all trials in all subjects. Each plot represents EMG data normalized by the maximum across all directions for each subjects. The solid red line represents EMG activity averaged across all subjects.
muscle synergies merely from muscle activity (Fig. 5), they appear to be organized among muscles active in the same direction. For example, postural control in our experimental condition is thought to be achieved by co-activation of SOL, LG, and MG in the rightanterior direction, but is controlled by TA as a ‘prime mover’ in the posterior direction. However, the g showed that in all directions in the horizontal plane, postural control was achieved by synergistic co-activation. This suggests that it is insufficient to investigate muscle coordination merely from EMG because many muscles are relevant to erect bipedal posture, and muscle synergies need to be examined from another component in addition to EMG. 4.2. Muscle contribution to postural control The present study used the EWA method proposed by Kutch et al. (2010) to investigate the contribution of individual muscles to postural control in each direction. We also estimated the net action direction of these muscles. Almost all muscles have a net action direction in either the anterior–posterior or diagonal direction (Fig. 6). This result agrees with the observation that g values in the anterior–posterior direction were significantly different from g values in the lateral directions. Although Fig. 6 indicates that GM would have a net action in the right direction, g in the right direction was low, suggesting that muscles other than GM contributed to keep COP position at the target direction (0°). This implies that postural control in our experimental task was imple-
Fig. 6. Polar histograms of EWA direction for TA, SOL, LG, MG, FL, and GM for all trials in all subjects. The radius represents trail counts and the solid line represents normalized EMG data averaged across all subjects from Fig. 5. The average net action directions of TA, SOL, LG, MG, and GM are 277.6°, 71.7°, 87.7°, 94.0°, and 2.2°, respectively.
mented using an ankle strategy. Moreover, the small contribution of GM to postural control in the right direction could not be estimated from EMG activity alone. If we estimate muscle contribution in accordance with the previous study (Kutch et al., 2010), GM was estimated to be a ‘prime mover’ in postural control in the right direction, which contradicts our results. In motor tasks such as postural control, in which many muscles are involved, the degree of contribution of each muscle could not be estimated from merely the EWA directions. However, in this study, some muscles have similar net action directions, e.g., SOL, LG, and MG. Referring back to the previous study by Kutch et al. (2008), if g is small, it suggests that the motor task is controlled by synergistic co-activation. However, there is a possibility that if muscles similar action directions organize muscle synergy, then g would be large in spite of the existence of muscle synergy, e.g., SOL, LG, and MG. Therefore, in motor tasks in which many muscles are involved, it would also be insufficient to investigate muscle coordination merely from g. Therefore, muscle contributions to such motor tasks need to be investigated by comparing the g and EWA directions. Using such a comparison, postural control in the anterior–posterior direction is suggested to occur by the co-activation of SOL, LG, and MG. Furthermore, postural control in the right direction is suggested to occur, not by the ‘prime mover’ GM, but through co-activation of TA, SOL, and MG. 4.3. EWA directions from different time-to-peaks When we used the EWA method, all muscles examined in this study showed an EWA time-to-peak with a time lag of approximately
H. Imagawa et al. / Journal of Electromyography and Kinesiology 23 (2013) 430–437
435
Fig. 7. Histograms of EWA time-to-peak for TA, SOL, LG, MG, FL, and GM for all trials and all subjects. Average time lags for first peaks are 42.2 ± 10.7 ms, 43.5 ± 14.1 ms, 45.6 ± 13.0 ms, 41.2 ± 12.0 ms, 46.4 ± 14.8 ms, and 47.2 ± 13.7 ms respectively.
50 ms. However, in some muscles, especially the lower limb muscles, a second time-to-peak occurred with a time lag of approximately 90 ms, which was not shown in a previous study (Kutch et al., 2010). This indicates that two different components are contained in fluctuations observed during postural maintenance. When we plotted each EWA component (anterior–posterior and left–right directions) at the first EWA time-to-peak, the EWA directions focused on directions where muscle activity was shown (Fig. 8B), i.e., this component performs the function of producing a force to lean towards the target directions. In contrast, when we accepted a second EWA time-to-peak, the EWA directions were concentrated in opposite directions from those using the first time-to-peak (Fig. 8C). Therefore, this component performs the function of maintaining postural stability in the target directions. This phenomenon, a second time-to-peak, was likely to be a characteristic of our experimental task. In the previous study (Kutch et al., 2008), endpoint force was produced in the cylinder and the position of the index finger did not change. On the other hand, in the present study, subjects changed their body position towards the target direction during force production. Therefore, subjects were affected by gravity in our experimental task, which gives rise to the second time-to-peak component. A possible explanation is
that these two components performed different roles. This explanation is supported by the observation that the recorded surface EMG contained action potentials from motor units having different (Vieira et al., 2011) unit size and firing rate indicating that they play different roles in postural control. Considering these results, the EWA method potentially reveals different functions performed by each component contained in force fluctuations observed in a motor task. 4.4. Validity of motor output fluctuations in studies of motor control In this study, we examined postural control from COP fluctuations derived from isometric endpoint force fluctuations. As a result of our study, we found that different components contained in force fluctuations performed different functions in postural control. Chvatal et al. (2011) investigated muscle synergies using surface translations. Although this study revealed several muscle synergies in postural control, these synergies did not contain force fluctuations resulting from postural control dynamics. Force fluctuations were examined in many studies, especially in relation to motor unit properties, firing rate, firing variability, or synchronization (Griffin et al., 2009; Mottram et al., 2005). This indicates that
436
H. Imagawa et al. / Journal of Electromyography and Kinesiology 23 (2013) 430–437
90º
180º
0º
270º
Fig. 8. Difference between EWA directions from the first time-to-peak and the second time-to-peak, exemplified by SOL. (A) The histogram of EWA time-to-peaks for SOL. (B) Polar histogram of EWA directions from the first time-to-peak. (C) Polar histogram from the second time-to-peak.
force fluctuations are involved in precise motor control and are therefore an important factor in investigating motor control. 5. Conclusion Multi-directional postural control is achieved by synergistic coactivation of multiple muscles. Moreover, using the EWA method, we can estimate muscle contribution to postural control in each direction. In contrast, muscle synergies need to be investigated from both the g and EWA directions in motor tasks involving many muscles. The different components observed in this study are shown to have different functions in postural control. The present results suggest that muscle synergies in human bipedal standing can be examined in more detail using COP fluctuations. Acknowledgement The authors thank Keisuke Fujii (Kyoto University) for stimulating discussions and useful suggestions. References Chvatal SA, Torres-Oviedo G, Safavynia SA, Ting LH. Common muscle synergies for control of center of mass and force in nonstepping and stepping postural behaviors. J Neurophysiol 2011;106(2):999–1015.
Collins JJ, De Luca CJ. Open-loop and closed loop control of posture: a random walk analysis of center of pressure. Exp Brain Res 1993;95(2):308–18. De Luca CJ, Gilmore LD, Kuznetsov M, Roy SH. Filtering the surface EMG signal: movement artifact and baseline noise contamination. J Biomechanics 2010;43(8):1573–9. Griffin L, Painter PE, Wadhwa A, Spirduso WW. Motor unit firing variability and synchronization during short-term light-load training in older adults. Exp Brain Res 2009;197(4):337–45. Henry SM, Fung J, Horak FB. EMG responses to maintain stance during multidirectional surface translations. J Neurophysiol 1998;80(4):1939–50. Kluzik JA, Peterka RJ, Horak FB. Adaptation of postural orientation to changes in surface inclination. Exp Brain Res 2007;178(1):1–17. Kouzaki M, Masani K. Reduced postural sway during quiet standing by light touch is due to finger tactile feedback but not mechanical support. Exp Brain Res 2008;88(1):153–8. Kouzaki M, Masani K. Postural sway during quiet standing is related to physiological tremor and muscle volume in young and elderly adults. Gait Posture 2012;35(1):11–7. Kouzaki M, Shinohara M. Steadiness in plantar flexor muscles and its relation to postural sway in young and elderly adults. Muscle Nerve 2010;42(1):78–87. Kutch JJ, Kuo AD, Bloch AM, Rymer WZ. Endpoint force fluctuations reveal flexible rather than synergistic patterns of muscle cooperation. J Neurophysiol 2008;100(5):2455–71. Kutch JJ, Kuo AD, Rymer WZ. Extraction of individual muscle mechanical action from endpoint force. J Neurophysiol 2010;103(6):3535–46. Masani K, Popovic MR, Nakazawa K, Kouzaki M, Nozaki D. Importance of body sway velocity information in controlling ankle extensor activities during quiet stance. J Neurophysiol 2003;90(6):3774–82. Morasso PG, Schieppati M. Can muscle stiffness alone stabilize upright standing? J Neurophysiol 1999;82(3):1622–6. Mottram CJ, Christou EA, Meyer FG, Enoka RM. Frequency modulation of motor unit discharge has task-dependent effects on fluctuations in motor output. J Neurophysiol 2005;94(4):2878–87.
H. Imagawa et al. / Journal of Electromyography and Kinesiology 23 (2013) 430–437 Muceli S, Boye AT, d’Avella A, Farina D. Identifying representative synergy matrices for describing muscular activation patterns during multidirectional reaching in horizontal plane. J Neurophysiol 2010;103(3):1532–42. Peterka RJ, Loughlin PJ. Dynamic regulation of sensorimotor integration in human postural control. J Neurophysiol 2004;91(1):410–23. Ting LH, Macpherson JN. A limited set of muscle synergies for force control during a postural task. J Neurophysiol 2005;93(1):609–13. Torres-Oviedo G, Ting LH. Muscle synergies characterizing human postural responses. J Neurophysiol 2007;98(4):2144–56. Vieira TMM, Loram ID, Muceli S, Merletti R, Farina D. Postural activation of the human medial gastrocnemius muscle: are the muscle units spatially localised? J Physiol 2011;589(2):431–43. Winter DA, Prince S, Frank JS, Powell C, Zabjek KF. Unified theory regarding A/P and M/L balance in quiet stance. J Neurophysiol 1996;75(6):1–10.
Hiroaki Imagawa received his bachelor’s degree in 2012 from the Faculty of Integrated Human Studies, Kyoto University. He is currently in master’s course of the Department of Physical and Health Education, Graduate School of Education, The University of Tokyo, where he is a member of the Laboratory of Human Movement Control and Learning. His major research interests are optimal control and neural mechanism underlying motor control.
437
Shota Hagio received bachelor’s degree in Faculty of Integrated Human Studies, Kyoto University, Kyoto, Japan. He is currently in master’s course of the Graduate School of Human and Environmental Studies, Kyoto University, where he is a member of the Laboratory of Neurophysiology. His major research interests focus on human motor control on the basis of ‘‘muscle synergy’’.
Motoki Kouzaki received his Ph.D. degree in 1999 from the Department of Life Science, The University of Tokyo. From 1999 to 2007, he was an assistant Professor of the Department of Life Science, The University of Tokyo. In 2007, he was appointed as an Associate Professor of the Graduate School of Human and Environmental Studies, Kyoto University, where he is currently Director of the Laboratory of Neurophysiology. His major research interests focus on neural mechanisms underlying human motor control as related to synergistic muscles.