The comparison of wavelet- and Fourier-based electromyographic indices of back muscle fatigue during dynamic contractions: validity and reliability results.

The purpose of this study was to compare the electromyographic (EMG) fatigue indices computed from short-time Fourier transform (STFT) and wavelet transform (WAV), by analyzing their criterion validity and test-retest reliability. The effect of averaging spectral estimates within and between repeated contractions (cycles) on EMG fatigue indices was also demonstrated. Thirty-one healthy subjects performed trunk flexion-extension cycles until exhaustion on a Biodex dynamometer. The load was determined theoretically as twice the L5-S1 moment produced by the trunk mass. To assess reliability, 10 subjects performed the same experimental protocol after a two-week interval. EMG signals were recorded bilaterally with 12 pairs of electrodes placed on the back muscles (at L4, L3, L1 and T10 levels), as well as on the gluteus maximus and biceps femoris. The endurance time and perceived muscle fatigue (Borg CR-10 scale) were used as fatigue criteria. EMG signals were processed using STFT and WAV to extract global (e.g, median frequency and instantaneous median frequency, respectively) or local (e.g., intensity contained in 8 frequency bands) information from the power spectrum. The slope values of these variables over time, obtained from regression analyses, were retained as EMG fatigue indices. EMG fatigue indices (STFT vs. WAV) were not significantly different within each muscle, had a variable association (Pearson's r range.: 0.06 to 0.68) with our fatigue criteria, and showed comparable reliability (Intra-class correlation range: 0.00 to 0.88), although they varied between muscles. The effect of averaging, within and between cycles, contributed to the strong association between EMG fatigue indices computed from STFT and WAV. As for EMG spectral indices of muscle fatigue, the conclusion is that both transforms carry essentially the same information.

Evaluation of measurement strategies to increase the reliability of EMG indices to assess back muscle fatigue and recovery.

The purpose of this study was to assess different measurement strategies to increase the reliability of different electromyographic (EMG) indices developed for the assessment of back muscle impairments. Forty male volunteers (20 controls and 20 chronic low back pain patients) were assessed on three sessions at least 2 days apart within 2 weeks. Surface EMG signals were recorded from four pairs (bilaterally) of back muscles (multifidus at the L5 level, iliocostalis lumborum at L3, and longissimus at L1 and T10) while the subjects performed, in a static dynamometer, two static trunk extension tasks at 75% of the maximal voluntary contraction separated by a 60 s rest period: (1) a 30 s fatigue task and (2) a 5 s recovery task. Different EMG indices (based on individual muscles or averaged across bilateral homologous muscles or across all muscles) were computed to evaluate muscular fatigue and recovery. Intra-class correlation coefficient (ICC) and standard error of measurement (SEM) in percentage of the grand mean were calculated for each EMG variable. Reliable EMG indices are achieved for both healthy and chronic low back pain subjects when (1) electrodes are positioned on medial back muscles (multifidus at the L5 level and longissimus at L1) and (2) measures are averaged across bilateral muscles and/or across two fatigue tests performed within a session. The most reliable EMG indices were the bilateral average of medial back muscles (ICC range: 0.68-0.91; SEM range: 5-35%) and the average of all back muscles (ICC range: 0.77-0.91; SEM range: 5-30%). The averaging of measures across two fatigue tests is predicted to increase the reliability by about 13%. With regards to EMG indices of fatigue, the identification of the most fatigable muscle also lead to satisfactory results (ICC range: 0.74-0.79; SEM range: 21-26%). The assessment of back muscle impairments through EMG analysis necessitates the use of multiple electrodes to achieve reliable results.

Effect of step and ramp static contractions on the median frequency of electromyograms of back muscles in humans.

The purpose of this study was to compare the electromyogram median frequency (MF) values from two contraction modes (ramp vs step) at different force levels of eight back muscles. A group of 20 healthy male subjects stood in a dynamometer with the trunk in a vertical position and performed trunk extension contractions using the displayed L5/S1 extension moment as visual feedback. The electromyogram (EMG) signals from four pairs of back muscles were collected at 4,096 Hz using active surface electrodes during two 7 s static ramp contractions ranging from 0% to 100% of the maximal voluntary contraction (MVC) and two 5 s static step contractions performed at five forces (10%, 20%, 40%, 60% and 80% MVC). The root mean square (RMS) and MF of the EMG signals corresponding to 250 ms windows were computed at each force level for both contraction modes. The RMS from the ramp contractions were significantly higher than from the step contractions in six muscles. The corresponding MF showed a significant (alpha = 0.05) contraction mode x force interaction in four muscles. A significant contraction mode main effect was obtained in four muscles having higher MF during step than during ramp contractions. These differences were more obvious (10-15 Hz) and more frequent at the lower (10%, 20% and 40% MVC) forces. It was suggested that mechanisms not related to motor unit recruitment might influence MF in contraction modes. These unknown mechanisms contaminate any possible relationship between the MF measurements and muscle composition.

Median frequency of the electromyographic signal: effect of time-window location on brief step contractions.

The purpose of this study was to determine, for different back muscles, if the median frequency (MF) of the electromyographic (EMG) power spectrum changes according to the position of the time window during a 5 s step contraction. Twenty males with no known back problems were standing upright in a dynamometer allowing lower limb and pelvis stabilization. Trunk extension efforts were performed by pushing on a force platform positioned at the T4 level while the extension moment at L5/S1 was displayed as visual feedback. The EMG signals from four homologous back muscles (multifidus at L5, ilicostalis lumborum at L3, and longissimus at L1 and T10) were collected using active surface electrodes during two 5 s static step contractions performed at five force levels (10, 20, 40, 60 and 80% of the maximal voluntary contraction). The root mean square (RMS) and MF values of the EMG signals corresponding to three 250 ms time windows (beginning, middle and end of each step contraction) were computed. The RMS values of several back muscles increased from the first to the third time window for contractions performed at high force levels only. However, a concomitant decrease in the MF values was observed only for the left multifidus muscle. It was concluded that muscle fatigue does not generally manifest itself during 5 s step contractions through the EMG signal. However, it is recommended to use step contractions lasting less than 5 s and to choose a time window located in the first 1-3 s to completely eliminate the possible effects of fatigue.

Physical impairments in cervicogenic headache: traumatic vs. nontraumatic onset.

In order to quantify the physical impairments associated with different types of headache, 77 subjects belonging to four different groups (postmotor vehicle accident cervicogenic headache subjects, cervicogenic headache subjects nontraumatic, migraine patients and control subjects) were evaluated using the following variables: posture, cervical range of motion, strength of the neck flexors and extensors, endurance of the short neck flexors, manual segmental mobility, proprioception of the neck, and pain (McGill Pain Questionnaire and the skin roll test). The results of this study showed that postmotor vehicle accident cervicogenic patients have significantly limited active cervical range of motion (in flexion/extension and rotations), present decreased strength and endurance of neck flexors and decreased strength of the extensor muscles. Our results suggest that there are enough differences between the postmotor vehicle accident and nontraumatic cervicogenic headache subjects to warrant caution when analysing the data of these two subgroups together, as several studies have done in the past. The onset of headache is therefore an important variable that should be controlled for when attempting to characterize the physical impairments associated with cervicogenic headache.

Effects of upper and lower limb static exertions on global synkineses in hemiparetic subjects.

OBJECTIVES: Global synkineses are nonpurposive pathological involuntary muscle activities or movements elicited at several or all of the joints of the affected limb or limbs during voluntary forceful resisted contractions. The purpose of this study was to assess the effect of upper and lower limb exertions on manifestations of upper limb global synkineses in hemiparetic subjects. DESIGN: Involuntary muscle activities on the affected upper limb of 11 hemiparetic subjects and on the left or right upper limb of 10 control subjects were recorded using surface electromyography during successive bilateral maximal ankle exertions and during contralateral grips. RESULTS: Significant differences in the level of involuntary electromyography (EMG) activities were observed between experimental conditions (ANOVAs, p < 0.05). EMG levels in hemiparetic subjects were significantly higher during contralateral grip tasks than during the ankle exertions. CONCLUSION: These results suggest that upper limb global synkinases are more prevalent in specific tasks and that this task specificity may reflect the neurophysiological mechanisms involved in the generation of global synkinases.

Analysis of the clinical factors determining natural and maximal gait speeds in adults with a stroke.

The objective of this study was to identify the most important clinical variables determining gait speed in persons with stroke. Sixteen chronic stroke subjects (mean age, 47.9 (+/-15.6) yr; mean time post-stroke, 43.9 (+/-36.5) mo) able to walk independently without a brace participated in the study. The impairments in motor function, sensation of the paretic lower limb, and balance were evaluated with the Fugl-Meyer Assessment. A spasticity index was used to assess the muscle tone of the plantarflexors. The maximal strengths in plantarflexion and hip flexion were measured with a Biodex dynamometric system. Cinematography and foot-contact data collected on the paretic side were used to determine the comfortable and maximal gait speeds. The level of association between gait speeds and the clinical variables were first examined with Pearson's correlation coefficients and, then, with multiple linear regression analyses using the stepwise method. Results revealed that the motor function of the lower limb, balance, and hip flexion strength were significantly related to comfortable and maximal gait speeds (0.5 < r < 0.88; P < 0.05). For the comfortable gait speed, the regression analysis selected only the hip flexor strength as a significant variable (R2 = 0.69). For maximal gait speed, the variables retained were hip flexor strength, sensation at the lower limb, and plantarflexor strength (R2 = 0.85). The present results suggest that strength and sensation at the lower limb are important factors to consider in determining the gait capacity of chronic stroke subjects.

Plantarflexor weakness as a limiting factor of gait speed in stroke subjects and the compensating role of hip flexors.

OBJECTIVE: To determine, using the Muscular Utilization Ratio (MUR) method, whether plantarflexor weakness is among the factors preventing stroke subjects from walking at faster speeds. Potential compensations by the hip flexors were also examined. DESIGN: A convenience sample of 17 chronic stroke subjects in a context of a descriptive study. BACKGROUND: Gait speed is correlated with the residual strength of the muscles involved in gait in stroke subjects. However, it has not been established if this residual strength limits gait speed. METHODS: Kinetic and kinematic data for comfortable and maximal gait speeds were collected on the paretic side, and were used to determine the moments in plantarflexion (mechanical demand: MUR numerator) during the push-off phase. The maximal potential moment (MUR denominator) of the plantarflexors during gait was predicted using an equation derived from dynamometric data collected with a Biodex system. The MURs of the plantarflexors were then calculated at every 1% interval of the push-off phase. The pull-off phase of gait and the hip flexor strength were also examined. RESULTS: Ten subjects of the sample had a MUR value between 80 and 150% at maximal gait speed. These subjects produced the lowest peak torques in plantarflexion. Each of the four fastest subjects of this group had a large hip flexion moment during the pull-off phase of gait and produced high hip flexion torque values on the dynamometer. Each of the seven remaining subjects had a MUR value under 70% when they walked at maximal speed. CONCLUSIONS: Weakness of the plantarflexors should be considered as one factor limiting gait speed in 10 hemiparetic subjects. Some subjects with weak plantarflexors could walk rapidly because they compensated with the hip flexors. For the remaining stroke subjects, factors other than weakness of the plantarflexors have to be considered in order to explain the reduction in their gait speed.

A static dynamometer measuring simultaneous torques exerted at the upper limb.

The majority of available dynamometers are designed to measure force or torque in one specific direction, one joint at a time. For the quantification of motor incoordination in neurological patient populations, these dynamometers provide limited information about the global behavior of the limb under investigation. This report describes the potential use and function of a static dynamometer measuring torques exerted simultaneously at the shoulder (flexion-extension, abduction-adduction, internal-external rotation), elbow (flexion-extension), and forearm (pronation-supination). Orthogonal forces were measured at the arm and wrist using strain gauge transducers interfaced with a laboratory computer. The lever arms were specified to a software program and the joint torques were calculated in real time according to static equilibrium equations. The use of the dynamometer is illustrated by characterizing for one hemiparetic subject, the joints torques recorded at the shoulder, elbow, and forearm during isolated submaximal grip exertions at different force levels on both sides. The torques generated at the shoulder, elbow and forearm during the hand grip tasks on the affected side were significantly higher than those obtained on the nonaffected side and increased with the grip force level. These differences probably reflect the loss of movement selectivity observed following a lesion in the central nervous system. Further studies are currently being undertaken in neurological patient populations to characterize and quantify motor deficits using this dynamometer. As a long term goal, we hope that the method and technologies described here will contribute to the evaluation and rehabilitation of these populations.

Effects of transcutaneous electrical nerve stimulation on the H-reflex of muscles of different fibre type composition.

Differential effects of repetitive stimulation of low threshold afferents on both the recruitment threshold and motoneuronal excitability of type I and type II motor units have been demonstrated. The present study was aimed at further investigating the differential effects of 30 minutes of transcutaneous electrical nerve stimulation (TENS) on the H-reflex amplitude (Hmax/2) of the Soleus (SO), gastrocnemius lateralis (GL) and medialis (GM) muscles. Eleven healthy subjects were tested in order to evaluate the effects of TENS on either the common peroneal (CPN), saphenous or sural nerve. The experimental session consisted of three consecutive 45 min periods. Within each of these periods, H-reflexes were recorded before, during and after the TENS was applied. It was hypothesized that repetitive low threshold afferent stimulation would either have inhibitory or facilitatory effects on the H-reflex amplitude of the SO or gastrocnemii muscles respectively. Non-parametric Friedman ANOVAs revealed a significant tendency (p < 0.05) toward inhibition of the H-reflex amplitude of the SO and GL muscle during TENS applied over either the CPN or sural nerve, as well as that of the GM during repetitive stimulation of the saphenous nerve. Although the present study failed to reveal any differential effects of TENS on the H-reflex amplitude of muscle on different fibre type content, the significant decrease in H-reflex observed on the triceps surae muscles during TENS applied over the CPN might have promising clinical outcomes for hyperreflexive subjects.

Dynamometric assessment of the plantarflexors in hemiparetic subjects: relations between muscular, gait and clinical parameters.

The aims of this study were to investigate, in 16 subjects with hemiparesis, the plantarflexor muscle performance of the paretic side and to determine the level of the relationships between muscular parameters, clinical measures and gait performance. A Biodex dynamometric system was used to evaluate static and dynamic torques, power and maximal rate of tension development of the plantarflexor muscles. The clinical measures included the Fugl-Meyer assessment (FMA), the "Up & Go" test and an evaluation of ankle muscle tone. Velocity, cadence, stride length and gait cycle duration were determined for each subject at both comfortable and maximal safe speeds using foot contacts and videographic data. Results indicated that dynamometric values produced by the hemiparetic subjects were reduced in comparison to those reported for healthy subjects. Their torque-angle curves had a curvilinear shape which indicated pronounced decrease of torque for plantarflexion efforts at the beginning of the movement. Torques produced at different velocities of testing did not demonstrate significant differences (MANOVAs: p > 0.05) but power values were significantly different. Results also showed that all the selected muscular parameters (torque, power and maximal rate of tension development) were moderately to highly interrelated (0.65 < r < 0.94; p < 0.01) suggesting that a common factor of muscular performance was assessed. Furthermore, the dynamometric data were significantly associated with some of the clinical measures (sensation and lower limb motor control scores of the FMA) but were not related to the gait variables (Pearson's r < 0.45; p > 0.05). This last finding suggests that the relationship between plantarflexor strength and the level of gait performance in adults with stroke is complex. The relationship may be influenced by other factors such as muscular compensations within and between limbs and motor control impairments.

Relationships between torque, velocity and power output during plantarflexion in healthy subjects.

This study investigated the relationships existing between torque, velocity and power output during plantarflexion. Using a Biodex dynamometric system, 15 healthy subjects performed three maximal dynamic tests, ranging from -12 degrees (-0.209 rad) of dorsiflexion to +47 degrees (+0.818 rad) of plantarflexion and one static test (test 4) at an angle of +10 degrees (+0.174 rad). The dynamic assessment included a 30 degrees s-1 (0.52 rad s-1) concentric isokinetic test (test 1) preceded by a 2-sec maximal pre-loading contraction. The other two dynamic tests were performed using the isotonic mode of testing with a selected torque of 27 N m; one of these tests was executed with pre-loading (test 2) while the other was performed without pre-loading (test 3). The results indicated that the dynamic peak torque, the peak power and the peak velocity were obtained in test 1, test 2 and test 3, respectively. These peak values, as well as the values of torque (test 1 and test 4), power (test 2) and velocity (test 3) obtained at a constant angle +10 degrees (+0.174 rad), were selected for the correlation analyses. The results showed that the torque, velocity and power output during plantarflexion were linearly related to one another with significant correlations (0.71 < r < 0.92; p < 0.01). This finding suggests that a common factor of muscular performance is assessed. Furthermore, these results indicated that the maximal torque produced by a subject can be predictive of his or her maximal velocity and power. Consequently, a stronger subject can generate higher velocity and power than a weaker subject when tested with the same load during maximal effort.

Normality and stationarity of EMG signals of elbow flexor muscles during ramp and step isometric contractions.

The purpose of this study was to test the stationarity and normality of electromyographic (EMG) signals obtained while exerting isometric contractions: (a) where a steady force level is maintained (step contractions); and (b) where the force level is increased linearly over time (ramp contractions). Ramp elbow flexions were performed from 0 to 100% of the maximum voluntary contraction (MVC) in a 5-s period. For the step contractions, four force levels (20, 40, 60 and 80% MVC) were maintained for a period of 3 s each. EMG signals of the biceps brachii (BB) and brachioradialis (BR) muscles of 16 subjects were recorded with surface electrodes and digitized at a sampling frequency of 2000 Hz. Tests of normality (Shapiro-Wilk test) and stationarity (reverse arrangement test) were performed locally on short finite time records (512-ms windows). Results show that, in general, EMG signals present a non-Gaussian amplitude distribution and are stationary. Furthermore, the amplitude distribution characteristics and the stationarity of the signal were not dependent on the muscle investigated, nor on the type of contraction or force level tested. The finding of local stationarity for both tasks is important, because it suggests that performing standard spectral analysis is applicable for both step and ramp contractions. It also allows a direct comparison between results obtained under both conditions.

A mechanical model to study the relationship between gait speed and muscular strength.

This study proposes a mechanical model to investigate the relationship between gait speed and strength of the ankle plantarflexor muscles. The model calculates the muscular utilization ratio (MUR) of the plantarflexor muscles during gait by comparing the plantarflexion moment used while walking to the maximal moment of the plantarflexors estimated from dynamometric measurements. To verify the model, MURs of the plantarflexor muscles were calculated for five healthy subjects and one hemiparetic subject walking at different speeds (slow, self-selected, and fast). Generally, the results of the healthy subjects revealed that MURs increase with an increasing gait speed: average (+/-SD) peak values of MUR reached 58.8% (+/-18.5), 65.6% (+/-17.2) and 71.0% (+/-17.8) for the slow, self-selected, and fast speeds, respectively. The average peak value of MURs at the self-selected speed corresponds to values reported in electromyographic studies of the plantarflexor muscles. At self-selected gait speed, the hemiparetic subject presented a higher peak MUR (80.5%) of the plantarflexors and a lower gait velocity when compared to healthy subjects. For the hemiparetic subject, peak values of MUR of the plantarflexor muscles at maximal walking speed reached 100% suggesting that full activation of the plantarflexors had been reached preventing him from walking faster. From these preliminary results, it appears that MURs calculated by the proposed model are sensitive to the mechanical demands imposed on a group of muscles during a task (eg., increase in gait speed) and to change in the maximal plantarflexor's strength (eg., weakness). The proposed model seems to have the potential to demonstrate whether muscle weakness limits maximal gait speed in hemiparetic subjects. However, considering the complexity of gait speed regulation in hemiparetic patients, the model should be tested on a large number of hemiparetic subjects.

Preloading and range of motion effect on plantarflexor muscle performance.

OBJECTIVE: To determine the effects of maximal preloading and range of motion (ROM) on the mechanical parameters of the plantarflexor muscles obtained while using the isotonic mode of testing of a Biodex dynamometer. DESIGN: A convenience sample of healthy subjects in the context of a descriptive comparative study. SETTING: Research laboratory in Canada. SUBJECTS: Fifteen volunteered subjects without history of injury or disorder to the right lower extremity. MEASUREMENTS: Four maximal isotonic tests were performed against a selected load of 27Nm. For the first two tests, the movement at the ankle ranged from -12 degrees (dorsiflexion) to +32 degrees (plantarflexion); one of these tests was preceded by a 2-second maximal preloading contraction, while the other was performed without preloading. For the other two tests, the ROM at the ankle was increased by 15 degrees of plantarflexion and thus ranged from -12 degrees to +47 degrees; again one of these tests was executed with preloading and the other without preloading. RESULTS: The four tests showed differences in the mechanical parameters (MANOVA p < .05). At angles of -10 degrees and +5 degrees, subjects produced higher torque and power but lower velocity values for the two tests preceded by a maximal preloading. The effect of ROM was demonstrated at +20 degrees where tests performed in a small amplitude reached a lower velocity than the corresponding tests performed in a larger amplitude. Based on the velocity profiles, the results also revealed that maximal preloading changed the selected isotonic movement of the Biodex dynamometer to an isoaccelerative movement characterized by high torque and power production. CONCLUSIONS: Isotonic assessment using the Biodex dynamometer provides different values of torque, velocity, and power depending on the testing conditions used. In clinical settings, it would be important to control these testing conditions.

Effects of transcutaneous electrical nerve stimulation on H-reflex and spinal spasticity.

The purpose of this study was to investigate the short-term effects of transcutaneous electrical nerve stimulation (TENS; 99 Hz; 250 ms pulses) on H-reflex and spinal spasticity. Considering the reflex hyperexcitability commonly displayed in spinal cord-injured subjects, it was hypothesized that repetitive low threshold afferent stimulation would have an inhibitory effect on the triceps surae H-reflexes which could also be reflected by a decrease in plantarflexor spasticity. Clonus, Achilles tendon reflex and modified Ashworth evaluations were performed on 14 spinal cord-injured subjects prior to and after 30 minutes' application of TENS. Non-parametric statistical analyses (n = 14; alpha = 0.05) failed to reveal significant effects of TENS on H-reflex amplitude. However, there was a significant decrease in scores for the Achilles tendon reflex and the modified Ashworth test. The clonus score decreased in most subjects post-TENS, although not in a statistically significant manner. The present pilot results thus suggest that TENS appears to be effective in reducing spinal spasticity, as measured clinically.

Changes in the electromyographic spectrum power distribution caused by a progressive increase in the force level.

The purpose of the present study was to determine the specific changes occurring in the power spectrum with an increasing force level during isometric contractions. Surface electromyographic signals of the triceps brachii (TB) and the anconeus (AN) of 29 normal subjects were recorded during isometric ramp contractions performed from 0 to 100% of the maximum voluntary contraction (MVC) in a 5-s period. Power spectra were obtained at 10, 20, 30, 40, 50, 60, 70, 80 and 90% MVC. Changes in the shape of these spectra were evaluated visually and with the calculation of several statistical parameters related to the distribution of power along the frequency axis, such as median frequency and mean power frequency, standard deviation, skewness, first and third quartiles and half-power range. For the AN, the behaviour of the spectrum was relatively similar across subjects, presenting a shift toward higher frequencies without any major change in the shape of the spectrum. For the TB, subjects with a thin skinfold thickness presented similar behaviours. In subjects with a thicker skinfold, however, a loss of power in the high frequency region paralleled the increase in the force level. Significant correlations were obtained between the extent of the change in the value of higher order statistical parameters across force and the thickness of the skin. This points out the importance of the skinfold layer when recording with surface electrodes. Furthermore, the use of a combination of several parameters appears to provide a better appreciation of the changes occurring in the spectrum than any single parameter taken alone.

Absence of consistent effects of repetitive transcutaneous electrical stimulation on soleus H-reflex in normal subjects.

Our purpose was to determine the effects of transcutaneous electrical nerve stimulation (TENS) on the soleus H-reflex amplitude in normal subjects. Eleven subjects were tested in five experimental sessions, the purpose of which was to compare the effects of 30 minutes of TENS delivered at either 50 or 99Hz (250 microseconds pulses) on a mixed (common peroneal nerve or CPN) versus a sensory (sural) nerve. The soleus H-reflex was elicited according to the classic protocol of Hugon (1973). Control values (Hctrl) were measured for 5 minutes prior to and for 10 minutes after the TENS was administered at twice the sensory threshold. No statistically significant session (treatment) effects (two-way ANOVAs for repeated measures; alpha = 0.05) resulted from the stimulation of the CPN or the sural nerve at 50 or 99Hz. However, although no specific trends were shown across all subjects, there was a definite tendency towards inhibition (> or = 10% Hctrl) of the H-reflex in 63% of the subjects after 30 minutes of TENS at 99Hz over the CPN, and in 50% of the subjects when TENS was applied over the sural nerve at 99Hz. The inherent variability of the H-reflex amplitude in normal subjects as well as the use of different stimulation paradigms and TENS parameters could explain the controversial findings present in the literature.

EMG power spectrum of elbow extensors: a reliability study.

In order to be considered a potential tool for the characterization of muscle activity, the reliability of EMG power spectral analysis should be demonstrated. In this study, the reliability of the mean power frequency (MPF) and of the median frequency (MF) of power spectra (triceps brachii (TB), anconeus (AN)) obtained at different force levels from both ramp and stepwise isometric contractions was tested across three similar sessions performed on three different days (N = 9). Two-way ANOVAs for repeated measures did not disclose any significant differences (p > 0.05) in the value of either the MF or the MPF across the different sessions for either type of contraction. In contrast, significant changes (p < 0.05) in both the MF and the MPF were found across force levels. No significant interactions (p > 0.05) were found between the session and the force factors, for any of the analyses performed. The present results indicate that the MPF and the MF of the EMG power spectrum, taken at a specific force level, are reliable measures across sessions performed on different days. Consequently, this supports the possible use of power spectral analysis of EMG signals as an evaluation technique that could monitor changes in the neuromuscular system that can occur over a given period.

Comparison of the EMG power spectrum of the human soleus and gastrocnemius muscles.

The purpose of this study was to compare the behaviour of electromyographic (EMG) power spectrum statistics, mean power frequency (MPF) and median frequency (MF), across increasing force levels of the soleus (SO), gastrocnemius medialis (GM) and gastrocnemius lateralis (GL) muscles. Surface EMG signals of these three muscles were recorded in 12 men and 10 women during both (1) ramp (single ongoing contractions with the force increasing linearly from 0 to 100% of the maximum voluntary contraction (MVC); and (2) step (steady force levels: 10, 20, 30, 40, 60 and 80% MVC) static (isometric) plantar flexions. Power spectral analysis of these signals was performed on single 256-ms windows at all of the above-mentioned force levels, for both types of contraction. The MF and MPF were calculated from each of the obtained spectra. A less pronounced increase in the MF or MPF was expected for the SO because of its higher type I fibre content. The main results are as follows: (1) similar behaviours were found in the value of MPF and MF across increasing force for the SO and GL muscles, while the GM gave rise to a different behaviour; (2) no difference was found between ramp and step contractions in the behaviour of either MF or MPF across force levels; and (3) different behaviours were observed between the MF and MPF across increasing force levels, for both ramp and step contractions. Our initial expectations were thus not confirmed. It is concluded that the present results support the hypothesis that the EMG power spectrum may be more sensitive to the diameter of the fibres than to the fibre type proportion of the triceps surae muscles. Furthermore, the sensitivity of the power spectrum statistics of a given muscle to the low-pass filter effect of its skin layer was also emphasized.