Journal of Applied Physiology Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J Appl Physiol 96: 1486-1495, 2004; doi:10.1152/japplphysiol.01070.2003
8750-7587/04 $5.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (148)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Farina, D.
Right arrow Articles by Enoka, R. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Farina, D.
Right arrow Articles by Enoka, R. M.

INVITED REVIEW

HIGHLIGHTED TOPICS
Neural Control of Movement

The extraction of neural strategies from the surface EMG

Dario Farina,1 Roberto Merletti,1 and Roger M. Enoka2

1Laboratorio di Ingegneria del Sistema Neuromuscolare, Dipartimento di Elettronica, Politecnico di Torino, Torino 10129, Italy; and 2Department of Integrative Physiology, University of Colorado, Boulder, Colorado 80309-0354


   ABSTRACT
TOP
 ABSTRACT
FACTORS THAT INFLUENCE THE...
 INFERRING NEURAL CONTROL...
 SUMMARY
 GRANTS
 REFERENCES
 
This brief review examines some of the methods used to infer central control strategies from surface electromyogram (EMG) recordings. Among the many uses of the surface EMG in studying the neural control of movement, the review critically evaluates only some of the applications. The focus is on the relations between global features of the surface EMG and the underlying physiological processes. Because direct measurements of motor unit activation are not available and many factors can influence the signal, these relations are frequently misinterpreted. These errors are compounded by the counterintuitive effects that some system parameters can have on the EMG signal. The phenomenon of crosstalk is used as an example of these problems. The review describes the limitations of techniques used to infer the level of muscle activation, the type of motor unit recruited, the upper limit of motor unit recruitment, the average discharge rate, and the degree of synchronization between motor units. Although the global surface EMG is a useful measure of muscle activation and assessment, there are limits to the information that can be extracted from this signal.

motor unit; spectral analysis; recruitment strategies; synchronization; amplitude cancellation; electromyogram

THE SURFACE ELECTROMYOGRAM (EMG) comprises the sum of the electrical contributions made by the active motor units (MUs) as detected by electrodes placed on the skin overlying the muscle. The information extracted from the surface EMG is often considered a global measure of MU activity, because of the inability of the traditional (2 electrode) recording configuration to detect activity at the level of single MUs. The global characteristics of the surface EMG, such as its amplitude and power spectrum, depend on the membrane properties of the muscle fibers as well as on the timing of the MU action potentials. Thus the surface EMG reflects both peripheral and central properties of the neuromuscular system.

Two approaches are available to study the relations between the surface EMG and the properties of the neuromuscular system: one forward and the other inverse. The forward approach, such as can be accomplished with modeling, allows us to predict the effect of various physiological processes on features of the surface EMG. The inverse approach uses the EMG to identify the underlying physiology. The inverse approach, however, requires simplifications to reduce the number of parameters and multiple solutions that influence the association. The relation between average conduction velocity of the muscle fibers and spectral frequencies of the surface EMG during isometric contractions sustained at a constant force (3, 8, 56) is an example of a forward association; the physiology determines the EMG characteristics. The inverse problem involves estimating changes in the average conduction velocity from the characteristic spectral frequencies and can be solved by approximating this relation with a linear equation. A limitation of inverse models is that approximations are valid for specific conditions and cannot be applied to more general situations, which can result in misleading conclusions. The situation is confounded when some of the factors that influence features of the signal have counterintuitive or unexpected effects.

The application of mathematical models (10, 27, 31) has proven useful in characterizing the sensitivity of the surface EMG to the parameters of the systems involved in the generation and detection of the signal. Structure-based models, for example, can now describe the generation of the surface EMG in complex volume conductors (31) as a consequence of various control strategies (35, 46, 84). Although the use of these tools has clarified the limitations associated with the measurement and interpretation of the surface EMG, these limitations are often not appreciated by the experimentalist.

The aim of this brief review was to characterize the strengths and weaknesses of some of the methods used to infer central control strategies from bipolar recordings of the surface EMG. The review focuses on the global surface EMG and does not relate, except for a brief note, to more advanced methods for extracting information on single MUs from noninvasive recordings, which have been addressed in recent reviews (19, 54, 55, 85, 86). The review is not exhaustive, but it examines topical issues related to the interpretation of the surface EMG by focusing on the limitations of some methods as they are used to extract information from the surface EMG.


   FACTORS THAT INFLUENCE THE SURFACE EMG
TOP
 ABSTRACT
 FACTORS THAT INFLUENCE THE...
INFERRING NEURAL CONTROL...
 SUMMARY
 GRANTS
 REFERENCES
 
The features of the EMG signal depend on many "nonphysiological" factors (Table 1). The influence of these factors has been measured, simulated, and discussed (22, 29, 30, 41, 49, 71). Some of these effects are not intuitive and vary with experimental conditions. Nonetheless, useful information can be extracted from the surface EMG, especially when the experimental protocol permits some of these factors to be minimized. The influence of some of these factors can be reduced significantly by appropriate placement of the electrodes. Because different electrode locations over the same muscle can provide signals with significantly different features, some locations are preferred over others (41, 71). For this reason, placement of the electrodes should always be reported in EMG studies.


View this table:
[in this window]
[in a new window]
  		Table 1. Factors that influence the surface EMG

 

Crosstalk

Of the nonphysiological factors listed in Table 1, crosstalk provides an example of how intuitive considerations can be incorrect when dealing with phenomena that are influenced by the properties of the volume conductor. Crosstalk refers to a signal recorded over one muscle that is actually generated by a nearby muscle and conducted through the intervening volume to the recording electrodes (15). Many investigators assume that a crosstalk signal has "a lower frequency spectrum because it originates further away and will be subject to additional low-pass filtering due to spatial filtering" (14). According to this rationale, high-pass filtering should reduce crosstalk; however, this is not a general finding. Recordings of muscle fiber action potentials are influenced by two events: propagation along the fiber and extinction at the end of the fiber. With an increase in the distance between the recording electrodes and the active muscle fibers, the nonpropagating components due to extinction of the action potentials (40) dominate those due to propagation (29). Because the high-frequency components of the nonpropagating signals are greater than those for the propagating signals, crosstalk signals can have a broader bandwidth than signals detected directly over an active muscle (17, 29). Because of this effect, high-pass filtering can have no effect on crosstalk (Fig. 1).



View larger version (39K):
[in this window]
[in a new window]
  		Fig. 1. The influence of the distance between the recording electrodes and active muscle fibers on the surface electromyogram. A: section of a volume conductor comprised of bone, muscle, fat, and skin layers (31) with pairs of electrodes placed on the skin at 2 locations. The muscle comprised 150 motor units (MUs) with the innervation number ranging from 50 to 800. B: the surface electromyogram was recorded with a pair of electrodes 10 mm apart and placed between the center of the innervation zone and the tendon junction. The muscle fibers were 110 mm in length, and there was a scatter of 5 mm for both the endplates and the tendon endings. C: simulated surface EMG signals as detected over (location 1) and far (location 2) from the active muscle. Signals detected concurrently at the two recording sites are reported as interference data (without any filtering) and after differentiation with respect to time, as suggested in Winter et al. (83) (sampling frequency 1,024 Hz). The amplitude of the signals is indicated in arbitrary units (au), with a scale of 1 au for the signals detected at location 1 and 0.05 au for those detected at location 2. The correlation coefficient (CC) between the recordings for the 2 locations was low, both before and after time differentiation. Moreover, high-pass filtering (differentiation) increased the amount of crosstalk, as indicated by the ratio of the average rectified value (ARV) of the signal detected over the muscle (ARVmus) relative to that far away (ARVcro). D: an action potential for a single MU as detected at the 2 locations. The shape of the potential recorded at location 1 is dominated by the propagating component, whereas the shape recorded at location 2 is largely due to the nonpropagating (extinction) component.

 

Alternatively, the presence of crosstalk has been estimated by calculating the cross-correlation coefficient between the signals detected from two muscles (1, 59, 61, 83). Journals that specialize in EMG measurements, for example, often recommend the use of cross-correlation analysis, as suggested by Winter et al. (83), as the standard for detecting crosstalk. However, the cross-correlation analysis is not a valid measure of the presence or absence of crosstalk (29, 51), as indicated in Fig. 1C.

A recent experimental study (29) and simulation analyses (17, 28, 51) suggest several conclusions about crosstalk: 1) signals detected far from the source are mainly due to the extinction of the action potentials at the ends of the fibers; 2) because of differences in the sources of the propagating and nonpropagating signals, the cross-correlation coefficient is generally not indicative of the amount of crosstalk; 3) the frequency content of an EMG signal does not identify the presence of crosstalk; 4) temporal high-pass filtering of surface EMG can fail to reduce crosstalk signals; and 5) the use of spatial filtering as a method to decrease crosstalk remains to be validated because the theoretical analyses of spatial filters have been limited to propagating signals (18, 69, 70).


   INFERRING NEURAL CONTROL STRATEGIES FROM THE SURFACE EMG
TOP
 ABSTRACT
 FACTORS THAT INFLUENCE THE...
 INFERRING NEURAL CONTROL...
SUMMARY
 GRANTS
 REFERENCES
 
Various methods have been used to infer details about the signals discharged from the spinal cord to activate muscle. These techniques, however, have various limitations that are often not appreciated.

Amplitude of the Surface EMG

The amplitude of the surface EMG can be estimated by a scheme of demodulation, smoothing, and relinearization. In this process, demodulation rectifies the EMG and then raises the result to a power (e.g., 1 for the average rectified value or 2 for the root mean square value), smoothing filters the signal, and relinearization inverts the power law applied during the demodulation stage and returns the signal to units of EMG amplitude.

The amplitude of the surface EMG is related to the net MU activity; that is, the recruitment and the discharge rates of the active MUs. Because of this relation, some investigators use EMG amplitude as an index of the level of activation provided by the spinal cord. As listed in Table 1, however, EMG amplitude is influenced by such factors as electrode location, thickness of the subcutaneous tissues, distribution of MU conduction velocities, and the detection system used to obtain the recording. Although some of these effects can be reduced by appropriate placement of the electrodes (30, 41), there remains a mismatch between the output of the spinal cord and the EMG amplitude.

Amplitude cancellation. The surface EMG underestimates the activation signal sent from the spinal cord to muscle as a result of the cancellation of positive and negative phases of MU action potentials (13) (Fig. 2). The amount of signal cancellation can be quantified by comparing the signal amplitude obtained by summing MU action potentials before and after the rectification of each potential (13, 44). Cancellation is present when the unrectified action potentials are summed but not when the rectified potentials are summed. These results indicate that, although the amplitude of the surface EMG increases monotonically with the neural drive to a muscle, a general relation cannot be determined.



View larger version (27K):
[in this window]
[in a new window]
  		Fig. 2. Simulation of amplitude cancellation. The simulations involved 2 of the MUs from the study described in Fig. 1 (nu = normalized units). An interference signal was simulated with each MU independently. Contraction intensity (Excitation level) was varied by repeating the number of times (from 50 to 1,000 in steps of 50) that the MU action potential occurred within a 5-s interval. Occurrence times were distributed randomly. A: the 2 MU potentials and representative signals generated at different levels of excitation with MU 2. Even at excitation level 2 (100 MU action potentials), examples of phase cancellation are clearly visible. B: the average rectified value of the signals, normalized relative to the first excitation level, is plotted against the excitation level. The signal amplitude does not increase linearly with excitation level because of the cancellation of positive and negative phases of the waveforms. Moreover, amplitude cancellation depended on the waveform selected and the assigned conduction velocity (3 or 5 m/s). At the maximal level of excitation, for example, amplitude could differ by ~30% in the different conditions, although the amount of excitation was the same in all cases. The magnitude of this effect is greater in an actual muscle (13).

 

Amplitude cancellation depends on many of the factors listed in Table 1 (44) and indicates that comparable levels of average rectified EMG, either in different subjects or on different occasions, do not necessarily indicate similar levels of output from the spinal cord (Fig. 2). Furthermore, changes in the average rectified or root mean square EMG values after an intervention may not rigorously reflect altered levels of neural drive to the muscle. Similarly, the so-called measure of neuromuscular efficiency (16, 58), which corresponds to the ratio between an exerted force and the amplitude of a surface EMG, and other such indexes suffer from the dependence of amplitude cancellation on factors that can have nothing to do with the amplitude of the activation signal (44). The extensive use of such indexes in basic and clinical studies underscores the lack of appreciation for the significance of signal cancellation.

Estimation of muscle force. Both the force exerted by a muscle and the amplitude of the surface EMG depend on the number of recruited MUs and the discharge rate of each active MU. Accordingly, it is reasonable to expect that muscle force can be estimated from the surface EMG. Because of amplitude cancellation, however, the amplitude change of the surface EMG actually underestimates the associated change in MU activity underlying the modulation of muscle force.

Many factors influence the relation between EMG amplitude and force (Table 1). When muscles and subjects are compared, these factors include the thickness of subcutaneous tissue, the recruitment strategy, the peak discharge rates of the different MUs (35), and so on. Moreover, the same control strategy may generate signals with different amplitude trends depending on the locations of the active MUs within the muscle (24). Because of the many factors that can influence this relation, there is no reason to expect that a specific EMG amplitude-force relation should have general validity. This relation should be identified on a subject-by-subject and muscle-by-muscle basis.

Spectral Analysis of the Surface EMG

Spectral analysis of surface EMG signals has been used to study muscle fatigue (56) and to infer changes in MU recruitment (6, 7, 73). Characteristic spectral frequencies can be computed by the classic periodogram (57) and autoregressive-based approaches (66) or by advanced methods such as Cohen's class time-frequency distributions (12) and wavelet analysis (43). The latter techniques have been used with dynamic (5, 12) and isometric contractions (42) and may be more appropriate than the classic approaches when the signals are nonstationary. However, the critical limitations in the spectral analysis of the surface EMG are intrinsic to the properties of the surface EMG signals and not the type of signal analysis.

Relation between spectral frequencies and conduction velocity. The relation between the average conduction velocity of the muscle fibers (4) and power spectral frequencies is used to study muscle fatigue, to identify the type of MUs recruited, and to describe the pattern of MU activity. The results, however, are only relevant when specific conditions are met. The relative changes in the spectral frequencies and conduction velocity during sustained isometric contractions at moderate to high intensities, for example, can be approximated by a linear relation (3, 56). Indeed, under ideal conditions, it can be shown that conduction velocity scales the power spectrum (50). Nonetheless, it is inappropriate to use this relation when the number of active MUs changes significantly during the contraction (25).

Spectral analysis of the surface EMG has also been used to estimate the activation of type I and II muscle fibers and the recruitment of MUs during a contraction (9, 37, 38, 74, 79). The rationale for these applications is that muscle fiber diameter, and hence conduction velocities of MUs, vary systematically with MU type and that the discharge rates have only a minor effect on the power spectrum (48). Four physiological details, however, confound this rationale. First, the two main fiber types do not have clearly distinct conduction velocities, but rather conduction velocity has a continuous distribution with a single peak (76) and in some muscles the diameter of type I fibers is larger than that of type II fibers. Thus distinguishing the proportion of the two fiber types on the basis of conduction velocity can be arbitrary. Second, the number of muscle fibers innervated by a MU has a skewed distribution (21), which must be considered when using the average conduction velocity to estimate the proportion of MU types active in a contraction. For example, the first dorsal interosseus muscle comprises an equal number of type I and II fibers, but ~84% of the MUs are slow contracting and fatigue resistant, whereas only 16% are fast contracting. Conversely, 75% of the motoneurons innervating the triceps brachii muscle innervate type I fibers, which represent only 33% of the fibers in the muscle (21). Third, the conduction velocity of single MUs changes with discharge rate (64), which means that the contribution of a MU to the average conduction velocity varies with discharge rate. Fourth, the shortening of muscle fibers during a dynamic contraction influences conduction velocity (47).

In addition to the physiology, the effect of the volume conductor on the action potentials further complicates the use of spectral analysis to estimate MU features (Fig. 3). This effect is not dependent on the MU type but only on the location and anatomy of the active muscle fibers within the muscle and the properties of the volume conductor. For example, differences in fiber length influence action potential shape (location 3 in Fig. 3B) and the associated power spectra (Fig. 3, C and D). Thus the use of spectral characteristics to infer the proportion of a MU type that is activated during a contraction is often flawed.



View larger version (35K):
[in this window]
[in a new window]
  		Fig. 3. Simulations of single MU action potentials as detected by a pair of electrodes located between the innervation zone and the tendon. The MU, which was located in 3 positions within the muscle, had a circular territory (2-mm radius) and innervated 250 muscle fibers with a density of 20 fibers/mm2. A: the volume conductor was cylindrical and layered (31). Locations 1 and 2 were at the same distance from the detection point, whereas location 3 was 4 mm deeper. Four anatomic conditions have been simulated corresponding to fiber lengths (L) of 130 and 50 mm and fat layers of 1 and 4 mm in thickness. Conduction velocity was 4 m/s. B: the detected MU potentials were normalized relative to the maximal value. The duration of the potentials and the relative weight of the nonpropagating (end-of-fiber) components were influenced by the anatomic parameters. C and D: the power spectral densities of the MU potentials were normalized relative to the maximal value. For each power spectrum, the mean frequency is shown to indicate the frequency around which the spectrum of the MU potential was concentrated. Because of the minor effect of discharge rate on the power spectrum, the power spectrum of the interference EMG signal can be approximated by the sum of the power spectra of the MU potentials (48). The bandwidth of the power spectrum was broader for shorter fibers (D vs. C). Because the distance from the detection system is not the only parameter to influence the power spectrum, the power spectra for locations 1 and 2 were not identical. Although the depth of the source may decrease mean frequency, the mean frequency may increase for large distances, especially when the fibers are short, because of the prevalence of nonpropagating components (see also Fig. 1). Despite a constant conduction velocity, the frequency bandwidths were significantly different in the 12 cases (2 fiber lengths x 3 locations x 2 fat thicknesses). Depending on the anatomy and the location of the MU, the power spectrum of the detected action potential will be concentrated between 60 and 150 Hz.

 

Because variation in muscle fiber conduction velocity has been proposed as an index of MU recruitment according to the size principle (2), it has been suggested that shifts in the mean or median frequencies should reflect the recruitment of progressively larger and faster MUs (6, 7). At the upper limit of MU recruitment, therefore, these frequencies should reach a plateau (or decrease) because subsequent increases in force are due to rate coding only (48). On the basis of these presumed associations, there are reports of differences in MU recruitment strategies during motor skill acquisition and between the dominant and nondominant arm (6, 7). Similar assumptions have been used when processing nonstationary surface EMG signals during force modulation and fatigue, such as the joint analysis of EMG spectrum and amplitude (JASA method) proposed by Luttmann et al. (52).

Because of volume conductor effects, however, the recruitment of a MU may contribute to either the high- or low-frequency bands depending on factors other than conduction velocity, such as the depth of the MU (Fig. 3). Thus the recruitment of a MU with a high conduction velocity does not necessarily increase the characteristic frequencies in the power spectrum. Accordingly, a modeling study has shown that different recruitment strategies may lead to similar mean frequency patterns (24). Furthermore, the diversity of experimental findings on changes in mean or median frequency with increasing force underscores the equivocal basis of this approach; there are reports of an increase in the spectral variables with increasing force level (36, 60), no increase in these variables with force (67, 78), and a decrease with increasing force (68, 81).

Discharge rate. From simple mathematical derivations (48, 77, 82), the power spectral density of a train of MU action potentials has two main components: the spectrum of the surface detected action potential and the repetition rate of action potentials. The second component represents peaks at frequencies that are multiples of the average discharge rate. Assuming that the trains of MU action potentials are independent, their power spectra will add to form the spectrum for the interference EMG signal. As a consequence, peaks at low frequencies should be observable in the surface EMG spectrum and indicate the average discharge rate of the active MUs. Although some authors have used low-frequency peaks to estimate mean discharge rate (39), at least five factors confound this association: 1) the scatter of the discharge rates decreases peak amplitudes; 2) the coefficient of variation for discharge times decreases the peaks at low frequencies; 3) the surface-detected MU potentials have low energy at low frequencies that correspond to the MU discharge rates; 4) the inherent variance of estimation in the power spectral density due to the stochastic nature of the surface EMG signals introduces spurious peaks (26); and 5) the nonstationary behavior in the discharge statistics can markedly influence the low-frequency portion of the EMG power spectral density (20).

Recently, rectification of the EMG signal has been suggested as a method to enhance the low-frequency peaks and hence the detection of average discharge rate (62). However, rectification is a nonlinear operation and, from the theoretical point of view, its use for enhancing the low-frequency peaks attributable to the discharge statistics is not justified. Because rectification is performed on the interference signal after the occurrence of amplitude cancellation, the presence of repeating waveforms in the EMG is not enhanced after rectification. As with amplitude cancellation, this effect is more pronounced with increasing numbers of MUs. Figure 4 demonstrates that signal rectification may increase the low-frequency peaks in some conditions, but not in general, and may easily lead to the detection of low-frequency peaks that are not related to the average discharge rate.



View larger version (46K):
[in this window]
[in a new window]
  		Fig. 4. Simulation of surface EMG signals at different contraction intensities with the model used in Fig. 1. The distribution of discharge rates of the active MUs was determined as described in Fuglevand et al. (35). The upper limit of MU recruitment was 80% of the excitation level, and discharge rate increased at 0.3 pulses per second (pps)/excitation level. Discharge rate at recruitment was 8 pps for all MUs. Four excitation levels were simulated: 15% (top row) and 35, 55, and 75% (bottom row). The number of MUs activated at the four levels of excitation was 92, 121, 137, and 147. Mean discharge rate is indicated to the left of each row. The coefficient of variation for discharge rate (variability) was assigned values of 0% (left), 10% (middle), and 20% (right). The estimated power spectra for the 20-s simulations are shown (only in the frequency range 0-80 Hz) for both the interference and rectified signals. The arrows indicate the mean discharge rates of the active MUs. Rectification enhanced the low-frequency peaks in the spectra for low coefficients of variations and few active MUs. As the coefficient of variation approached values observed experimentally (65), however, it was difficult to detect the peaks in the rectified EMG spectra because of discharge rate variability. In many cases, peaks were evident in the interference signal but not the rectified signal spectra, because the model of the surface electromyogram as the summation of action potential trains does not hold for the rectified signal.

 

Cross-Correlation Analysis of the Surface EMG

When many MUs are synchronized between muscles, it has been proposed that cross-correlation analysis between the surface EMG recordings for the two muscles may be used to assess the degree of synchronization (45). This technique assumes a similarity in the shapes of the action potentials discharged by the MUs in each muscle. However, the action potentials detected on the skin over the two muscles are presumably generated by MUs at different distances from the detection surfaces and include differences in fiber length, locations of the innervation and tendon zones, local properties of the volume conductor (e.g., fat thickness), and anatomic territories. Because these factors can all influence the shape of surface-detected potentials, it seems unlikely that the cross-correlation between the surface EMG signals can identify coincidental discharges by the MUs in the two muscles.

If the "synchronized" MUs are both located close enough to the detection points so that the signals are dominated by the propagating components, the shapes of the action potentials of MUs in the two muscles will be rather similar because of the filtering effect of the volume conductor, despite differences in the properties of the two MUs. As an example, the three MUs simulated in Fig. 3B have a similar shape despite different locations, fiber lengths, and thicknesses of the fat layer. The cross-correlation coefficient between the MU potential in the top left of Fig. 3B (location 1, Fat 1 mm, L 130 mm) and the MU potential in the bottom right (location 3, Fat 4 mm, L 50 mm) was 0.88, which indicates that MUs with very different properties may easily generate surface potentials with similar shapes. Similarly, the potentials in Fig. 3 generated by the MU with short fibers at the three locations have minimal correlation coefficients of 0.78 or 0.89 for the thin or thick fat layers, respectively. This is why a large correlation between surface EMG signals detected from different muscles was found in a simulation study that examined the synchronization of MUs between the muscles (51).

The cross-correlation function between signals detected from two distant muscles is not influenced by crosstalk; when the nonpropagating components dominate, the crosstalk components are not correlated to the signals detected over the two muscles (Fig. 1). For example, high peaks of the cross-correlation function between surface EMG signals from different muscles were observed when synchronization was simulated, whereas the presence of crosstalk did not have a major influence on the cross-correlation function (51). As a consequence of this finding, methods based on blind source separation for reducing the effect of crosstalk on the measurement of synchronization from a cross-correlation analysis are not necessary (45). Moreover, the relatively large reduction in the cross-correlation peak when these methods were applied to signals detected from distant muscles (45) was probably an effect introduced by the algorithm rather than due to a reduction in crosstalk.

Nonlinear Analysis of the Surface EMG

It has been proposed that central strategies can be examined with variables extracted from the surface EMG by nonlinear recurrence methods (63, 80). Webber et al. (80) reported that subtle changes in surface EMG, because of force modulation, could have a greater influence on the percentage of determinism in the signal compared with an effect on the spectral frequencies. It has also been speculated (33, 34) that the percentage of determinism during fatiguing contractions should be able to detect within-muscle synchronization of MUs because the measurement reveals embedded determinisms in an otherwise stochastic signal. Contrary to these expectations, the percentage of determinism is highly correlated with characteristic spectral frequencies and does not provide additional information when synchronization is limited to values observed in muscles of healthy subjects (23). In the presence of high levels of synchronization, however, the recurrence analysis may provide more information about the degree of synchronization than the spectral analysis (D. Farina and L. Fattorini, unpublished observations). Although these results suggest potential applications for this method, the recurrence plot analysis must be validated with modeling studies before it can be used to infer central control strategies (23).

Detecting Single MUs from the Surface EMG

Techniques for the analysis of single MU properties from surface EMG signals have been developed recently, at least for specific conditions (Fig. 5). These approaches are based on multichannel surface EMG recordings (54, 86), which allow the location of the innervation zones, fiber length, muscle fiber conduction velocity, and the timing of action potentials for single MUs to be determined (32, 53, 72, 75). With such measurements, direct observation of the MU control strategies can be assessed, which avoids the limitations associated with the inverse modeling approach that is typical of the global surface EMG analysis.



View larger version (80K):
[in this window]
[in a new window]
  		Fig. 5. Multichannel surface electromyogram to detect the activity of single MUs. A: surface electromyogram signals detected by a matrix of 61 electrodes (5 columns x 13 rows with a 5-mm interelectrode distance in both the longitudinal and transverse directions) from the biceps brachii muscle during a voluntary contraction at 5% of the maximal voluntary contraction force. The signals were acquired with a single differential filter in the longitudinal direction. B-D: a double differentiation was applied offline to the single differential signals to increase spatial resolution, and single MU potentials were automatically extracted and classified as belonging to different MUs by a decomposition method for surface EMG signals (35a). Results for 3 MUs are shown. The potentials classified as belonging to a specific MU are superimposed to show the shape similarity. The column of the matrix with the potentials of maximum amplitude corresponds to the location of the MU over the skin plane. Increasing distance from the source in the transverse direction decreases the amplitude of the potentials. Potentials detected along the fiber direction (i.e., by the electrodes of a column) are similar in shape and delayed between each other when the electrodes are between the innervation zone and the tendon. The detection with the 2-dimensional array allowed the identification of the location of the MUs over the skin plane and the estimation of their anatomic properties, such as the location of the innervation zones. [Reprinted from Gazzoni et al. (35a), with permission from Elsevier.]

 


   SUMMARY
TOP
 ABSTRACT
 FACTORS THAT INFLUENCE THE...
 INFERRING NEURAL CONTROL...
 SUMMARY
GRANTS
 REFERENCES
 
This brief review examined some of the methods used to extract information on the neural control of movement from surface EMG signals. The focus was on the global surface EMG analysis. The topics addressed in the review comprised methods that are currently used extensively in this field, with an emphasis on the critical assumptions underlying some of these approaches. The limitations of these techniques are often not appreciated, which sometimes leads to erroneous interpretation of the results and conflicting reports in the literature. The review was intended to highlight the limitations of some surface-EMG measurements that are used to identify the physiological mechanisms responsible for the neural control of movement.


   GRANTS
TOP
 ABSTRACT
 FACTORS THAT INFLUENCE THE...
 INFERRING NEURAL CONTROL...
 SUMMARY
 GRANTS
REFERENCES
 
This work was partially supported by the European Shared Cost Project "Neuromuscular Assessment in the Elderly Worker" (NEW) (contract no. QLRT-2000-00139) (to D. Farina and R. Merletti) and by National Institute of Neurological Disorders and Stroke Award NS-42734 (to R. M. Enoka).


   FOOTNOTES
 

Address for reprint requests and other correspondence: D. Farina, Dipartimento di Elettronica, Politecnico di Torino, Corso Duca degli Abruzzi 24, Torino 10129, Italy (E-mail: dario.farina{at}polito.it).


   REFERENCES
TOP
 ABSTRACT
 FACTORS THAT INFLUENCE THE...
 INFERRING NEURAL CONTROL...
 SUMMARY
 GRANTS
 REFERENCES
 

  1. Aagaard P, Simonsen EB, Andersen JL, Magnusson SP, Bojsen-Møller F, and Dyhre-Poulsen P. Antagonist muscle coactivation during isokinetic knee extension. Scand J Med Sci Sports 10: 58-67, 2000.
  2. Andreassen S and Arendt-Nielsen L. Muscle fibre conduction velocity in motor units of the human anterior tibial muscle: a new size principle parameter. J Physiol 391: 561-571, 1987.
  3. Arendt-Nielsen L, Mills KR, and Forster A. Changes in muscle fiber conduction velocity, mean power frequency, and mean EMG voltage during prolonged submaximal contractions. Muscle Nerve 12: 493-497, 1989.
  4. Arendt-Nielsen L and Zwarts M. Measurement of muscle fiber conduction velocity in humans: techniques and applications. Electroencephalogr Clin Neurophysiol 6: 173-190, 1989.
  5. Balestra G, Frassinelli S, Knaflitz M, and Molinari F. Time-frequency analysis of surface myoelectric signals during athletic movement. IEEE Eng Med Biol Mag 20: 106-115, 2001.
  6. Bernardi M, Felici F, Marchetti M, Montellanico F, Piacentini MF, and Solomonow M. Force generation performance and motor unit recruitment strategy in muscles of contralateral limbs. J Electromyogr Kinesiol 9: 121-130, 1999.
  7. Bernardi M, Solomonow M, Nguyen G, Smith A, and Baratta R. Motor unit recruitment strategies changes with skill acquisition. Eur J Appl Physiol 74: 52-59, 1996.
  8. Bigland-Ritchie B, Donovan EF, and Roussos CS. Conduction velocity and EMG power spectrum changes in fatigue of sustained maximal efforts. J Appl Physiol 51: 1300-1305, 1981.
  9. Bilodeau M, Schindler-Ivens S, Williams DM, Chandran R, and Sharma SS. EMG frequency content changes with increasing force and during fatigue in the quadriceps femoris muscle of men and women. J Electromyogr Kinesiol 13: 83-92, 2003.
  10. Blok JH, Stegeman DF, and van Oosterom A. Three-layer volume conductor model and software package for applications in surface electromyography. Ann Biomed Eng 30: 566-577, 2002.
  11. Bonato P, Cheng MS, Gonzalez-Cueto J, Leardini A, O'Connor J, and Roy SH. EMG-based measures of fatigue during a repetitive squat exercise. IEEE Eng Med Biol Mag 20: 133-143, 2001.
  12. Bonato P, Roy SH, Knaflitz M, and DeLuca CJ. Time-frequency parameters of the surface myoelectric signal for assessing muscle fatigue during cyclic dynamic contractions. IEEE Trans Biomed Eng 48: 745-753, 2001.
  13. Day SJ and Hulliger M. Experimental simulation of cat electromyogram: evidence for algebraic summation of motor-unit action-potential trains. J Neurophysiol 86: 2144-2158, 2001.
  14. De Luca CJ. The use of surface electromyography in biomechanics. J Appl Biomech 13: 135-163, 1997.
  15. De Luca CJ and Merletti R. Surface myoelectric signal cross-talk among muscles of the leg. Electroencephalogr Clin Neurophysiol 69: 568-575, 1988.
  16. Deschenes MR, Giles JA, McCoy RW, Volek JS, Gomez AL, and Kraemer WJ. Neural factors account for strength decrements observed after short-term muscle unloading. Am J Physiol Regul Integr Comp Physiol 282: R578-R583, 2002.
  17. Dimitrova NA, Dimitrov V, and Nikitin OA. Neither high-pass filtering nor mathematical differentiation of the EMG signals can considerably reduce cross-talk. J Electromyogr Kinesiol 12: 235-246, 2002.
  18. Disselhorst C, Silny J, and Rau G. Improvement of spatial resolution in surface-EMG: a theoretical and experimental comparison of different spatial filters. IEEE Trans Biomed Eng 44: 567-574, 1997.
  19. Disselhorst-Klug C, Bahm J, Ramaekers V, Trachterna A, and Rau G. Non-invasive approach of motor unit recording during muscle contractions in humans. Eur J Appl Physiol 83: 144-150, 2000.
  20. Englehart KB and Parker PA. Single motor unit myoelectric signal analysis with nonstationary data. IEEE Trans Biomed Eng 41: 168-180, 1994.
  21. Enoka RM and Fuglevand AJ. Motor unit physiology: some unresolved issues. Muscle Nerve 24: 4-17, 2001.
  22. Farina D, Cescon C, and Merletti R. Influence of anatomical, physical and detection-system parameters on surface EMG. Biol Cybern 86: 445-456, 2002.
  23. Farina D, Fattorini L, Felici F, and Filligoi G. Nonlinear surface EMG analysis to detect changes of motor unit conduction velocity and synchronization. J Appl Physiol 93: 1753-1763, 2002.
  24. Farina D, Fosci M, and Merletti R. Motor unit recruitment strategies investigated by surface EMG variables. J Appl Physiol 92: 235-247, 2002.
  25. Farina D, Gazzoni M, and Merletti R. Assessment of low back muscle fatigue by surface EMG signal analysis: methodological aspects. J Electromyogr Kinesiol 13: 319-332, 2003.
  26. Farina D and Merletti R. Comparison of algorithms for estimation of EMG variables during voluntary isometric contractions. J Electromyogr Kinesiol 10: 337-350, 2000.
  27. Farina D and Merletti RR. A novel approach for precise simulation of the EMG signal detected by surface electrodes. IEEE Trans Biomed Eng 48: 637-646, 2001.
  28. Farina D, Merletti R, Indino B, and Graven-Nielsen T. Surface EMG crosstalk evaluated from experimental recordings and simulated signals Reflections on crosstalk interpretation, quantification and reduction. Meth Inform Med. In press.
  29. Farina D, Merletti R, Indino B, Nazzaro M, and Pozzo M. Cross-talk between knee extensor muscles. Experimental and model results. Muscle Nerve 26: 681-695, 2002.
  30. Farina D, Merletti R, Nazzaro M, and Caruso I. Effect of joint angle on surface EMG variables for the muscles of the leg and thigh. IEEE Eng Med Biol Mag 20: 62-71, 2001.
  31. Farina D, Mesin L, Martina S, and Merletti R. A surface EMG generation model with multi-layer cylindrical description of the volume conductor. IEEE Trans Biomed Eng. In press.
  32. Farina D, Muhammad W, Fortunato E, Meste O, Merletti R, and Rix H. Estimation of single motor unit conduction velocity from surface electromyogram signals detected with linear electrode arrays. Med Biol Eng Comput 39: 225-236, 2001.
  33. Felici F, Rosponi A, Sbriccoli P, Filligoi GC, Fattorini L, and Marchetti M. Linear and non-linear analysis of surface electromyograms in weightlifters. Eur J Appl Physiol 84: 337-342, 2001.
  34. Filligoi G and Felici F. Detection of hidden rhythms in surface EMG signals with a non-linear time-series tool. Med Eng Phys 21: 439-448, 1999.
  35. Fuglevand AJ, Winter DA, and Patla AE. Models of recruitment and rate coding organization in motor unit pools. J Neurophysiol 70: 2470-2488, 1993.
  36. Gazzoni M, Farina D, and Merletti R. A new method for the extraction and classification of single motor unit action potentials from surface EMG signals. J Neurosci Meth. In press.
  37. Gerdle B, Eriksson NE, and Brundin L. The behaviour of the mean power frequency of the surface electromyogram in biceps brachii with increasing force and during fatigue. With special regard to the electrode distance. Electromyogr Clin Neurophysiol 30: 483-489, 1990.
  38. Gerdle B, Henriksson-Larsen K, Lorentzon R, and Wretling ML. Dependence of the mean power frequency of the electromyogram on muscle force and fibre type. Acta Physiol Scand 142: 457-465, 1991.
  39. Gerdle B, Wretling ML, and Henriksson-Larsen K. Do the fibre-type proportion and the angular velocity influence the mean power frequency of the electromyogram? Acta Physiol Scand 134: 341-346, 1988.
  40. Gibbs J, Harrison LM, and Stephens JA. Cross-correlation analysis of motor unit activity recorded from two separate thumb muscles during development in man. J Physiol 499: 255-266, 1997.
  41. Gydikov A, Gerilovski L, Radicheva N, and Troyanova N. Influence of the muscle fiber end geometry on the extracellular potentials. Biol Cybern 54: 1-8, 1986.
  42. Jensen C, Vasseljen O, and Westgaard RH. The influence of electrode position on bipolar surface electromyogram recordings of the upper trapezius muscle. Eur J Appl Physiol 67: 266-273, 1993.
  43. Karlsson S and Gerdle B. Mean frequency and signal amplitude of the surface EMG of the quadriceps muscles increase with increasing torque-a study using the continuous wavelet transform. J Electromyogr Kinesiol 11: 131-140, 2001.
  44. Karlsson S, Yu J, and Akay M. Time-frequency analysis of myoelectric signals during dynamic contractions: a comparative study. IEEE Trans Biomed Eng 47: 228-238, 2000.
  45. Keenan KG, Farina D, Maluf KS, Merletti R, and Enoka RM. Age-associated changes in motor unit properties reduce signal cancellation in the simulated electromyogram (Abstract). Soc Neurosci Abstr #914.11, 2003.
  46. Kilner JM, Baker SN, and Lemon RN. A novel algorithm to remove electrical cross-talk between surface EMG recordings and its application to the measurement of short-term synchronisation in humans. J Physiol 538: 919-930, 2002.
  47. Kleine BU, Stegeman DF, Mund D, and Anders C. Influence of motorneuron firing synchronization on SEMG characteristics in dependence of electrode position. J Appl Physiol 91: 1588-1599, 2001.
  48. Kossev A, Gantchev N, Gydikov A, Gerasimenko Y, and Christova P. The effect of muscle fiber length change on motor units potentials propagation velocity. Electromyogr Clin Neurophysiol 32: 287-294, 1992.
  49. Lago P and Jones NB. Effect of motor unit firing time statistics on EMG spectra. Med Biol Eng Comput 15: 648-655, 1977.
  50. Li W and Sakamoto K. The influence of location of electrode on muscle fiber conduction velocity and EMG power spectrum during voluntary isometric contractions measured with surface array electrodes. Appl Human Sci 15: 25-32, 1996.
  51. Lindstrom L and Magnusson R. Interpretation of myoelectric power spectra: a model and its applications. Proc IEEE 65: 653-662, 1977.
  52. Lowery MM, Stoykov NS, and Kuiken TA. A simulation study to examine the use of cross-correlation as an estimate of surface EMG cross talk. J Appl Physiol 94: 1324-1334, 2003.
  53. Luttmann A, Jäger M, Sökeland J, and Laurig W. Joint analysis of aspectrum and amplitude (JASA) of electromyograms applied for the indication of muscular fatigue among surgeons in urology. In: Advances in Occupational Ergonomics and Safety, edited by Mital A, Krueger H, Kumar S, Menozzi M, and Fernandez JE. Cincinnati, OH: Int. Soc. for Occupational Ergonomics and Safety, 1996, p. 523-528.
  54. Masuda T and Sadoyama T. Skeletal muscles from which the propagation of motor unit action potentials is detectable with surface electrode arrays. Electroencephalogr Clin Neurophysiol 67: 421-427, 1987.
  55. Merletti R, Farina D, and Gazzoni M. The linear electrode array: a useful tool with many applications. J Electromyogr Kinesiol 13: 37-47, 2003.
  56. Merletti R, Farina D, and Granata A. Non-invasive assessment of motor unit properties with linear electrode arrays. Electroencephalogr Clin Neurophysiol 50: 293-300, 1999.
  57. Merletti R, Knaflitz M, and De Luca CJ. Myoelectric manifestations of fatigue in voluntary and electrically elicited contractions. J Appl Physiol 69: 1810-1820, 1990.
  58. Merletti R and Lo Conte L. Surface EMG signal processing during isometric contractions. J Electromyogr Kinesiol 7: 241-250, 1997.
  59. Milner-Brown HS, Mellenthin M, and Miller RG. Quantifying human muscle strength, endurance and fatigue. Arch Phys Med Rehabil 67: 530-535, 1986.
  60. Mogk JP and Keir PJ. Crosstalk in surface electromyography of the proximal forearm during gripping tasks. J Electromyogr Kinesiol 13: 63-71, 2003.
  61. Moritani T and Muro M. Motor unit activity and surface electromyogram power spectrum during increasing force of contraction. Eur J Appl Physiol 56: 260-265, 1987.
  62. Morrenhof JW and Abbink HJ. Cross-correlation and cross-talk in surface electromyography. Electromyogr Clin Neurophysiol 25: 73-79, 1985.
  63. Myers LJ, Lowery M, O'Malley M, Vaughan CL, Heneghan C, St Clair Gibson A., Harley YXR, and Sreenivasan R. Rectification and non-linear pre-processing of EMG signals for cortico-muscular analysis. J Neurosci Methods 124: 157-165, 2003.
  64. Nieminen H and Takala EP. Evidence of deterministic chaos in the myoelectric signal. Electromyogr Clin Neurophysiol 36: 49-58, 1996.
  65. Nishizono H, Kurata H, and Miyashita M. Muscle fiber conduction velocity related to stimulation rate. Electroencephalogr Clin Neurophysiol 72: 529-534, 1989.
  66. Nordstrom MA, Fuglevand AJ, and Enoka RM. Estimating the strength of common input to human motoneurons from the cross-correlogram. J Physiol 453: 547-574, 1992.
  67. Paiss O and Inbar G. Autoregressive modelling of surface EMG and its spectrum with application to fatigue. IEEE Trans Biomed Eng 34: 761-770, 1987.
  68. Petrofsky JS and Lind AR. Frequency analysis of the surface EMG during sustained isometric contractions. Eur J Appl Physiol 43: 173-182, 1980.
  69. Rainoldi A, Galardi G, Maderna L, Comi G, Lo Conte L, and Merletti R. Repeatability of surface EMG variables during voluntary isometric contractions of the biceps brachii muscle. J Electromyogr Kinesiol 9: 105-119, 1999.
  70. Reucher H, Rau G, and Silny J. Spatial filtering of noninvasive multi-electrode EMG. Part I. Introduction to measuring technique and applications. IEEE Trans Biomed Eng 34: 98-105, 1987.
  71. Reucher H, Silny J, and Rau G. Spatial filtering of noninvasive multi-electrode EMG. Part II. Filter performance in theory and modeling. IEEE Trans Biomed Eng 34: 106-113, 1987.
  72. Roy SH, De Luca CJ, and Schneider J. Effects of electrode location on myoelectric conduction velocity and median frequency estimates. J Appl Physiol 61: 1510-1517, 1986.
  73. Saitou K, Masuda T, Michikami D, Kojima R, and Okada M. Innervation zones of the upper and lower limb muscles estimated by using multichannel surface EMG. J Hum Ergol (Tokyo) 29: 35-52, 2000.
  74. Solomonow M, Baten C, Smith J, Baratta R, Hermens H, D'Ambrosia R, and Shoji H. Electromyogram power spectra frequencies associated with motor unit recruitment strategies. J Appl Physiol 68: 1177-1185, 1990.
  75. Tesch PA, Komi PV, Jacobs I, Karlsson J, and Viitasalo JT. Influence of lactate accumulation of EMG frequency spectrum during repeated concentric contractions. Acta Physiol Scand 119: 61-67, 1983.
  76. Tokunaga T, Baba S, Tanaka M, Kashiwagi K, Kimura K, and Kawazoe T. Two-dimensional configuration of the myoneural junctions of human masticatory muscle detected with matrix electrode. J Oral Rehabil 25: 329-334, 1998.
  77. Troni W, Cantello R, and Rainero I. Conduction velocity along human muscle fibers in situ. Neurology 33: 1453-1459, 1983.
  78. Van Boxtel A and Shomaker LRB. Influence of motor unit firing statistics on the median frequency of the EMG power spectrum. Eur J Appl Physiol 52: 207-213, 1984.
  79. Viitasalo JT and Komi PV. Interrelationships of EMG signal characteristics at different levels of muscle tension during fatigue. Electromyogr Clin Neurophysiol 18: 167-178, 1978.
  80. Wakeling JM, Pascual SA, Nigg BM, and von Tscharner V. Surface EMG shows distinct populations of muscle activity when measured during sustained sub-maximal exercise. Eur J Appl Physiol 86: 40-47, 2001.
  81. Webber CL, Schmidt MA, and Walsh SM. Influence of isometric loading on biceps EMG dynamics as assessed by linear and non linear tools. J Appl Physiol 78: 814-822, 1995.
  82. Westbury JR and Shaughnessy TG. Associations between spectral representation of the surface electromyogram and fiber type distribution and size in human masseter muscle. Electromyogr Clin Neurophysiol 27: 427-435, 1987.
  83. Weytjens JL and van Steenberghe D. The effects of motor unit synchronization on the power spectrum of the electromyogram. Biol Cybern 51: 71-77, 1984.
  84. Winter DA, Fuglevand AJ, and Archer SE. Crosstalk in surface electromyography: theoretical and practical estimates. J Electromyogr Kinesiol 4: 15-26, 1994.
  85. Yao W, Fuglevand AJ, and Enoka RM. Motor-unit synchronization increases EMG amplitude and decreases force steadiness of simulated contractions. J Neurophysiol 83: 441-452, 2000.
  86. Zwarts MJ, Drost G, and Stegeman DF. Recent progress in the diagnostic use of surface EMG for neurological diseases. J Electromyogr Kinesiol 10: 287-291, 2000.
  87. Zwarts MJ and Stegeman DF. Multichannel surface EMG: basic aspects and clinical utility. Muscle Nerve 28: 1-17, 2003.



This article has been cited by other articles:


Home page
 J. Neurophysiol.Home page
T. W. Boonstra, M. Roerdink, A. Daffertshofer, B. van Vugt, G. van Werven, and P. J. Beek
Low-Alcohol Doses Reduce Common 10- to 15-Hz Input to Bilateral Leg Muscles During Quiet Standing
J Neurophysiol, October 1, 2008; 100(4): 2158 - 2164.
[Abstract] [Full Text] [PDF]

Home page
 J. Appl. Physiol.Home page
B. C. Clark, L. C. Issac, J. L. Lane, L. A. Damron, and R. L. Hoffman
Neuromuscular plasticity during and following 3 wk of human forearm cast immobilization
J Appl Physiol, September 1, 2008; 105(3): 868 - 878.
[Abstract] [Full Text] [PDF]

Home page
 J. Appl. Physiol.Home page
S. K. Hunter, T. Yoon, J. Farinella, E. E. Griffith, and A. V. Ng
Time to task failure and muscle activation vary with load type for a submaximal fatiguing contraction with the lower leg
J Appl Physiol, August 1, 2008; 105(2): 463 - 472.
[Abstract] [Full Text] [PDF]

Home page
 Integr. Comp. Biol.Home page
T. J. Roberts and A. M. Gabaldon
Interpreting muscle function from EMG: lessons learned from direct measurements of muscle force
Integr. Comp. Biol., August 1, 2008; 48(2): 312 - 320.
[Abstract] [Full Text] [PDF]

Home page
 J. Appl. Physiol.Home page
T. Rudroff, D. Staudenmann, and R. M. Enoka
Electromyographic measures of muscle activation and changes in muscle architecture of human elbow flexors during fatiguing contractions
J Appl Physiol, June 1, 2008; 104(6): 1720 - 1726.
[Abstract] [Full Text] [PDF]

Home page
 ptjournalHome page
L. L Andersen, M. Kjaer, C. H Andersen, P. B Hansen, M. K Zebis, K. Hansen, and G. Sjogaard
Muscle Activation During Selected Strength Exercises in Women With Chronic Neck Muscle Pain
Physical Therapy, June 1, 2008; 88(6): 703 - 711.
[Abstract] [Full Text] [PDF]

Home page
 J. Appl. Physiol.Home page
C. R. Abbiss, J. J. Peiffer, J. M. Peake, K. Nosaka, K. Suzuki, D. T. Martin, and P. B. Laursen
Effect of carbohydrate ingestion and ambient temperature on muscle fatigue development in endurance-trained male cyclists
J Appl Physiol, April 1, 2008; 104(4): 1021 - 1028.
[Abstract] [Full Text] [PDF]

Home page
 J. Neurophysiol.Home page
W. Taube, C. Leukel, M. Schubert, M. Gruber, T. Rantalainen, and A. Gollhofer
Differential Modulation of Spinal and Corticospinal Excitability During Drop Jumps
J Neurophysiol, March 1, 2008; 99(3): 1243 - 1252.
[Abstract] [Full Text] [PDF]

Home page
 J. Neurophysiol.Home page
T. J. Dartnall, M. A. Nordstrom, and J. G. Semmler
Motor Unit Synchronization Is Increased in Biceps Brachii After Exercise-Induced Damage to Elbow Flexor Muscles
J Neurophysiol, February 1, 2008; 99(2): 1008 - 1019.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Regul. Integr. Comp. Physiol.Home page
A. M. Gabaldon, F. E. Nelson, and T. J. Roberts
Relative shortening velocity in locomotor muscles: turkey ankle extensors operate at low V/Vmax
Am J Physiol Regulatory Integrative Comp Physiol, January 1, 2008; 294(1): R200 - R210.
[Abstract] [Full Text] [PDF]

Home page
 J. Neurophysiol.Home page
K. G. Keenan and F. J. Valero-Cuevas
Experimentally Valid Predictions of Muscle Force and EMG in Models of Motor-Unit Function Are Most Sensitive to Neural Properties
J Neurophysiol, September 1, 2007; 98(3): 1581 - 1590.
[Abstract] [Full Text] [PDF]

Home page
 J. Appl. Physiol.Home page
J. G. Semmler, K. J. Tucker, T. J. Allen, and U. Proske
Eccentric exercise increases EMG amplitude and force fluctuations during submaximal contractions of elbow flexor muscles
J Appl Physiol, September 1, 2007; 103(3): 979 - 989.
[Abstract] [Full Text] [PDF]

Home page
 J. Appl. Physiol.Home page
J.-F. Grosset, I. Mora, D. Lambertz, and C. Perot
Changes in stretch reflexes and muscle stiffness with age in prepubescent children
J Appl Physiol, June 1, 2007; 102(6): 2352 - 2360.
[Abstract] [Full Text] [PDF]

Home page
 ptjournalHome page
R. L Segal
Use of Imaging to Assess Normal and Adaptive Muscle Function
Physical Therapy, June 1, 2007; 87(6): 704 - 718.
[Abstract] [Full Text] [PDF]