- Email: [email protected]
The change in spatial distribution of upper trapezius muscle activity is correlated to contraction duration
Available online at www.sciencedirect.com
Journal of Electromyography and Kinesiology 18 (2008) 16–25 www.elsevier.com/locate/jelekin
The change in spatial distribution of upper trapezius muscle activity is correlated to contraction duration Dario Farina a,*, Fre´de´ric Leclerc a,b,c, Lars Arendt-Nielsen a, Olivier Buttelli b, Pascal Madeleine a a
b
Center for Sensory-Motor Interaction (SMI), Department of Health Science and Technology, Aalborg University, Fredrik Bajers Vej 7D-3, 9220 Aalborg East, Denmark Laboratoire Activite´ Motrice et Conception ergonomique, Universite´ d’Orle´ans, rue de Vendoˆme, BP 6237, 45062 Orle´ans cedex 2, France c Laboratoire d’Electronique, Signaux et Images, Universite´ d’Orle´ans, 12 rue de Blois, BP 6744, 45067 Orle´ans cedex 2, France Received 13 March 2006; received in revised form 3 August 2006; accepted 14 August 2006
Abstract The aim of the study was to confirm the hypothesis that the longer a contraction is sustained, the larger are the changes in the spatial distribution of muscle activity. For this purpose, surface electromyographic (EMG) signals were recorded with a 13 · 5 grid of electrodes from the upper trapezius muscle of 11 healthy male subjects during static contractions with shoulders 90 abducted until endurance. The entropy (degree of uniformity) and center of gravity of the EMG root mean square map were computed to assess spatial inhomogeneity in muscle activation and changes over time in EMG amplitude spatial distribution. At the endurance time, entropy decreased (mean ± SD, percent change 2.0 ± 1.6%; P < 0.0001) and the center of gravity moved in the cranial direction (shift 11.2 ± 6.1 mm; P < 0.0001) with respect to the beginning of the contraction. The shift in the center of gravity was positively correlated with endurance time (R2 = 0.46, P < 0.05), thus subjects with larger shift in the activity map showed longer endurance time. The percent variation in average (over the grid) root mean square was positively correlated with the shift in the center of gravity (R2 = 0.51, P < 0.05). Moreover, the shift in the center of gravity was negatively correlated to both initial and final (at the endurance) entropy (R2 = 0.54 and R2 = 0.56, respectively; P < 0.01 in both cases), indicating that subjects with less uniform root mean square maps had larger shift of the center of gravity over time. The spatial changes in root mean square EMG were likely due to spatially-dependent changes in motor unit activation during the sustained contraction. It was concluded that the changes in spatial muscle activity distribution play a role in the ability to maintain a static contraction. 2006 Elsevier Ltd. All rights reserved. Keywords: Multi-channel EMG; Muscle topography; Fatigue
1. Introduction During sustained contractions, fiber membrane and motor unit control properties change. Muscle fiber conduction velocity decreases (Bigland-Ritchie et al., 1981) and
*
Corresponding author. Tel.: +45 96358821; fax: +45 98154008. E-mail address: [email protected] (D. Farina).
1050-6411/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.jelekin.2006.08.005
the intracellular action potential becomes progressively longer in duration and smaller in amplitude (Hanson and Persson, 1971). Moreover, in submaximal contractions, motor units are additionally recruited to compensate for the decrease in twitch force (Fallentin et al., 1993; Garland et al., 1994), discharge rates usually decrease (De Luca et al., 1996), and de-recruitment may occur (Westgaard and De Luca, 2001). Surface electromyography (EMG) reflects these changes and usually shows an increase in
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
amplitude over time due to the increased number of active motor units and duration of the extracellular action potentials (Merletti et al., 1990). As compared to classic single-channel bipolar recording, high-density, two-dimensional surface EMG (Scholle et al., 1994; Zwarts and Stegeman, 2003) provides information on the spatial distribution of electric potential over the skin surface during muscle contraction. The technique consists in the placement of a number of closely located electrodes over the same muscle and allows a topographical representation of the muscle electrical activity over the skin plane (Holtermann et al., 2005; Lapatki et al., 2004; Zwarts and Stegeman, 2003). From surface EMG maps, it was shown that the spatial distribution of EMG amplitude is inhomogeneous (Holtermann et al., 2005) which underlines heterogeneity either in the distribution of the motor units within the muscle or in the strategy with which motor units are recruited (Holtermann et al., 2005). This may reflect muscle compartmentalization which seems particularly convenient in muscles with highly diversified biomechanical functions, e.g., upper trapezius (Johnson et al., 1994). The topographical EMG maps change with force level (Holtermann et al., 2005; Holtermann and Roeleveld, 2006) indicating that motor unit recruitment is not spatially uniform in the muscle (Hermans and Spaepen, 1997). Moreover, in the upper trapezius muscle, the skin spatial distribution of muscle activity varied over time with sustained contraction (Holtermann and Roeleveld, 2006), consistently with the expected changes due to orderly motor unit recruitment (Henneman, 1957). Changes in the spatial distribution of EMG amplitude maps were also observed in response to nociceptive afferent input (Madeleine et al., in press). Modifications in the surface EMG maps during sustained contraction indicate relative modifications in the intensity of activity within muscle regions. This may be due to dependence of fiber membrane properties on fiber location into the muscle, inhomogeneous motor unit recruitment or substitution. The control strategy has a major effect in the modifications of the spatial distribution of muscle activity during sustained contractions since similar changes in EMG maps were observed with increasing force and during fatigue development (Holtermann and Roeleveld, 2006). Variability in EMG activity among subdivisions of the erector spinae muscle resulted in longer endurance (van Dieen et al., 1993), highlighting the role of modifications in the spatial distribution of EMG activity in the maintenance of the force output. We hypothesized that sustaining a given force level for a longer period should correspond to a larger change in spatial distribution of trapezius muscle activity, which reflects sharing of the load among regions of the same muscle. Thus, in this study we tested the hypothesis that the degree of change in the spatial distribution of upper trapezius muscle activity over time is correlated to endurance time.
17
2. Methods 2.1. Subjects Eleven healthy, right-handed sedentary male subjects (age, mean ± SD, 28.2 ± 5.7 years, body mass 71.9 ± 6.6 kg, stature 1.79 ± 0.06 m, body mass index 22.6 ± 2.2) participated to the study. The study was conducted in accordance with the declaration of Helsinki, approved by the Local Ethics Committee (approval number VN 2004/56), and written informed consent was obtained from all participants prior to inclusion. 2.2. Surface EMG recording Surface EMG signals were detected from the right upper trapezius with a semi-disposable adhesive grid of 64 electrodes (LISiN-OT Bioelectronica, Torino, Italy, model ELSCH064). The grid is made of 13 rows and 5 columns of electrodes (1-mm diameter, 8-mm interelectrode distance in both directions) with one missing electrode at the upper right corner (Fig. 1). The position corresponding to the missing electrode was used as the origin of the coordinate system to define electrode location. The grid was connected to the amplifier through 4 connectors which were fixed at the subject skin by adhesive tape. The bipolar EMG signals were amplified (64-channel surface EMG amplifier, SEA 64, LISiN-OT Bioelectronica, Torino, Italy; 3 dB bandwidth 10– 500 Hz) by a factor 5000, sampled at 4096 Hz per channel, and converted to digital form by a 12-bit A/D converter. Before placement of the grid, the main innervation zone location of the upper trapezius was identified with an array of 8 electrodes (silver bars, 5-mm long, 1-mm diameter, 5-mm interelectrode distance), located along the C7-acromion line, as previously described (Farina et al., 2002). The distance from acromion to C7 was (mean ± SD) 22.4 ± 0.7 cm and the distance from C7 to the main innervation zone was 11.1 ± 0.5 cm. The electrode grid was placed with the 4th row along the C7-acromion line and with the most medial electrode column 10-mm distant from the innervation zone location (Fig. 1). The part of the skin where the grid was located was slightly abraded with abrasive paste (Medic-Every, Parma, Italy). 30 lL of conductive gel were inserted into the cavities of the grid to assure proper electrodeskin contact. A reference electrode was placed at the right wrist. 2.3. General procedures The subject sat comfortably on a chair with feet parallel on the floor at a distance approximately equal to the distance between the left and right acromion. After placement of the EMG electrode grid (see below), the subject rested for 5-min. He was then asked to abduct both arms at 90 with elbows fully extended and forearms 90 pronated, with palm facing toward the ground and without hand load. Two flexible bars placed at shoulder level were used to provide tactile position feedback to the subject. Moreover, two templates placed behind and on the side of the subject were used by the experimenter to ensure the same position of the neck and head during the contraction. The static contraction was maintained until endurance, i.e., until it was not possible for the subject to maintain the arms 90 abducted in touch with the flexible bars after strong verbal encouragement. According to previous data, the shoulder moment generated by the active muscles was 14.5 ± 1.0 Nm, corresponding to 15–20% of the maximum voluntary contraction of the trapezius muscle
18
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
Fig. 1. (a) The EMG electrode grid placed on the right upper trapezius of a subject. (b) Schematic representation of the electrode grid with the indication of the coordinate axes and origin for defining electrode position.
(Mathiassen et al., 1995). Skin temperature was monitored (Ellab, Copenhagen, Denmark) from two locations over the upper trapezius, proximal and distal with respect to the electrode grid along the seventh cervical vertebra (C7)-acromion line. Room temperature was kept stable at 24 C (±1 C). 2.4. Data processing EMG signals were off-line band-pass filtered (second order Butterworth filter; 3 dB bandwidth, 10–400 Hz). Fifty-one bipolar EMG signals were obtained from the grid (13 · 4 bipolar recordings with one missing electrode). The bipolar systems were parallel to fiber direction according to Jensen and Westgaard (1997) (Fig. 1). Mean power spectral frequency and root mean square value were computed from each bipolar recording from adjacent, non-overlapping signal epochs of 1-s duration, as described previously (Merletti et al., 1990; Welch, 1967). For graphical representation, the 51 values were interpolated by a factor 8 but only the original values were used for data processing and statistics. To characterize the spatial distribution of muscle activity, the following variables were extracted from the 51 bipolar signals: root mean square and mean power spectral frequency averaged over the 51 signals (RMSmean, MNFmean), root mean square and mean frequency corresponding to the bipolar signal with maximum root mean square in the first second of the contraction (RMSmax, MNFmax), the two coordinates of the center of gravity of the root mean square map (Gx and Gy for the lateral–medial and cranial–caudal direction, respectively), and the modified entropy of the 51 root mean square values. RMSmax and MNFmax corresponded to a fixed electrode location over the contraction. Modified entropy was computed as follows:
Entr ¼
51 X i¼1
p2 ðiÞlog2 p2 ðiÞ
where p2(i) is the square of the root mean square value at electrode i normalized by the summation of the squares of the 51 root mean square values. Entropy is a measure of uniformity of values (Proakis, 2001), used for source coding in communication engineering. The entropy of a set of M values is maximum if all values are the same (uniform distribution) and takes the value log2 M while its minimum value is zero. In the context of spatial muscle activity distribution, entropy indicates the degree of homogeneity in activation, with higher values corresponding to more uniform distribution of the root mean square values over the grid. Five consecutive values over time of RMSmean, RMSmax, MNFmean, MNFmax, Gx, Gy, and entropy were averaged to obtain mean values corresponding to 0%, 25%, 50%, 75%, and 100% of the endurance time. Percent changes of RMSmean, RMSmax, MNFmean, and MNFmax over the endurance time were defined as the difference between initial (0% endurance time) and final (100% endurance time) values normalized by the corresponding initial values and expressed in percentage. The shift of the center of gravity was defined as the modulus of the center of gravity displacement in medial–lateral and cranial–caudal direction at the beginning (0% endurance time) and end (100% endurance time) of the contraction. 2.5. Statistical analysis Linear regression analysis was used to investigate the relation between the shift of the center of gravity and endurance time, rate of change of root mean square and mean frequency. One-way repeated measures analysis of variance (ANOVA) was used to assess dependency of the EMG variables on time (0–100% endurance time, 25% increments) and was followed by post–hoc Student–Newman–Keuls (SNK) pair-wise comparisons, when appropriate. Student t-test was used to compare skin temperature at the beginning and end of the contraction. Significance was accepted for P-values less than 0.05. Results are reported as mean and standard deviation (SD) in the text and standard error (SE) in the figures.
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
3. Results Average endurance time was 382.6 ± 163.7 s. Skin temperature did not change significantly from the beginning to the end of the contraction (from 32.7 ± 1.5 C to 33.2 ± 1.1 C) and with respect to location (33.2 ± 1.7 C and 33.7 ± 1.1 C for proximal and distal location, respectively). The root mean square maps showed two main areas of high activity and a line of low activity, corresponding to the location of the innervation zones (Fig. 2), as confirmed by the inversion of the action potential polarity (visual
19
inspection of the signals). Mean power spectral frequency maps showed high values in correspondence to the innervation zone and lower values in the high-amplitude region, in agreement with previous results (Farina et al., 2002; Kleine et al., 2000) (Fig. 2). 3.1. Trends over time of EMG variables Average RMSmax across subjects did not significantly change over time due to a large variability of behaviors across subjects (percent increase at the end with respect to
Fig. 2. Topographical map (interpolation by a factor 8) of root mean square (a) and mean power spectral frequency (b) at 50% endurance time in one subject. Innervation zone location corresponds to low amplitude and high frequency values.
Fig. 3. Mean (±SE) root mean square (RMS) averaged over the grid (a), mean power spectral frequency (MNF) averaged over the grid (b), x- (empty circles) and y- (filled circles) coordinate of the center of gravity (c), and entropy (d), for 0%, 25%, 50%, 75%, and 100% of the endurance time.
20
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
Fig. 4. Root mean square topographical maps at 0%, 25%, 50%, 75%, and 100% endurance time in one subject. The black circle in the maps indicates the center of gravity. The grey scale is the same for the five maps.
the beginning, 7.5 ± 41.2%). For seven subjects RMSmax increased while the opposite was observed for the other subjects. On the contrary, RMSmean increased over time (F = 9.9, P < 0.0001) and was larger at 100% contraction time than at all other contraction times (P < 0.05), and at 0% smaller than at 50%, 75%, and 100% (P < 0.05) (Fig. 3a). Average MNFmax across subjects did not depend on time which was due, as for RMSmax, to the large variability across subjects (percent decrease, 2.3 ± 15.5%). MNFmean decreased over time (F = 10.5, P < 0.0001), and was larger at 0% contraction time than at the other contraction times (P < 0.001) (Fig. 3b).
Center of gravity shift was also positively correlated with the percent change in MNFmax (R2 = 0.44, P < 0.05), with subjects showing large shift increasing MNFmax, as opposite to subjects with smaller changes (Fig. 5c). Center of gravity shift was not correlated with MNFmean (R2 = 0.03) while it was negatively correlated to both initial and final (contraction times 0% and 100% of the endurance time) entropy (R2 = 0.54 and R2 = 0.56, respectively; P < 0.01 in both cases) (Fig. 5d), indicating that subjects with less uniform root mean square maps had larger shifts of the center of gravity. 4. Discussion
3.2. Changes over time of the root mean square map Fig. 4 shows an example of root mean square maps at the five time instants considered. The y-coordinate of the root mean square map decreased (i.e., cranial shift of the center of gravity) over time (F = 25.9, P < 0.0001), with all time instants different between each other (P < 0.01), except for 50/75% and 75/100% (Fig. 3c). The x-coordinate increased (i.e., lateral shift of the center of gravity) over time (F = 18.6, P < 0.0001), with differences between all time instants except 50/75% and 75/100% (Fig. 3c). However, the absolute shift of the x-coordinate (1.0 ± 0.6 mm over the endurance time) was negligible with respect to the shift in the y-coordinate (11.2 ± 6.1 mm). Entropy of the root mean square map significantly decreased over time (less uniform map) (F = 8.5, P < 0.0001), with 25% contraction time different from 75% (P < 0.05) and 100% smaller than all others (P < 0.05) (Fig. 3d). 3.3. Correlations between changes in EMG variables and endurance time Subjects whose activity map shifted more (displacement of the center of gravity) had longer endurance times (R2 = 0.46, P < 0.05) (Fig. 5a). Subjects with large shift showed a decreased RMSmax at the endurance with respect to the beginning of the contraction while the opposite was observed in subjects with small shift (R2 = 0.39, P < 0.05) (Fig. 5b). Percent change in RMSmean was on the contrary positively correlated with the shift in the center of gravity (R2 = 0.51, P < 0.05) and was larger at the endurance time with respect to the beginning of the contraction in all subjects.
This study shows for the first time that the amount of shift in the spatial distribution of EMG activity of the upper trapezius muscle is related to the time during which a static contraction is maintained. This indicates that changes in spatial muscle activity distribution, estimated by changes in the center of gravity of the root mean square map and entropy, play a role in the ability to maintain a static contraction. 4.1. Spatial changes in topographical EMG maps The topographical map of EMG activity changed over time. This has been quantified by the position of the center of gravity and by the entropy of root mean square values. The changes included a cranial and lateral shift of the center of gravity and a decrease in entropy, reflecting less uniform map. The observed cranial shift of muscle activity is in agreement with previous work (Kleine et al., 2000; Madeleine et al., in press). The observed heterogeneity in muscle activation supports the hypothesis of functional subdivision of the upper trapezius muscle during low level activation (Jensen and Westgaard, 1997; Mathiassen et al., 1995). The temporal changes in spatial activity distribution may have been due to both peripheral and central mechanisms. Peripheral mechanisms include the change in muscle fiber membrane properties with fatigue that has an effect on EMG amplitude (Merletti et al., 1990). Muscle fiber conduction velocity decreases with sustained contraction due to the increase in interstitial potassium concentration with repetitive fiber activation (Kossler et al., 1991). The decrease in conduction velocity corresponds to increased
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
21
Fig. 5. Scatter plots of the shift of the center of gravity vs endurance time (a), percent change in maximum root mean square (RMSmax, see text for its definition) (b), percent change in maximum mean power spectral frequency (MNFmax, see text for its definition) (c), and entropy at the beginning (0% endurance time, filled circles) and end (100% endurance time, empty circles) of the contraction (d).
duration of the action potentials (Lindstrom and Magnusson, 1977) and thus to increased root mean square in case of unchanged intracellular potential (Merletti et al., 1990). On the other hand, the increase in potential duration also increases amplitude cancellation (Keenan et al., 2005), although this phenomenon is more limited than the direct effect of conduction velocity on action potential area (Keenan et al., 2005). Finally, the intracellular action potential shape may change with sustained contraction (Hanson and Persson, 1971) which affects EMG amplitude (Dimitrova and Dimitrov, 2003). The change in root mean square distribution may be due to an inhomogeneous modification over time in any of the variables related to fiber membrane properties. Inhomogeneous changes in membrane properties may be due to a different distribution of fiber types in different regions of the upper trapezius (Lindman et al., 1990) or to a distribution of intramuscular pressure which may affect the rate of metabolic wash-out during sustained contraction. However, a pure peripheral effect in the modification of EMG maps is unlikely since in four subjects there was a reduction in root mean square in the location where at the beginning it was at its maximum (Fig. 5b). The decrease in conduction velocity is probably the main determinant, among the peripheral effects, of the changes in EMG amplitude (Merletti et al., 1990) and mainly determines an increase in amplitude. Moreover, for subjects who showed large changes in the location of the center of gravity, mean power spectral frequency at the location with maximal initial root mean square increased over time
(Fig. 5c). This cannot be explained by decreased conduction velocity without assuming a change in motor control strategies. Indeed, without any change in the number of active motor units and in their discharge rates, mean frequency and conduction velocity show similar trends over time (Lindstrom and Magnusson, 1977). Thus, spatial changes in root mean square maps cannot solely be explained by changes at the muscle level. Motor unit control strategies change during sustained contraction. In particular, motor unit discharge rate usually decreases with fatigue (Bigland-Ritchie et al., 1983; De Luca et al., 1996), probably due to reflex inhibition from small diameter muscle afferents (Bigland-Ritchie et al., 1986; Gandevia, 2001; Woods et al., 1987). EMG amplitude is directly related to motor unit discharge rate, thus space-dependent decrease in discharge rate may explain variations in the root mean square map. There is also evidence of additional recruitment of motor units during sustained static contractions (Fallentin et al., 1993; Garland et al., 1994) and some reports indicate motor unit derecruitment and substitution in order to cope with fatigue development (Westad et al., 2003; Westgaard and De Luca, 2001). The way in which motor units are recruited during a static contraction was related to the endurance time, as observed comparing similar tasks with different loads (Maluf and Enoka, 2005). Recruitment/derecruitment of motor units likely change the spatial activity distribution since distribution of motor unit types and diameters in the upper trapezius is not homogeneous (Lindman et al., 1990). Moreover, the changes in root mean square maps
22
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
over time in static contractions resemble the changes which occur with increasing force in non-fatiguing contractions (Holtermann and Roeleveld, 2006), indicating an orderly recruitment with sustained contraction. Degree of motor unit synchronization may also vary with fatigue although the effect of synchronization on surface EMG amplitude may be limited (Zhou and Rymer, 2004). 4.2. Time to task failure and EMG maps The present study showed a correlation between endurance time and the amount of shift in the center of gravity of root mean square maps. Subjects with larger shift had longer endurance time, in agreement with results reported for different subdivisions of the erector spinae muscle using multiple bipolar recordings (van Dieen et al., 1993). This indicates a functional significance of modifications in muscle activity spatial distribution. The range of variation for the shift of the center of gravity varied from approximately 2 to 17 mm (Fig. 5). The subjects who showed large shift were those with the most heterogeneous root mean square map (minimum entropy) (Fig. 5d). The differences among subjects in the amount of shift in the center of gravity may be related to a different strategy to cope with fatigue, in particular in relation to motor unit recruitment/derecruitment. During the contraction, for some subjects and locations over the muscle, the EMG activity decreased over time (Fig. 5b), as opposed to an average increase in root mean square (Fig. 3a), which indicates localized decreased excitation level to the muscle. This may be due to either derecruitment of motor units or spatially dependent decrease in discharge rate. In both cases, the results indicate a space-dependant modification in the neural drive to the muscle which allowed different parts of the muscle to be active at different stages of the fatigue development in order to maintain the same force output. One possible interpretation of these results is that the reorganization of muscle activity with sustained contraction had the functional significance of reducing fatigue development. This is in agreement with the hypothesis that spatial reorganization of muscle activity reduces muscle fiber overload and leads to more uniform changes in the extra-cellular environment in different muscle regions. For example, motor strategies with increased degree of freedom (Madeleine et al., 2003; Mathiassen et al., 2003) are less susceptible to lead to work related musculoskeletal disorders. However, the correlation between endurance time and changes in spatial EMG distribution may also be due to different activation levels of the muscle at the beginning of the contraction. The relative load sustained with respect to maximal force may have indeed been different among subjects. Thus, subjects exerting more relative force may have sustained the contraction more than subjects exerting lower force. Since maximal force was not measured in this study, this interpretation cannot be completely excluded. However, it is not in agreement with the results in van
Dieen et al. (1993) and with the observed higher uniformity in the maps of the subjects who maintained the task for shorter time. If these subjects were also those for whom the relative force level was higher, a larger heterogeneity, corresponding to higher force would have been expected. The degree of heterogeneity increased with fatigue and since similar changes in surface EMG maps are observed with increasing force and with fatigue (Holtermann and Roeleveld, 2006), higher force should correspond to higher degree of heterogeneity. From the concomitant observation of a correlation between shift of the maps and entropy with endurance time, the explanation based purely on differences in initial relative forces seems very unlikely. 4.3. Interpretation of surface EMG variables The present results underline difficulties in the interpretation of EMG variables estimated from signals recorded from a single muscle site. This is due to the large variability among subjects in spatial changes over time. Usually, trends over time of mean power spectral frequency and root mean square are used as indexes of muscle fatigue. This assumption is based on the relation between muscle fiber conduction velocity, mean power frequency, and EMG amplitude (Lindstrom and Magnusson, 1977; Merletti et al., 1990). However, when the number of motor units or distribution of discharge rates is not the same over time, these relations are not valid. Depending on the amount of change in the position of the center of gravity, RMSmax and MNFmax decreased or increased (Fig. 5b and c). Subjects with larger shift of the center of gravity showed decreased RMSmax and increased MNFmax, contrary to the average trend of root mean square and mean frequency over the grid (Fig. 3a and b), while the opposite trends were observed for subjects with smaller shift. This was probably related to a larger variation in the number of active motor units over time for subjects with larger changes in the spatial distribution of root mean square. Variations in the number of active motor units may indeed compensate for the decrease in mean frequency (Farina et al., 2006). Variability in the spatial changes of muscle activity among subjects may thus explain the controversial results reported in the literature on trends over time of EMG variables during static contractions of the upper trapezius (compare, e.g., Farina et al., 2002; Madeleine et al., ¨ berg et al., 1992; Sadoyama 2002; Mathiassen et al., 1995; O et al., 2000). This problem is more important at low contraction levels when the number of active motor units at the beginning of the task does not correspond to full recruitment. Topographical mapping allows us to take this variability into consideration. 4.4. Limitations The lateral shift (x-coordinate) of the center of gravity of root mean square maps was not expected in case the grid was placed exactly in line with the fiber orientation and
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
between the innervation zone and tendon. Indeed, in these conditions the x-direction of the grid would correspond to the direction of propagation of the action potentials and the distribution of amplitude should not be altered (if the fiber is activated, the action potential always propagates along the entire fiber semilength, from innervation zone to tendon). The shift of the center of gravity in the medial–lateral direction, which was anyway negligible compared with the cranial–caudal shift, was probably due to slight misalignment of the grid transversal side with respect to the fiber orientation or to changes in location of the innervation zone over time. This does not affect the modulus of the center of gravity displacement and thus the conclusions drawn. The position of the center of gravity and the entropy were computed from the entire maps, although some parts of the grid covered the innervation zones of the muscle. Since the action potentials have lower amplitude in correspondence to innervation zone, spatial heterogeneity may have been simply resulted from anatomical factors. Absolute values of descriptors of heterogeneity (e.g., entropy) are affected by this problem. However, the relative changes in the variables investigated, on which most of the conclusions are based, were not influenced by this issue. The type of analysis performed is global and aims at assessing spatial modifications. Even in case the maps changed due to recruitment of motor units with different locations of the innervation zones over time (thus affecting the map for changed anatomical properties), the global variables used would have been valuable in underlining strategies in motor unit recruitment. The selected task is not specific for the upper trapezius, thus endurance may have been limited by other muscles rather the investigated one, for example the deltoid. However, even if this was the case, the observed correlations between spatial modifications and duration of the contraction indicate that the larger the modifications in upper trapezius spatial activity distribution, the longer the contraction was sustained, which is the main conclusion of the study. 4.5. Conclusion The study shows correlation between change in the spatial distribution of muscle activity over time and endurance time. Subjects with more heterogeneous activity and larger shift in the upper trapezius activity toward the cranial direction could sustain the static contraction longer than subjects with more uniform activity distribution and smaller changes in the EMG map over time. A likely explanation of these results is that heterogeneity in muscle activity and spatial adaptation over time have the functional role of prolonging sustenance of a static task. 5. Grants The Marie Curie fellowship of the European Commission program ‘‘Improving the human research potential
23
and the socio-economic knowledge’’ based under the contract number HPMT-CT-2000-00092 supported Frederic Leclerc. The Danish Technical Research Council (Program ‘‘Centre for Neuroengineering (CEN)’’, contract number 26-04-0100) partly supported the study. Acknowledgements The system for two-dimensional EMG detection (amplifier and sensor) has been developed at the Laboratory for Neuromuscular System Engineering (LISiN), Torino, Italy. The authors are sincerely grateful to Roberto Merletti and Andrea Bottin (LISiN) for the support in the use of the system. References Bigland-Ritchie B, Donovan EF, Roussos CS. Conduction velocity and EMG power spectrum changes in fatigue of sustained maximal efforts. J Appl Physiol 1981;51:1300–5. Bigland-Ritchie B, Johansson R, Lippold OC, Smith S, Woods JJ. Changes in motoneurone firing rates during sustained maximal voluntary contractions. J Physiol 1983;340:335–46. Bigland-Ritchie BR, Dawson NJ, Johansson RS, Lippold OC. Reflex origin for the slowing of motoneurone firing rates in fatigue of human voluntary contractions. J Physiol 1986;379:451–9. De Luca CJ, Foley PJ, Erim Z. Motor unit control properties in constantforce isometric contractions. J Neurophysiol 1996;76:1503–16. Dimitrova NA, Dimitrov GV. Interpretation of EMG changes with fatigue: facts, pitfalls, and fallacies. J Electromyogr Kinesiol 2003;13: 13–36. Fallentin N, Jorgensen K, Simonsen EB. Motor unit recruitment during prolonged isometric contractions. Eur J Appl Physiol Occup Physiol 1993;67:335–41. Farina D, Madeleine P, Graven-Nielsen T, Merletti R, Arendt-Nielsen L. Standardising surface electromyogram recordings for assessment of activity and fatigue in the human upper trapezius muscle. Eur J Appl Physiol 2002;86:469–78. Farina D, Zennaro D, Pozzo M, Merletti R, Laubli T. Single motor unit and spectral surface EMG analysis during low-force, sustained contractions of the upper trapezius muscle. Eur J Appl Physiol 2006;96:157–64. Gandevia SC. Spinal and supraspinal factors in human muscle fatigue. Physiol Rev 2001;81:1725–89. Garland SJ, Enoka RM, Serrano LP, Robinson GA. Behavior of motor units in human biceps brachii during a submaximal fatiguing contraction. J Appl Physiol 1994;76:2411–9. Hanson J, Persson A. Changes in the action potential and contraction of isolated frog muscle after repetitive stimulation. Acta Physiol Scand 1971;81:340–8. Henneman E. Relation between size of neurons and their susceptibility to discharge. Science 1957;126:1345–7. Hermans V, Spaepen AJ. Influence of electrode position on changes in electromyograph parameters of the upper trapezius muscle during submaximal sustained contractions. Eur J App Physiol 1997;75: 319–25. Holtermann A, Roeleveld K, Karlsson JS. Inhomogeneities in muscle activation reveal motor unit recruitment. J Electromyogr Kinesiol 2005;15:131–7. Holtermann A, Roeleveld K. Global muscle activation in sustained contractions. Acta Physiol (Oxf) 2006;186:159–68. Jensen C, Westgaard RH. Functional subdivision of the upper trapezius muscle during low-level activation. Eur J Appl Physiol 1997;76: 335–9.
24
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
Johnson G, Bogduk N, Nowitzke A, House D. Anatomy and actions of the trapezius muscle. Clin Biomech 1994;9:44–50. Keenan KG, Farina D, Maluf KS, Merletti R, Enoka RM. Influence of amplitude cancellation on the simulated surface electromyogram. J Appl Physiol 2005;98:120–31. Kleine BU, Schumann NP, Stegeman DF, Scholle HC. Surface EMG mapping of the human trapezius muscle: the topography of monopolar and bipolar surface EMG amplitude and spectrum parameters at varied forces and in fatigue. Clin Neurophysiol 2000;111:686–93. Kossler F, Lange F, Caffier G, Kuchler G. External potassium and action potential propagation in rat fast and slow twitch muscles. Gen Physiol Biophys 1991;10:485–98. Lapatki BG, Van Dijk JP, Jonas IE, Zwarts MJ, Stegeman DF. A thin, flexible multielectrode grid for high-density surface EMG. J Appl Physiol 2004;96:327–36. Lindman R, Eriksson A, Thornell LE. Fiber type composition of the human male trapezius muscle: enzyme-histochemical characteristics. Am J Anat 1990;189:236–44. Lindstrom L, Magnusson R. Interpretation of myoelectric power spectra: a model and its applications. Proc IEEE 1977;65:653–62. Madeleine P, Farina D, Merletti R, Arendt-Nielsen L. Upper trapezius muscle mechanomyographic and electromyographic activity in humans during low force fatiguing and non-fatiguing contractions. Eur J Appl Physiol 2002;87:327–36. Madeleine P, Lundager B, Voigt M, Arendt-Nielsen L. The effects of neckshoulder pain development on sensory-motor interactions among female workers in the poultry and fish industries. A prospective study. Int Arch Occup Environ Health 2003;76:39–49. Madeleine P, Leclerc F, Arendt-Nielsen L, Ravier P, Farina D. Experimental muscle pain changes the spatial distribution of upper trapezius muscle activity during sustained contraction. Clin Neurophysiol [in press]. Maluf KS, Enoka RM. Task failure during fatiguing contractions performed by humans. J Appl Physiol 2005;99:389–96. Mathiassen SE, Winkel J, Ha¨gg GM. Normalisation of surface EMG ampltude from the upper trapezius muscle in ergonomic studies – A review. J Electromyograph Kinesiol 1995;5:197–226. Mathiassen SE, Moller T, Forsman M. Variability in mechanical exposure within and between individuals performing a highly constrained industrial work task. Ergonomics 2003;46:800–24. Merletti R, Knaflitz M, De Luca CJ. Myoelectric manifestations of fatigue in voluntary and electrically elicited contractions. J Appl Physiol 1990;69:1810–20. ¨ berg T, Sandsjo L, Kadefors R, Larsson SE. Electromyographic changes O in work-related myalgia of the trapezius muscle. Eur J-Appl Physiol 1992;65:251–7. Proakis J. Digital communications. New York: McGraw Hill; 2001. Sadoyama T, Sakai K, Hosoya S , Kamijyo M, Masuda T. Changes in surface EMG during static fatiguing contractions in the upper trapezius muscle. In: Proceedings of the 13th ISEK congress, Sapporo; 2000. Scholle HC, Schumann NP, Anders C. Quantitative-topographic and temporal characterization of myoelectrical activation patterns: new diagnostic possibilities in neurology, physiotherapy and orthopaedics. Funct Neurol 1994;9:35–45. van Dieen JH, Oude Vrielink HH, Toussaint HM. An investigation into the relevance of the pattern of temporal activation with respect to erector spinae muscle endurance. Eur J Appl Physiol Occup Physiol 1993;66:70–5. Welch PD. Use of fast Fourier transform for estimation of power spectra A method based on time averaging over short modified periodograms. IEEE Trans Audio Electroacoust 1967;15:70–3. Westad C, Westgaard RH, De Luca CJ. Motor unit recruitment and derecruitment induced by brief increase in contraction amplitude of the human trapezius muscle. J Physiol 2003;552(Pt 2):645–56. Westgaard RH, De Luca CJ. Motor control of low-threshold motor units in the human trapezius muscle. J Neurophysiol 2001;85:1777–81. Woods JJ, Furbush F, Bigland-Ritchie B. Evidence for a fatigue-induced reflex inhibition of motoneuron firing rates. J Neurophysiol 1987;58: 125–37.
Zhou P, Rymer WZ. Factors governing the form of the relation between muscle force and the EMG: a simulation study. J Neurophysiol 2004;92:2878–86. Zwarts MJ, Stegeman DF. Multichannel surface EMG: basic aspects and clinical utility. Muscle Nerve 2003;28:1–17.
Dario Farina graduated summa cum laude in Electronics Engineering from Politecnico di Torino, Torino, Italy, in February 1998. During 1998 he was a Fellow of the Laboratory for Neuromuscular System Engineering in Torino. In 2001 and 2002 he obtained the Ph.D. degrees in Automatic Control and Computer Science and in Electronics and Communications Engineering from the Ecole Centrale de Nantes, Nantes, France, and Politecnico di Torino, respectively. In 1999–2004 he taught courses in Electronics and Mathematics at Politecnico di Torino and in 2002–2004 he was Research Assistant Professor at the same University. Since 2004, he is Associate Professor in Biomedical Engineering at the Faculty of Engineering and Science, Department of Health Science and Technology of Aalborg University, Aalborg, Denmark, where he teaches courses on biomedical signal processing, modeling, and neuromuscular physiology. He regularly acts as referee for approximately 20 scientific International Journals, is an Associate Editor of IEEE Transactions on Biomedical Engineering, is on the Editorial Boards of the Journal of Neuroscience Methods, the Journal of Electromyography and Kinesiology, and Medical & Biological Engineering & Computing, and member of the Council ISEK (International Society of Electrophysiology and Kinesiology). His main research interests are in the areas of signal processing applied to biomedical signals, modeling of biological systems, basic and applied physiology of the neuromuscular system, and brain-computer interfaces. Dr. Farina is a Registered Professional Engineer in Italy.
Fre´de´ric Leclerc was born in 1980. He graduated in Electronics and Signal-Processing Engineering from Polytech’Orleans in September 2004. The same year he obtained the Master of Science Degree in Signal-Images in Biology and Medicine from Angers University where he developed an Holter system. Currently he is a Ph.D. candidate in the Laboratory of Electronic Signals Images of Orleans, France, working on a project on the development of methods for information extraction and interpretation from electrophysiological measures. His main interests are related to the development of acquisition systems, software and processing techniques for electromyographic and electrocardiographic signals.
Lars Arendt-Nielsen, born in 1958, received the M.Sc.E.E. degree from Aalborg University, Denmark, in 1983, with specialisation in biomedical engineering, and the Ph.D. degree in 1992. In 1994 he received his Dr. Sci. degree in Medicine from the Medical Faculty, Aarhus University, Denmark. From 1983 to 1984 he was a Research fellow, Department of Clinical Neurophysiology, The National Hospital for Nervous Diseases, London. Since 1988 he has been with the Department of Medical Informatics and Image Analysis, Aalborg University as an Associated Professor. In 1993 he was appointed Professor in Biomedical Engineering and Principal investigator at Center for Sensory-Motor Interaction, which was established in 1993 at Aalborg University and in 1997 Head of the International Doctoral
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25 School in Biomedical Science and Engineering, Aalborg University, with 55 Ph.D. students enrolled. During his career he has worked as guest professor in Japan and Australia. He is member of the Danish Research Council and the Danish Research Education Council. He has published approx. Five hundred and ten scientific papers within neuroscience with focus on motor control and pain research and given more than 110 keynote lectures at international conferences. Olivier Buttelli was born in 1966. He received the MS and Ph.D. degrees in biomechanics and motor physiology in 1991 and 1996, respectively, both from the University of Paris 11 Orsay. Since 1997, he is Assistant Professor in biomechanics and exercise physiology at the sport exercise faculty of Orleans University. His researches focus on fatigue and exercise.
25
Pascal Madeleine was born in Toulouse, France, in 1969. He received the M.Sc. degree in biomedical engineering with an electronic specialisation in 1991 from Paul Sabatier University, Toulouse, France, the Ph.D. qualifying degree in biomedical engineering with a biosignal processing specialisation in 1993 from Paris XII-Val de Marne University, Creteil, France and the Ph.D. degree in 1998 from Aalborg University, Denmark. He is currently employed as an associate professor at the Center for Sensory-Motor Interaction (SMI), Department of Health Science and Technology at Aalborg University, Denmark. His main areas of research interests are the interactions between muscle pain and motor control in relation to work related musculo-skeletal disorders via mechanomyographic, electromyographic, kinetic and kinematic data signal processing.
Journal of Electromyography and Kinesiology 18 (2008) 16–25 www.elsevier.com/locate/jelekin
The change in spatial distribution of upper trapezius muscle activity is correlated to contraction duration Dario Farina a,*, Fre´de´ric Leclerc a,b,c, Lars Arendt-Nielsen a, Olivier Buttelli b, Pascal Madeleine a a
b
Center for Sensory-Motor Interaction (SMI), Department of Health Science and Technology, Aalborg University, Fredrik Bajers Vej 7D-3, 9220 Aalborg East, Denmark Laboratoire Activite´ Motrice et Conception ergonomique, Universite´ d’Orle´ans, rue de Vendoˆme, BP 6237, 45062 Orle´ans cedex 2, France c Laboratoire d’Electronique, Signaux et Images, Universite´ d’Orle´ans, 12 rue de Blois, BP 6744, 45067 Orle´ans cedex 2, France Received 13 March 2006; received in revised form 3 August 2006; accepted 14 August 2006
Abstract The aim of the study was to confirm the hypothesis that the longer a contraction is sustained, the larger are the changes in the spatial distribution of muscle activity. For this purpose, surface electromyographic (EMG) signals were recorded with a 13 · 5 grid of electrodes from the upper trapezius muscle of 11 healthy male subjects during static contractions with shoulders 90 abducted until endurance. The entropy (degree of uniformity) and center of gravity of the EMG root mean square map were computed to assess spatial inhomogeneity in muscle activation and changes over time in EMG amplitude spatial distribution. At the endurance time, entropy decreased (mean ± SD, percent change 2.0 ± 1.6%; P < 0.0001) and the center of gravity moved in the cranial direction (shift 11.2 ± 6.1 mm; P < 0.0001) with respect to the beginning of the contraction. The shift in the center of gravity was positively correlated with endurance time (R2 = 0.46, P < 0.05), thus subjects with larger shift in the activity map showed longer endurance time. The percent variation in average (over the grid) root mean square was positively correlated with the shift in the center of gravity (R2 = 0.51, P < 0.05). Moreover, the shift in the center of gravity was negatively correlated to both initial and final (at the endurance) entropy (R2 = 0.54 and R2 = 0.56, respectively; P < 0.01 in both cases), indicating that subjects with less uniform root mean square maps had larger shift of the center of gravity over time. The spatial changes in root mean square EMG were likely due to spatially-dependent changes in motor unit activation during the sustained contraction. It was concluded that the changes in spatial muscle activity distribution play a role in the ability to maintain a static contraction. 2006 Elsevier Ltd. All rights reserved. Keywords: Multi-channel EMG; Muscle topography; Fatigue
1. Introduction During sustained contractions, fiber membrane and motor unit control properties change. Muscle fiber conduction velocity decreases (Bigland-Ritchie et al., 1981) and
*
Corresponding author. Tel.: +45 96358821; fax: +45 98154008. E-mail address: [email protected] (D. Farina).
1050-6411/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.jelekin.2006.08.005
the intracellular action potential becomes progressively longer in duration and smaller in amplitude (Hanson and Persson, 1971). Moreover, in submaximal contractions, motor units are additionally recruited to compensate for the decrease in twitch force (Fallentin et al., 1993; Garland et al., 1994), discharge rates usually decrease (De Luca et al., 1996), and de-recruitment may occur (Westgaard and De Luca, 2001). Surface electromyography (EMG) reflects these changes and usually shows an increase in
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
amplitude over time due to the increased number of active motor units and duration of the extracellular action potentials (Merletti et al., 1990). As compared to classic single-channel bipolar recording, high-density, two-dimensional surface EMG (Scholle et al., 1994; Zwarts and Stegeman, 2003) provides information on the spatial distribution of electric potential over the skin surface during muscle contraction. The technique consists in the placement of a number of closely located electrodes over the same muscle and allows a topographical representation of the muscle electrical activity over the skin plane (Holtermann et al., 2005; Lapatki et al., 2004; Zwarts and Stegeman, 2003). From surface EMG maps, it was shown that the spatial distribution of EMG amplitude is inhomogeneous (Holtermann et al., 2005) which underlines heterogeneity either in the distribution of the motor units within the muscle or in the strategy with which motor units are recruited (Holtermann et al., 2005). This may reflect muscle compartmentalization which seems particularly convenient in muscles with highly diversified biomechanical functions, e.g., upper trapezius (Johnson et al., 1994). The topographical EMG maps change with force level (Holtermann et al., 2005; Holtermann and Roeleveld, 2006) indicating that motor unit recruitment is not spatially uniform in the muscle (Hermans and Spaepen, 1997). Moreover, in the upper trapezius muscle, the skin spatial distribution of muscle activity varied over time with sustained contraction (Holtermann and Roeleveld, 2006), consistently with the expected changes due to orderly motor unit recruitment (Henneman, 1957). Changes in the spatial distribution of EMG amplitude maps were also observed in response to nociceptive afferent input (Madeleine et al., in press). Modifications in the surface EMG maps during sustained contraction indicate relative modifications in the intensity of activity within muscle regions. This may be due to dependence of fiber membrane properties on fiber location into the muscle, inhomogeneous motor unit recruitment or substitution. The control strategy has a major effect in the modifications of the spatial distribution of muscle activity during sustained contractions since similar changes in EMG maps were observed with increasing force and during fatigue development (Holtermann and Roeleveld, 2006). Variability in EMG activity among subdivisions of the erector spinae muscle resulted in longer endurance (van Dieen et al., 1993), highlighting the role of modifications in the spatial distribution of EMG activity in the maintenance of the force output. We hypothesized that sustaining a given force level for a longer period should correspond to a larger change in spatial distribution of trapezius muscle activity, which reflects sharing of the load among regions of the same muscle. Thus, in this study we tested the hypothesis that the degree of change in the spatial distribution of upper trapezius muscle activity over time is correlated to endurance time.
17
2. Methods 2.1. Subjects Eleven healthy, right-handed sedentary male subjects (age, mean ± SD, 28.2 ± 5.7 years, body mass 71.9 ± 6.6 kg, stature 1.79 ± 0.06 m, body mass index 22.6 ± 2.2) participated to the study. The study was conducted in accordance with the declaration of Helsinki, approved by the Local Ethics Committee (approval number VN 2004/56), and written informed consent was obtained from all participants prior to inclusion. 2.2. Surface EMG recording Surface EMG signals were detected from the right upper trapezius with a semi-disposable adhesive grid of 64 electrodes (LISiN-OT Bioelectronica, Torino, Italy, model ELSCH064). The grid is made of 13 rows and 5 columns of electrodes (1-mm diameter, 8-mm interelectrode distance in both directions) with one missing electrode at the upper right corner (Fig. 1). The position corresponding to the missing electrode was used as the origin of the coordinate system to define electrode location. The grid was connected to the amplifier through 4 connectors which were fixed at the subject skin by adhesive tape. The bipolar EMG signals were amplified (64-channel surface EMG amplifier, SEA 64, LISiN-OT Bioelectronica, Torino, Italy; 3 dB bandwidth 10– 500 Hz) by a factor 5000, sampled at 4096 Hz per channel, and converted to digital form by a 12-bit A/D converter. Before placement of the grid, the main innervation zone location of the upper trapezius was identified with an array of 8 electrodes (silver bars, 5-mm long, 1-mm diameter, 5-mm interelectrode distance), located along the C7-acromion line, as previously described (Farina et al., 2002). The distance from acromion to C7 was (mean ± SD) 22.4 ± 0.7 cm and the distance from C7 to the main innervation zone was 11.1 ± 0.5 cm. The electrode grid was placed with the 4th row along the C7-acromion line and with the most medial electrode column 10-mm distant from the innervation zone location (Fig. 1). The part of the skin where the grid was located was slightly abraded with abrasive paste (Medic-Every, Parma, Italy). 30 lL of conductive gel were inserted into the cavities of the grid to assure proper electrodeskin contact. A reference electrode was placed at the right wrist. 2.3. General procedures The subject sat comfortably on a chair with feet parallel on the floor at a distance approximately equal to the distance between the left and right acromion. After placement of the EMG electrode grid (see below), the subject rested for 5-min. He was then asked to abduct both arms at 90 with elbows fully extended and forearms 90 pronated, with palm facing toward the ground and without hand load. Two flexible bars placed at shoulder level were used to provide tactile position feedback to the subject. Moreover, two templates placed behind and on the side of the subject were used by the experimenter to ensure the same position of the neck and head during the contraction. The static contraction was maintained until endurance, i.e., until it was not possible for the subject to maintain the arms 90 abducted in touch with the flexible bars after strong verbal encouragement. According to previous data, the shoulder moment generated by the active muscles was 14.5 ± 1.0 Nm, corresponding to 15–20% of the maximum voluntary contraction of the trapezius muscle
18
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
Fig. 1. (a) The EMG electrode grid placed on the right upper trapezius of a subject. (b) Schematic representation of the electrode grid with the indication of the coordinate axes and origin for defining electrode position.
(Mathiassen et al., 1995). Skin temperature was monitored (Ellab, Copenhagen, Denmark) from two locations over the upper trapezius, proximal and distal with respect to the electrode grid along the seventh cervical vertebra (C7)-acromion line. Room temperature was kept stable at 24 C (±1 C). 2.4. Data processing EMG signals were off-line band-pass filtered (second order Butterworth filter; 3 dB bandwidth, 10–400 Hz). Fifty-one bipolar EMG signals were obtained from the grid (13 · 4 bipolar recordings with one missing electrode). The bipolar systems were parallel to fiber direction according to Jensen and Westgaard (1997) (Fig. 1). Mean power spectral frequency and root mean square value were computed from each bipolar recording from adjacent, non-overlapping signal epochs of 1-s duration, as described previously (Merletti et al., 1990; Welch, 1967). For graphical representation, the 51 values were interpolated by a factor 8 but only the original values were used for data processing and statistics. To characterize the spatial distribution of muscle activity, the following variables were extracted from the 51 bipolar signals: root mean square and mean power spectral frequency averaged over the 51 signals (RMSmean, MNFmean), root mean square and mean frequency corresponding to the bipolar signal with maximum root mean square in the first second of the contraction (RMSmax, MNFmax), the two coordinates of the center of gravity of the root mean square map (Gx and Gy for the lateral–medial and cranial–caudal direction, respectively), and the modified entropy of the 51 root mean square values. RMSmax and MNFmax corresponded to a fixed electrode location over the contraction. Modified entropy was computed as follows:
Entr ¼
51 X i¼1
p2 ðiÞlog2 p2 ðiÞ
where p2(i) is the square of the root mean square value at electrode i normalized by the summation of the squares of the 51 root mean square values. Entropy is a measure of uniformity of values (Proakis, 2001), used for source coding in communication engineering. The entropy of a set of M values is maximum if all values are the same (uniform distribution) and takes the value log2 M while its minimum value is zero. In the context of spatial muscle activity distribution, entropy indicates the degree of homogeneity in activation, with higher values corresponding to more uniform distribution of the root mean square values over the grid. Five consecutive values over time of RMSmean, RMSmax, MNFmean, MNFmax, Gx, Gy, and entropy were averaged to obtain mean values corresponding to 0%, 25%, 50%, 75%, and 100% of the endurance time. Percent changes of RMSmean, RMSmax, MNFmean, and MNFmax over the endurance time were defined as the difference between initial (0% endurance time) and final (100% endurance time) values normalized by the corresponding initial values and expressed in percentage. The shift of the center of gravity was defined as the modulus of the center of gravity displacement in medial–lateral and cranial–caudal direction at the beginning (0% endurance time) and end (100% endurance time) of the contraction. 2.5. Statistical analysis Linear regression analysis was used to investigate the relation between the shift of the center of gravity and endurance time, rate of change of root mean square and mean frequency. One-way repeated measures analysis of variance (ANOVA) was used to assess dependency of the EMG variables on time (0–100% endurance time, 25% increments) and was followed by post–hoc Student–Newman–Keuls (SNK) pair-wise comparisons, when appropriate. Student t-test was used to compare skin temperature at the beginning and end of the contraction. Significance was accepted for P-values less than 0.05. Results are reported as mean and standard deviation (SD) in the text and standard error (SE) in the figures.
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
3. Results Average endurance time was 382.6 ± 163.7 s. Skin temperature did not change significantly from the beginning to the end of the contraction (from 32.7 ± 1.5 C to 33.2 ± 1.1 C) and with respect to location (33.2 ± 1.7 C and 33.7 ± 1.1 C for proximal and distal location, respectively). The root mean square maps showed two main areas of high activity and a line of low activity, corresponding to the location of the innervation zones (Fig. 2), as confirmed by the inversion of the action potential polarity (visual
19
inspection of the signals). Mean power spectral frequency maps showed high values in correspondence to the innervation zone and lower values in the high-amplitude region, in agreement with previous results (Farina et al., 2002; Kleine et al., 2000) (Fig. 2). 3.1. Trends over time of EMG variables Average RMSmax across subjects did not significantly change over time due to a large variability of behaviors across subjects (percent increase at the end with respect to
Fig. 2. Topographical map (interpolation by a factor 8) of root mean square (a) and mean power spectral frequency (b) at 50% endurance time in one subject. Innervation zone location corresponds to low amplitude and high frequency values.
Fig. 3. Mean (±SE) root mean square (RMS) averaged over the grid (a), mean power spectral frequency (MNF) averaged over the grid (b), x- (empty circles) and y- (filled circles) coordinate of the center of gravity (c), and entropy (d), for 0%, 25%, 50%, 75%, and 100% of the endurance time.
20
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
Fig. 4. Root mean square topographical maps at 0%, 25%, 50%, 75%, and 100% endurance time in one subject. The black circle in the maps indicates the center of gravity. The grey scale is the same for the five maps.
the beginning, 7.5 ± 41.2%). For seven subjects RMSmax increased while the opposite was observed for the other subjects. On the contrary, RMSmean increased over time (F = 9.9, P < 0.0001) and was larger at 100% contraction time than at all other contraction times (P < 0.05), and at 0% smaller than at 50%, 75%, and 100% (P < 0.05) (Fig. 3a). Average MNFmax across subjects did not depend on time which was due, as for RMSmax, to the large variability across subjects (percent decrease, 2.3 ± 15.5%). MNFmean decreased over time (F = 10.5, P < 0.0001), and was larger at 0% contraction time than at the other contraction times (P < 0.001) (Fig. 3b).
Center of gravity shift was also positively correlated with the percent change in MNFmax (R2 = 0.44, P < 0.05), with subjects showing large shift increasing MNFmax, as opposite to subjects with smaller changes (Fig. 5c). Center of gravity shift was not correlated with MNFmean (R2 = 0.03) while it was negatively correlated to both initial and final (contraction times 0% and 100% of the endurance time) entropy (R2 = 0.54 and R2 = 0.56, respectively; P < 0.01 in both cases) (Fig. 5d), indicating that subjects with less uniform root mean square maps had larger shifts of the center of gravity. 4. Discussion
3.2. Changes over time of the root mean square map Fig. 4 shows an example of root mean square maps at the five time instants considered. The y-coordinate of the root mean square map decreased (i.e., cranial shift of the center of gravity) over time (F = 25.9, P < 0.0001), with all time instants different between each other (P < 0.01), except for 50/75% and 75/100% (Fig. 3c). The x-coordinate increased (i.e., lateral shift of the center of gravity) over time (F = 18.6, P < 0.0001), with differences between all time instants except 50/75% and 75/100% (Fig. 3c). However, the absolute shift of the x-coordinate (1.0 ± 0.6 mm over the endurance time) was negligible with respect to the shift in the y-coordinate (11.2 ± 6.1 mm). Entropy of the root mean square map significantly decreased over time (less uniform map) (F = 8.5, P < 0.0001), with 25% contraction time different from 75% (P < 0.05) and 100% smaller than all others (P < 0.05) (Fig. 3d). 3.3. Correlations between changes in EMG variables and endurance time Subjects whose activity map shifted more (displacement of the center of gravity) had longer endurance times (R2 = 0.46, P < 0.05) (Fig. 5a). Subjects with large shift showed a decreased RMSmax at the endurance with respect to the beginning of the contraction while the opposite was observed in subjects with small shift (R2 = 0.39, P < 0.05) (Fig. 5b). Percent change in RMSmean was on the contrary positively correlated with the shift in the center of gravity (R2 = 0.51, P < 0.05) and was larger at the endurance time with respect to the beginning of the contraction in all subjects.
This study shows for the first time that the amount of shift in the spatial distribution of EMG activity of the upper trapezius muscle is related to the time during which a static contraction is maintained. This indicates that changes in spatial muscle activity distribution, estimated by changes in the center of gravity of the root mean square map and entropy, play a role in the ability to maintain a static contraction. 4.1. Spatial changes in topographical EMG maps The topographical map of EMG activity changed over time. This has been quantified by the position of the center of gravity and by the entropy of root mean square values. The changes included a cranial and lateral shift of the center of gravity and a decrease in entropy, reflecting less uniform map. The observed cranial shift of muscle activity is in agreement with previous work (Kleine et al., 2000; Madeleine et al., in press). The observed heterogeneity in muscle activation supports the hypothesis of functional subdivision of the upper trapezius muscle during low level activation (Jensen and Westgaard, 1997; Mathiassen et al., 1995). The temporal changes in spatial activity distribution may have been due to both peripheral and central mechanisms. Peripheral mechanisms include the change in muscle fiber membrane properties with fatigue that has an effect on EMG amplitude (Merletti et al., 1990). Muscle fiber conduction velocity decreases with sustained contraction due to the increase in interstitial potassium concentration with repetitive fiber activation (Kossler et al., 1991). The decrease in conduction velocity corresponds to increased
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
21
Fig. 5. Scatter plots of the shift of the center of gravity vs endurance time (a), percent change in maximum root mean square (RMSmax, see text for its definition) (b), percent change in maximum mean power spectral frequency (MNFmax, see text for its definition) (c), and entropy at the beginning (0% endurance time, filled circles) and end (100% endurance time, empty circles) of the contraction (d).
duration of the action potentials (Lindstrom and Magnusson, 1977) and thus to increased root mean square in case of unchanged intracellular potential (Merletti et al., 1990). On the other hand, the increase in potential duration also increases amplitude cancellation (Keenan et al., 2005), although this phenomenon is more limited than the direct effect of conduction velocity on action potential area (Keenan et al., 2005). Finally, the intracellular action potential shape may change with sustained contraction (Hanson and Persson, 1971) which affects EMG amplitude (Dimitrova and Dimitrov, 2003). The change in root mean square distribution may be due to an inhomogeneous modification over time in any of the variables related to fiber membrane properties. Inhomogeneous changes in membrane properties may be due to a different distribution of fiber types in different regions of the upper trapezius (Lindman et al., 1990) or to a distribution of intramuscular pressure which may affect the rate of metabolic wash-out during sustained contraction. However, a pure peripheral effect in the modification of EMG maps is unlikely since in four subjects there was a reduction in root mean square in the location where at the beginning it was at its maximum (Fig. 5b). The decrease in conduction velocity is probably the main determinant, among the peripheral effects, of the changes in EMG amplitude (Merletti et al., 1990) and mainly determines an increase in amplitude. Moreover, for subjects who showed large changes in the location of the center of gravity, mean power spectral frequency at the location with maximal initial root mean square increased over time
(Fig. 5c). This cannot be explained by decreased conduction velocity without assuming a change in motor control strategies. Indeed, without any change in the number of active motor units and in their discharge rates, mean frequency and conduction velocity show similar trends over time (Lindstrom and Magnusson, 1977). Thus, spatial changes in root mean square maps cannot solely be explained by changes at the muscle level. Motor unit control strategies change during sustained contraction. In particular, motor unit discharge rate usually decreases with fatigue (Bigland-Ritchie et al., 1983; De Luca et al., 1996), probably due to reflex inhibition from small diameter muscle afferents (Bigland-Ritchie et al., 1986; Gandevia, 2001; Woods et al., 1987). EMG amplitude is directly related to motor unit discharge rate, thus space-dependent decrease in discharge rate may explain variations in the root mean square map. There is also evidence of additional recruitment of motor units during sustained static contractions (Fallentin et al., 1993; Garland et al., 1994) and some reports indicate motor unit derecruitment and substitution in order to cope with fatigue development (Westad et al., 2003; Westgaard and De Luca, 2001). The way in which motor units are recruited during a static contraction was related to the endurance time, as observed comparing similar tasks with different loads (Maluf and Enoka, 2005). Recruitment/derecruitment of motor units likely change the spatial activity distribution since distribution of motor unit types and diameters in the upper trapezius is not homogeneous (Lindman et al., 1990). Moreover, the changes in root mean square maps
22
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
over time in static contractions resemble the changes which occur with increasing force in non-fatiguing contractions (Holtermann and Roeleveld, 2006), indicating an orderly recruitment with sustained contraction. Degree of motor unit synchronization may also vary with fatigue although the effect of synchronization on surface EMG amplitude may be limited (Zhou and Rymer, 2004). 4.2. Time to task failure and EMG maps The present study showed a correlation between endurance time and the amount of shift in the center of gravity of root mean square maps. Subjects with larger shift had longer endurance time, in agreement with results reported for different subdivisions of the erector spinae muscle using multiple bipolar recordings (van Dieen et al., 1993). This indicates a functional significance of modifications in muscle activity spatial distribution. The range of variation for the shift of the center of gravity varied from approximately 2 to 17 mm (Fig. 5). The subjects who showed large shift were those with the most heterogeneous root mean square map (minimum entropy) (Fig. 5d). The differences among subjects in the amount of shift in the center of gravity may be related to a different strategy to cope with fatigue, in particular in relation to motor unit recruitment/derecruitment. During the contraction, for some subjects and locations over the muscle, the EMG activity decreased over time (Fig. 5b), as opposed to an average increase in root mean square (Fig. 3a), which indicates localized decreased excitation level to the muscle. This may be due to either derecruitment of motor units or spatially dependent decrease in discharge rate. In both cases, the results indicate a space-dependant modification in the neural drive to the muscle which allowed different parts of the muscle to be active at different stages of the fatigue development in order to maintain the same force output. One possible interpretation of these results is that the reorganization of muscle activity with sustained contraction had the functional significance of reducing fatigue development. This is in agreement with the hypothesis that spatial reorganization of muscle activity reduces muscle fiber overload and leads to more uniform changes in the extra-cellular environment in different muscle regions. For example, motor strategies with increased degree of freedom (Madeleine et al., 2003; Mathiassen et al., 2003) are less susceptible to lead to work related musculoskeletal disorders. However, the correlation between endurance time and changes in spatial EMG distribution may also be due to different activation levels of the muscle at the beginning of the contraction. The relative load sustained with respect to maximal force may have indeed been different among subjects. Thus, subjects exerting more relative force may have sustained the contraction more than subjects exerting lower force. Since maximal force was not measured in this study, this interpretation cannot be completely excluded. However, it is not in agreement with the results in van
Dieen et al. (1993) and with the observed higher uniformity in the maps of the subjects who maintained the task for shorter time. If these subjects were also those for whom the relative force level was higher, a larger heterogeneity, corresponding to higher force would have been expected. The degree of heterogeneity increased with fatigue and since similar changes in surface EMG maps are observed with increasing force and with fatigue (Holtermann and Roeleveld, 2006), higher force should correspond to higher degree of heterogeneity. From the concomitant observation of a correlation between shift of the maps and entropy with endurance time, the explanation based purely on differences in initial relative forces seems very unlikely. 4.3. Interpretation of surface EMG variables The present results underline difficulties in the interpretation of EMG variables estimated from signals recorded from a single muscle site. This is due to the large variability among subjects in spatial changes over time. Usually, trends over time of mean power spectral frequency and root mean square are used as indexes of muscle fatigue. This assumption is based on the relation between muscle fiber conduction velocity, mean power frequency, and EMG amplitude (Lindstrom and Magnusson, 1977; Merletti et al., 1990). However, when the number of motor units or distribution of discharge rates is not the same over time, these relations are not valid. Depending on the amount of change in the position of the center of gravity, RMSmax and MNFmax decreased or increased (Fig. 5b and c). Subjects with larger shift of the center of gravity showed decreased RMSmax and increased MNFmax, contrary to the average trend of root mean square and mean frequency over the grid (Fig. 3a and b), while the opposite trends were observed for subjects with smaller shift. This was probably related to a larger variation in the number of active motor units over time for subjects with larger changes in the spatial distribution of root mean square. Variations in the number of active motor units may indeed compensate for the decrease in mean frequency (Farina et al., 2006). Variability in the spatial changes of muscle activity among subjects may thus explain the controversial results reported in the literature on trends over time of EMG variables during static contractions of the upper trapezius (compare, e.g., Farina et al., 2002; Madeleine et al., ¨ berg et al., 1992; Sadoyama 2002; Mathiassen et al., 1995; O et al., 2000). This problem is more important at low contraction levels when the number of active motor units at the beginning of the task does not correspond to full recruitment. Topographical mapping allows us to take this variability into consideration. 4.4. Limitations The lateral shift (x-coordinate) of the center of gravity of root mean square maps was not expected in case the grid was placed exactly in line with the fiber orientation and
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
between the innervation zone and tendon. Indeed, in these conditions the x-direction of the grid would correspond to the direction of propagation of the action potentials and the distribution of amplitude should not be altered (if the fiber is activated, the action potential always propagates along the entire fiber semilength, from innervation zone to tendon). The shift of the center of gravity in the medial–lateral direction, which was anyway negligible compared with the cranial–caudal shift, was probably due to slight misalignment of the grid transversal side with respect to the fiber orientation or to changes in location of the innervation zone over time. This does not affect the modulus of the center of gravity displacement and thus the conclusions drawn. The position of the center of gravity and the entropy were computed from the entire maps, although some parts of the grid covered the innervation zones of the muscle. Since the action potentials have lower amplitude in correspondence to innervation zone, spatial heterogeneity may have been simply resulted from anatomical factors. Absolute values of descriptors of heterogeneity (e.g., entropy) are affected by this problem. However, the relative changes in the variables investigated, on which most of the conclusions are based, were not influenced by this issue. The type of analysis performed is global and aims at assessing spatial modifications. Even in case the maps changed due to recruitment of motor units with different locations of the innervation zones over time (thus affecting the map for changed anatomical properties), the global variables used would have been valuable in underlining strategies in motor unit recruitment. The selected task is not specific for the upper trapezius, thus endurance may have been limited by other muscles rather the investigated one, for example the deltoid. However, even if this was the case, the observed correlations between spatial modifications and duration of the contraction indicate that the larger the modifications in upper trapezius spatial activity distribution, the longer the contraction was sustained, which is the main conclusion of the study. 4.5. Conclusion The study shows correlation between change in the spatial distribution of muscle activity over time and endurance time. Subjects with more heterogeneous activity and larger shift in the upper trapezius activity toward the cranial direction could sustain the static contraction longer than subjects with more uniform activity distribution and smaller changes in the EMG map over time. A likely explanation of these results is that heterogeneity in muscle activity and spatial adaptation over time have the functional role of prolonging sustenance of a static task. 5. Grants The Marie Curie fellowship of the European Commission program ‘‘Improving the human research potential
23
and the socio-economic knowledge’’ based under the contract number HPMT-CT-2000-00092 supported Frederic Leclerc. The Danish Technical Research Council (Program ‘‘Centre for Neuroengineering (CEN)’’, contract number 26-04-0100) partly supported the study. Acknowledgements The system for two-dimensional EMG detection (amplifier and sensor) has been developed at the Laboratory for Neuromuscular System Engineering (LISiN), Torino, Italy. The authors are sincerely grateful to Roberto Merletti and Andrea Bottin (LISiN) for the support in the use of the system. References Bigland-Ritchie B, Donovan EF, Roussos CS. Conduction velocity and EMG power spectrum changes in fatigue of sustained maximal efforts. J Appl Physiol 1981;51:1300–5. Bigland-Ritchie B, Johansson R, Lippold OC, Smith S, Woods JJ. Changes in motoneurone firing rates during sustained maximal voluntary contractions. J Physiol 1983;340:335–46. Bigland-Ritchie BR, Dawson NJ, Johansson RS, Lippold OC. Reflex origin for the slowing of motoneurone firing rates in fatigue of human voluntary contractions. J Physiol 1986;379:451–9. De Luca CJ, Foley PJ, Erim Z. Motor unit control properties in constantforce isometric contractions. J Neurophysiol 1996;76:1503–16. Dimitrova NA, Dimitrov GV. Interpretation of EMG changes with fatigue: facts, pitfalls, and fallacies. J Electromyogr Kinesiol 2003;13: 13–36. Fallentin N, Jorgensen K, Simonsen EB. Motor unit recruitment during prolonged isometric contractions. Eur J Appl Physiol Occup Physiol 1993;67:335–41. Farina D, Madeleine P, Graven-Nielsen T, Merletti R, Arendt-Nielsen L. Standardising surface electromyogram recordings for assessment of activity and fatigue in the human upper trapezius muscle. Eur J Appl Physiol 2002;86:469–78. Farina D, Zennaro D, Pozzo M, Merletti R, Laubli T. Single motor unit and spectral surface EMG analysis during low-force, sustained contractions of the upper trapezius muscle. Eur J Appl Physiol 2006;96:157–64. Gandevia SC. Spinal and supraspinal factors in human muscle fatigue. Physiol Rev 2001;81:1725–89. Garland SJ, Enoka RM, Serrano LP, Robinson GA. Behavior of motor units in human biceps brachii during a submaximal fatiguing contraction. J Appl Physiol 1994;76:2411–9. Hanson J, Persson A. Changes in the action potential and contraction of isolated frog muscle after repetitive stimulation. Acta Physiol Scand 1971;81:340–8. Henneman E. Relation between size of neurons and their susceptibility to discharge. Science 1957;126:1345–7. Hermans V, Spaepen AJ. Influence of electrode position on changes in electromyograph parameters of the upper trapezius muscle during submaximal sustained contractions. Eur J App Physiol 1997;75: 319–25. Holtermann A, Roeleveld K, Karlsson JS. Inhomogeneities in muscle activation reveal motor unit recruitment. J Electromyogr Kinesiol 2005;15:131–7. Holtermann A, Roeleveld K. Global muscle activation in sustained contractions. Acta Physiol (Oxf) 2006;186:159–68. Jensen C, Westgaard RH. Functional subdivision of the upper trapezius muscle during low-level activation. Eur J Appl Physiol 1997;76: 335–9.
24
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25
Johnson G, Bogduk N, Nowitzke A, House D. Anatomy and actions of the trapezius muscle. Clin Biomech 1994;9:44–50. Keenan KG, Farina D, Maluf KS, Merletti R, Enoka RM. Influence of amplitude cancellation on the simulated surface electromyogram. J Appl Physiol 2005;98:120–31. Kleine BU, Schumann NP, Stegeman DF, Scholle HC. Surface EMG mapping of the human trapezius muscle: the topography of monopolar and bipolar surface EMG amplitude and spectrum parameters at varied forces and in fatigue. Clin Neurophysiol 2000;111:686–93. Kossler F, Lange F, Caffier G, Kuchler G. External potassium and action potential propagation in rat fast and slow twitch muscles. Gen Physiol Biophys 1991;10:485–98. Lapatki BG, Van Dijk JP, Jonas IE, Zwarts MJ, Stegeman DF. A thin, flexible multielectrode grid for high-density surface EMG. J Appl Physiol 2004;96:327–36. Lindman R, Eriksson A, Thornell LE. Fiber type composition of the human male trapezius muscle: enzyme-histochemical characteristics. Am J Anat 1990;189:236–44. Lindstrom L, Magnusson R. Interpretation of myoelectric power spectra: a model and its applications. Proc IEEE 1977;65:653–62. Madeleine P, Farina D, Merletti R, Arendt-Nielsen L. Upper trapezius muscle mechanomyographic and electromyographic activity in humans during low force fatiguing and non-fatiguing contractions. Eur J Appl Physiol 2002;87:327–36. Madeleine P, Lundager B, Voigt M, Arendt-Nielsen L. The effects of neckshoulder pain development on sensory-motor interactions among female workers in the poultry and fish industries. A prospective study. Int Arch Occup Environ Health 2003;76:39–49. Madeleine P, Leclerc F, Arendt-Nielsen L, Ravier P, Farina D. Experimental muscle pain changes the spatial distribution of upper trapezius muscle activity during sustained contraction. Clin Neurophysiol [in press]. Maluf KS, Enoka RM. Task failure during fatiguing contractions performed by humans. J Appl Physiol 2005;99:389–96. Mathiassen SE, Winkel J, Ha¨gg GM. Normalisation of surface EMG ampltude from the upper trapezius muscle in ergonomic studies – A review. J Electromyograph Kinesiol 1995;5:197–226. Mathiassen SE, Moller T, Forsman M. Variability in mechanical exposure within and between individuals performing a highly constrained industrial work task. Ergonomics 2003;46:800–24. Merletti R, Knaflitz M, De Luca CJ. Myoelectric manifestations of fatigue in voluntary and electrically elicited contractions. J Appl Physiol 1990;69:1810–20. ¨ berg T, Sandsjo L, Kadefors R, Larsson SE. Electromyographic changes O in work-related myalgia of the trapezius muscle. Eur J-Appl Physiol 1992;65:251–7. Proakis J. Digital communications. New York: McGraw Hill; 2001. Sadoyama T, Sakai K, Hosoya S , Kamijyo M, Masuda T. Changes in surface EMG during static fatiguing contractions in the upper trapezius muscle. In: Proceedings of the 13th ISEK congress, Sapporo; 2000. Scholle HC, Schumann NP, Anders C. Quantitative-topographic and temporal characterization of myoelectrical activation patterns: new diagnostic possibilities in neurology, physiotherapy and orthopaedics. Funct Neurol 1994;9:35–45. van Dieen JH, Oude Vrielink HH, Toussaint HM. An investigation into the relevance of the pattern of temporal activation with respect to erector spinae muscle endurance. Eur J Appl Physiol Occup Physiol 1993;66:70–5. Welch PD. Use of fast Fourier transform for estimation of power spectra A method based on time averaging over short modified periodograms. IEEE Trans Audio Electroacoust 1967;15:70–3. Westad C, Westgaard RH, De Luca CJ. Motor unit recruitment and derecruitment induced by brief increase in contraction amplitude of the human trapezius muscle. J Physiol 2003;552(Pt 2):645–56. Westgaard RH, De Luca CJ. Motor control of low-threshold motor units in the human trapezius muscle. J Neurophysiol 2001;85:1777–81. Woods JJ, Furbush F, Bigland-Ritchie B. Evidence for a fatigue-induced reflex inhibition of motoneuron firing rates. J Neurophysiol 1987;58: 125–37.
Zhou P, Rymer WZ. Factors governing the form of the relation between muscle force and the EMG: a simulation study. J Neurophysiol 2004;92:2878–86. Zwarts MJ, Stegeman DF. Multichannel surface EMG: basic aspects and clinical utility. Muscle Nerve 2003;28:1–17.
Dario Farina graduated summa cum laude in Electronics Engineering from Politecnico di Torino, Torino, Italy, in February 1998. During 1998 he was a Fellow of the Laboratory for Neuromuscular System Engineering in Torino. In 2001 and 2002 he obtained the Ph.D. degrees in Automatic Control and Computer Science and in Electronics and Communications Engineering from the Ecole Centrale de Nantes, Nantes, France, and Politecnico di Torino, respectively. In 1999–2004 he taught courses in Electronics and Mathematics at Politecnico di Torino and in 2002–2004 he was Research Assistant Professor at the same University. Since 2004, he is Associate Professor in Biomedical Engineering at the Faculty of Engineering and Science, Department of Health Science and Technology of Aalborg University, Aalborg, Denmark, where he teaches courses on biomedical signal processing, modeling, and neuromuscular physiology. He regularly acts as referee for approximately 20 scientific International Journals, is an Associate Editor of IEEE Transactions on Biomedical Engineering, is on the Editorial Boards of the Journal of Neuroscience Methods, the Journal of Electromyography and Kinesiology, and Medical & Biological Engineering & Computing, and member of the Council ISEK (International Society of Electrophysiology and Kinesiology). His main research interests are in the areas of signal processing applied to biomedical signals, modeling of biological systems, basic and applied physiology of the neuromuscular system, and brain-computer interfaces. Dr. Farina is a Registered Professional Engineer in Italy.
Fre´de´ric Leclerc was born in 1980. He graduated in Electronics and Signal-Processing Engineering from Polytech’Orleans in September 2004. The same year he obtained the Master of Science Degree in Signal-Images in Biology and Medicine from Angers University where he developed an Holter system. Currently he is a Ph.D. candidate in the Laboratory of Electronic Signals Images of Orleans, France, working on a project on the development of methods for information extraction and interpretation from electrophysiological measures. His main interests are related to the development of acquisition systems, software and processing techniques for electromyographic and electrocardiographic signals.
Lars Arendt-Nielsen, born in 1958, received the M.Sc.E.E. degree from Aalborg University, Denmark, in 1983, with specialisation in biomedical engineering, and the Ph.D. degree in 1992. In 1994 he received his Dr. Sci. degree in Medicine from the Medical Faculty, Aarhus University, Denmark. From 1983 to 1984 he was a Research fellow, Department of Clinical Neurophysiology, The National Hospital for Nervous Diseases, London. Since 1988 he has been with the Department of Medical Informatics and Image Analysis, Aalborg University as an Associated Professor. In 1993 he was appointed Professor in Biomedical Engineering and Principal investigator at Center for Sensory-Motor Interaction, which was established in 1993 at Aalborg University and in 1997 Head of the International Doctoral
D. Farina et al. / Journal of Electromyography and Kinesiology 18 (2008) 16–25 School in Biomedical Science and Engineering, Aalborg University, with 55 Ph.D. students enrolled. During his career he has worked as guest professor in Japan and Australia. He is member of the Danish Research Council and the Danish Research Education Council. He has published approx. Five hundred and ten scientific papers within neuroscience with focus on motor control and pain research and given more than 110 keynote lectures at international conferences. Olivier Buttelli was born in 1966. He received the MS and Ph.D. degrees in biomechanics and motor physiology in 1991 and 1996, respectively, both from the University of Paris 11 Orsay. Since 1997, he is Assistant Professor in biomechanics and exercise physiology at the sport exercise faculty of Orleans University. His researches focus on fatigue and exercise.
25
Pascal Madeleine was born in Toulouse, France, in 1969. He received the M.Sc. degree in biomedical engineering with an electronic specialisation in 1991 from Paul Sabatier University, Toulouse, France, the Ph.D. qualifying degree in biomedical engineering with a biosignal processing specialisation in 1993 from Paris XII-Val de Marne University, Creteil, France and the Ph.D. degree in 1998 from Aalborg University, Denmark. He is currently employed as an associate professor at the Center for Sensory-Motor Interaction (SMI), Department of Health Science and Technology at Aalborg University, Denmark. His main areas of research interests are the interactions between muscle pain and motor control in relation to work related musculo-skeletal disorders via mechanomyographic, electromyographic, kinetic and kinematic data signal processing.