Median frequency of the myoelectric signal

Summary A study was performed to investigate the changes that occur in the median frequency of the myoelectric signal during local ischemia or reduction of intramuscular temperature produced by surface cooling. Data was obtained from experiments which involved the first dorsal interosseous muscle of 10 female and 16 male subjects. These subjects were asked to perform isometric constant-force abduction contractions of the index finger at 20% and 80% of maximal voluntary contraction level. The initial median frequency (IMF) of the myoelectric signal during the first 0.5 s of contraction was calculated. Results showed a significant reduction of the IMF in contractions performed under ischemic conditions; upon release, the IMF recovered quickly. At 80% maximal voluntary level of contraction, a greater decrease of the IMF was recorded. Similar results were demonstrated during reduction of intramuscular temperature with gradual recovery of the IMF after cooling. These results demonstrate that the median frequency of the myoelectric signal displays behavior similar to that reported for conduction velocity and this is consistent with the notion that accumulation of metabolic byproducts in muscle tissue causes a decrease in the conduction velocity of the muscle fibers.Dr. R. Merletti was on a leave of absence from the Institute of Electrical Engineering, Politecnico di Torino, Italy     

Relationship between myoelectric and mechanical manifestations of fatigue in the quadriceps femoris muscle group

Fatigue is commonly defined as “the failure to maintain the required force”. As such, it may be argued that the use of electromyographic (EMG) power spectral statistics to monitor muscle fatigue is inappropriate, because, during the maintenance of a submaximal force of contraction, EMG changes are readily observable in the absence of any decline in the muscle's mechanical output. However, it is possible that the EMG changes reflect the changing metabolic status of the muscle and hence its inability to generate its normal maximal force. The present study sought to examine whether the decline in EMG median frequency, which occurs during the maintenance of a submaximal force, is correlated with a reduction in the muscle's maximum force-generating capacity. The maximum voluntary contraction (MVC) of the knee extensors in ten young, healthy subjects was determined. On five separate occasions, randomly assigned forces of 20, 30, 40, 50 and 60% MVC were held to the limit of endurance. At intervals throughout the sustained contractions, subjects were required to rapidly generate an MVC for 1–2 s, then return to the fixed submaxial target force. Surface EMG signals were recorded throughout the contractions from the rectus lemons and vastus lateralis muscles, from which the power spectrum median frequency was calculated. Regression analysis revealed highly significant relationships between the rate of decline in MF and the rate of decline in MVC, and between each of these parameters and endurance time to fatigue (P = 0.0001, in each case). It is concluded that the decline in MF can be used to monitor fatigue, where fatigue is defined as the inability to generate the maximum force that can be produced by the muscle in its fresh state.     

Using risk factors, myoelectric signal, and finger tremor to distinguish computer users with and without musculoskeletal symptoms

Most of previous studies use questionnaire to assess risk factors for cumulative trauma disorders of the upper extremity (CTDUE) for computer workers. Few studies combine both physical examination and questionnaire to assess musculoskeletal symptoms. Fifteen symptomatic and 15 non-symptomatic computer users were recruited. Both of them were asked to perform a repetitive tapping task (200 taps/min) as the fatigue task. Tremor of the index finger and surface electromyography (EMG) of the flexor digitorum superficial (FDS) were collected prior and after the tapping task. Muscle strength and range of motion for right wrist were collected before the tapping task. All subjects were asked to fill out the questionnaire about risk factors of CTDUE. Female users in the symptomatic group had weaker wrist extensor strength than those in the non-symptomatic group (P < 0.05). After performing the tapping task, FDS strength and median frequency of the FDS EMG at 25%, and 100% maximal voluntary contraction (MVC) for the symptomatic group decreased (P < 0.05). However, no significant differences were found in strength and median frequency of the FDS EMG between prior and after tapping task in the non-symptomatic group. There was no significant difference for root mean square of the finger tremor between and within groups. Regression analysis revealed that median frequency of the FDS EMG at 25% MVC, age, total time spent on computer, and mouse position were better factors to classify computer users into the symptomatic group compared to other factors obtained from questionnaire and physical exam. Besides, symptomatic computer users seem to have longer experience of computer use than non-symptomatic users. After the fatigue task, the symptomatic users decreased their muscle strength to a larger extent than the non-symptomatic users.     

Control performance characteristics of myoelectric signal with additive interference

Myoelectric signal is an important source of control information for powered prostheses. A commonly used performance measure for the signal processors of such control systems is the ratio of processor output mean to variance. This ratio (SNR) is a function of a number of factors including physiological parameters and additive interference. The paper investigates the effects of motor unit physiological parameters and interference on control performance, with particular reference to SNR. Performance equations are derived and verified within vivo experiments. The results show a complex interaction among the physiological parameters and interference. A particular point of interest is the misleading SNR values that can occur under certain recruitment and interference conditions.     

Influence of handedness on motor unit discharge properties and force tremor

Discharge properties of motor units (MUs) in the first dorsal interosseous muscle (FDI) were studied in the dominant and non-dominant hands of six right-handed (RH) and six left-handed (LH) individuals. MU discharge rates and variability were similar in each hand in RH (186 MUs) and LH (160 MUs) subjects. MU synchronization was less prominent in the dominant hand of RH subjects, with 51% (45/88) of cross-correlograms of MU discharge having significant central peaks, compared with 81% (90/111) for the non-dominant hand. The strength of MU synchronization (expressed as the frequency of extra synchronous discharges above chance) was weaker in the dominant hand of right-handers (0.23 ± 0.03 s^-1 vs 0.39 ± 0.03 s^-1), and synchronous peaks from that hand were slightly broader. Four of six RH subjects had significant differences in synchronization between hands (weaker in dominant hand). In contrast, left-handers had similar incidence (80 vs 82%, n = 161) and strength (0.41 ± 0.03 s^-1 vs 0.37 ± 0.03 s^-1) of MU synchrony in dominant and non-dominant hands. No LH subject had a significant difference in synchronization between hands. Force tremor was quantified in each hand in the same subjects during isometric abduction of FDI at 0.5 N and 3.5 N, and directly correlated with the extent of MU synchronization in the muscle. Tremor root mean square amplitude was similar in dominant and non-dominant hands. Power spectral analysis of the tremor force revealed that the peak frequency in the power spectrum was not influenced by handedness, but power at the peak frequency was higher in the non-dominant hand of RH subjects. Correlations between MU discharge variability and synchrony with measures of tremor amplitude were weak. The reduced MU synchronization in the dominant hand of right-handers may reflect a more restricted distribution of direct projections from motor cortical neurons within the FDI motoneuron pool, or reduced excitability of the cortical neurons during the task. These differences in MU synchronization, however, had an insignificant influence on the magnitude of physiological tremor in the FDI.     

Feature-based classification of myoelectric signals using artificial neural networks

A pattern classification system, designed to separate myoelectric signal records based on contraction tasks, is described. The amplitude of the myoelectric signal during the first 200 ms following the onset of a contraction has a non-random structure that is specific to the task performed. This permits the application of advanced pattern recognition techniques to separate these signals. The pattern classification system described consists of a spectrographic preprocessor, a feature extraction stage and a classifier stage. The preprocessor creates a spectrogram by generating a series of power spectral densities over adjacent time segments of the input signal. The feature extraction stage reduces the dimensionality of the spectrogram by identifying features that correspond to subtle underlying structures in the input signal data. This is realised by a self-organising artificial neural network (ANN) that performs an advanced statistical analysis procedure known as exploratory projection pursuit. The extracted features are then classified by a supervised-learning ANN. An evaluation of the system, in terms of system performance and the complexity of the ANNs, is presented.     

Bilateral deficit expressions and myoelectric signal activity during submaximal and maximal isometric knee extensions in young,athletic males

The bilateral limb deficit (BLD) describes the difference in maximal or near maximal force generating capacity of muscles when they are contracted alone or in combination with the contralateral muscles. A deficit occurs when the summed unilateral force is greater than the bilateral force. This study examined the presence of the BLD during submaximal (25, 50, 75% of MVC) and maximal (100% MVC) isometric knee extensions in a group of young, athletic males (n = 6, mean age of 22 ± 3 years, mean height = 177.7 ± 6.4 cm, mean weight = 72.4 ± 5.2 kg). Torque and myoelectric signal (MES) data were collected from three superficial muscles of the quadriceps (vastus lateralis, vastus medialis and rectus femoris) during submaximal and maximal isometric knee extensions and it was found that a similar BLD exists using either torque or MES data. MES data showed that there were differences between bilateral and the total unilateral isometric knee extension regardless of percent contraction. This suggests that the BLD may be due to neural mechanisms and that future studies should examine the relationship between torque and the corresponding MES activity.     

A comparison of the mechanical properties of the first dorsal interosseous in the dominant and non-dominant hand

Summary The electrically evoked and voluntary contractile properties of the first dorsal interosseous muscle were measured on both hands in 10 healthy adults. The force of abduction of the index finger interosseous muscle was measured using a transducer resting against the lateral side of the proximal interphalangeal joint.The mean values of time to peak tension measured on the dominant hands were significantly slower than the values on the non-dominant hands (P<0.01) in a pairedt-test. Maximal tetanic tension, maximal voluntary contraction strength, and maximal twitch tension are not significantly different. Fatigue indices on the dominant hands in each subject were higher than those on the non-dominant hands. The correlation coefficient between fatigue indices on the dominant and the non-dominant hand was 0.92 (P<0.01).     

Real-time compression of myoelectric data utilising adaptive differential pulse code modulation

The myoelectric signal, obtained by either surface or needle electrodes, is used in many areas of clinical research and diagnosis. The conventional method of storing such information is in digitised form on a computer. However, the bandwidth of the signal and the required resolution result in large memory requirements. Adaptive differential pulse code modulation is investigated as a method of reducing the memory requirements for myoelectric data storage. In this scheme, a 12-bit sample is reduced to four bits, thus reducing the memory requirements by a factor of three. In reality, this compression ratio is closer to 4∶1 owing to the fact that the widths of most memories are organised as multiples of eight bits.     

Electromyogram median power frequency in dynamic exercise at medium exercise intensities

The electromyogram (EMG) median power frequency of the calf muscles was investigated during an exhausting treadmill exercise and a 20-min recovery period. The exercise was an uphill run at a speed of 5 km · h^–1 and a gradient of 20%. During exercise there was no decrease of EMG median power frequency. In contrast, EMG median power frequency in isometric contractions of the same muscles decreased by 7% for the soleus muscle and 16 to 18% for gastrocnemius muscles immediately after the exercise. During the recovery period the isometric median power frequency of the gastrocnemius muscles increased to pre-exercise levels in about 5 min. The isometric median power frequency of the soleus muscle also increased but had not reached pre-exercise values by 20 min. The observations from this study and from a previous uphill treadmill investigation at a steeper gradient gave evidence that two types of exhaustion can be distinguished during dynamic exercise; exhaustion at lower exercise intensities without a decrease in frequency during exercise and exhaustion at high intensities accompanied by a decline of frequency. The reason for this difference remains unclear.     

Further insight into the task-dependent excitability of motor evoked potentials in first dorsal interosseous muscle in humans

We have reexamined the contradictory evidence in which task-dependent excitation of motor evoked potentials (MEPs) in the first dorsal interosseous (FDI) muscle was stronger with increasingly more complex finger tasks than with individual finger movement tasks. In the first step of the experiment, based on previous findings, we investigated remarkable functional differences between intrinsic and extrinsic hand muscles during complex finger tasks (precision and power grip). During the performance of the tasks, the optimal stimulus intensity of the transcranial magnetic stimulation (TMS) was applied to the contralateral motor cortex. MEPs of the FDI, extensor carpi radialis (ECR), and flexor carpi radialis (FCR) muscles were recorded simultaneously with increased background EMG activity step by step in both tasks. The intensity threshold of TMS was lower in the precision grip. Furthermore, the MEP amplitudes of FDI muscle dependent on the background EMG activity were different between these two tasks, i.e., MEP amplitudes and regression coefficients in a precision grip were larger than those in a power grip. Although our results for MEP amplitude and threshold in the FDI muscle were similar to previous reported evidence, the different contributions of a synergistic muscle (in particular, the ECR muscle) during performance in these tasks was new evidence. Since there were no differences in cutaneous afferent effects on both tasks, corticomotoneuronal (CM) cells connected to FDI motoneurons seemed generally to be more active during precision than power gripping, and there were different contributions from synergistic muscles during the performance of these tasks. In the second part of the experiment, the results obtained from the complex tasks were compared with those from a simple task (isolated index finger flexion). MEP amplitudes, dependent on the background EMG activity during isolated index finger flexion, varied among subjects, i.e., the relationship between the MEP amplitude and the background EMG of the FDI muscle showed individual, strategy-dependent modulation. There were several kinds of individual motor strategies for performing the isolated finger movement. The present results may explain the previous contradictory evidence related to the contribution of the CM system during coordinated finger movement.     

Estimation of muscle force from intramuscular total pressure

Estimation of muscle force from the root mean square value of the myoelectric signal is limited to nonfatiguing contractions because of the change in the signal value that occurs with fatigue. In this paper the total intramuscular pressure is proposed as a muscle force estimator. Using a catheter, intramuscular pressure was measured for isometric time-varying force contractions and compared with the myoelectric signal mean absolute value. Results show that the pressure provides a linear estimator for load with an estimation error which is less than that obtained from the myoelectric signal. The myoelectric load estimator changes with fatigue but there is no significant change in the pressure load estimator.     

Motor unit discharge and force tremor in skill- and strength-trained individuals

We examined motor unit (MU) discharge properties (mean interspike interval, ISI, discharge variability, short-term synchronization, common drive) and force tremor in the first dorsal interosseous (FDI) muscle of five musicians (skill-trained), five weight-lifters (strength-trained) and six untrained subjects during low-force isometric abduction of the index finger. Mean MU ISI was slightly shorter in skill-trained subjects than in untrained subjects. Discharge variability of FDI MUs did not differ significantly between groups. The mean strength of MU synchronization (expressed as the frequency of extra synchronous discharges above chance) was different in skill-trained (0.22±0.02 s^–1, 162 MU pairs), untrained (0.32±0.02 s^–1, 199 MU pairs) and strength-trained subjects (0.44±0.03 s^–1, 183 MU pairs). FDI MU synchrony was weak and of equivalent strength in both hands of skill-trained subjects and the dominant (skilled) hand of untrained subjects. The stronger FDI MU synchrony in the non-dominant hand of untrained subjects was equivalent to that found in both hands of strength-trained subjects. The extent of common modulation of firing rates (common drive) was assessed for a subset of MU pairs and was weaker in skill-trained subjects (0.30±0.04, n=14) than untrained (0.43±0.3, n=14) and strength-trained (0.48±0.03, n=21) subjects. Force tremor was quantified for each hand in the same subjects during isometric index finger abduction at target forces of 0.5 N and 3.5 N. Tremor rms amplitude and peak power in the force frequency spectrum were significantly lower in skill-trained subjects than strength-trained subjects with the 3.5-N target force. The peak tremor frequency was similar in the three groups. The relatively more independent discharge of pairs of FDI MUs in skill-trained subjects was not responsible for the reduced tremor amplitudes in these subjects. Correlations between the overall extent of MU synchrony and common drive in FDI muscles and tremor measures obtained during the same experimental session were all non-significant. Differences in the central descending command signals are the most likely explanation for the more independent discharge of FDI MUs in skill-trained hands, while neural or peripheral muscular factors may be responsible for the weaker tremor. Received: 15 May 1997 / Accepted: 15 August 1997     

Changes in muscle strength and EMG median frequency after anterior cruciate ligament reconstruction

The purpose of this study was to monitor neuromuscular changes in quadriceps femoris muscle at 1 and 3 months after anterior cruciate ligament reconstruction (ACLR). Changes in isometric muscle strength (MVC), voluntary activation and surface electromyogram (EMG) parameters were examined in relation to knee stability, pain and swelling in 31 patients (25 M, 6 F) mean (SD) 30(8) years. Physically inactive (RC) and sports participants (SC) acted as controls. Median frequency (Hz) and amplitude (mV) of rectus femoris using Fast Fourier Transform (FFT) 2(11) was calculated during 5-s isometric contractions at 100, 75, 50 and 25% of MVC. One month after surgery, a significant correlation (P < 0.01) was found between activation (%) and MVC of injured knee extensors. By 3 months, most patients were pain free and had achieved full activation but still had muscle weakness. At 1 and 3 months post-surgery and for all levels of MVC contraction, the median frequencies of the injured limbs were significantly lower (P < 0.05) compared to the SC group as were those of the RC group. There was a significant lowering of the median frequencies of the uninjured limbs compared to the SC group at 75 and 100% of MVC. The EMG amplitude of the uninjured and injured limbs mirrored those of the SC and RC groups, respectively. These results support the view that muscle activation patterns were altered following ACL injury and surgical repair and may contribute to subsequent changes in muscle fibre properties during detraining and subsequent retraining.     

Statistics of the myoelectric signal from monopolar and bipolar electrodes

With the aid of some simplifying assumptions, a model is proposed that describes certain statistics of the myoelectric signal detected with a bipolar electrode. It is shown that these statistics are determined by the statistics of the monopolar signal and the spacing of the bipolar electrode pair. In particular, the dependence of the bipolar-signal bandwidth on electrode spacing is derived. The experimental results agree reasonably well with those predicted by the model.     

Effect of side dominance on myoelectric manifestations of muscle fatigue in the human upper trapezius muscle

The purpose of this study was to investigate whether differences in the peripheral and control properties of the neuromuscular system due to long-term preferential use, related to side dominance, affect postural muscles, such as the upper trapezius. Therefore, fatigability properties of the upper trapezius muscles of the dominant and non-dominant side were assessed. Surface EMG signals were detected from the upper trapezius muscles of both sides of nine right- and five left-handed subjects with adhesive linear electrode arrays consisting of eight contact points. Static constant force contractions with the arms 90° abducted were performed by the subjects while holding hand loads of 0 kg, 0.5 kg, and 1 kg. Surface EMG spectral and amplitude variables were computed from the recorded signals. EMG spectral variable rate of change (indicating fatigue) showed a statistically significant difference between the two sides, with the dominant side less fatigable than the non-dominant one. The observed differences held for both the right- and left-handed subject group. A possible explanation for the results is that long preferential use of one side with respect to the other leads to changes in muscle fiber membrane and control properties, in agreement with previous results on limb muscles.     

Reduced intracortical inhibition and facilitation of corticospinal neurons in musicians

Interhemispheric inhibition between motor cortices is reduced in musicians. In the present study we have assessed intracortical inhibition (ICI) and facilitation (ICF) within ipsilateral motor cortex in 15 musicians and 15 non-musician controls. Transcranial magnetic stimulation (TMS) was used to elicit muscle evoked potentials (MEPs) from left first dorsal interosseous (FDI) muscle at rest, and during voluntary index finger abduction (0.5 N). Paired TMS with subthreshold conditioning was used to test early ICI with interstimulus intervals (ISIs) 1-5 ms, and ICF with ISIs 8-15 ms. Suprathreshold conditioning was used to test late ICI with ISIs 100-200 ms. TMS thresholds were similar in musicians and controls both at rest and with weak voluntary activation of FDI, indicating that postsynaptic excitability of corticospinal neurons was similar in both groups. ICI was less effective in musicians with FDI at rest and active, but only with an ISI of 3 ms. ICF was less effective in musicians under both rest and active conditions, and this was independent of ISI. There were no differences in late ICI between musicians and controls. We conclude that ICI and ICF circuits which are activated by weak TMS have less influence on corticospinal neuron excitability in musicians. Because of the dependence on ISI, the most likely explanation for the reduced ICI in musicians is an alteration of the interaction between the ICI circuit and neural elements responsible for the later I-waves evoked in corticospinal neurons by TMS. Excitability of the neural elements producing early and late ICI is not altered in musicians. Reduced ICF in musicians could be due to reduced excitability of neurons responsible for ICF, or an altered balance of excitatory inputs to corticospinal neurons which favours neurons that are not acted upon by the ICF circuit. The reduced influence of ICI and ICF circuits on corticospinal neuron excitability in musicians is likely to reflect a training-induced adaptation. It is not clear at present whether these differences represent an adaptive change related to their extraordinary control of finger movements, or alternatively a maladaptive change induced by "overuse" of the hands from extensive training.     

Research summary of a scheme to ascertain the availability of speech information in the myoelectric signals of neck and head muscles using surface electrodes

The availability of speech information in the myoelectric signals (MES) of neck and head muscles was observed during five experiments conducted on two subjects. The MES of four channels, obtained using surface electrodes, was analog amplified, filtered and enhanced prior to digitization. Information was extracted at the rate of 20 points per second per channel using an average magnitude algorithm. The presence of speech related information was statistically verified with a pattern recognition algorithm based on the maximum likelihood algorithm. The results indicated that statistically significant (p = 0.01) information was present and that this scheme might be valuable in the future development of a vocal prosthesis.     

Evidence that myoelectric complexes in the isolated mouse colon may not be of myogenic origin

The hypothesis that spontaneous depolarisations (myoelectric complexes, MCs) can occur in the absence of neuronal activity, depending on the level of the membrane potential, was systematically studied. In control Krebs' solution, MCs were recorded approximately every 5 min and were abolished by tetrodotoxin (TTX, 1.6 μM). However, TTX also induced sustained membrane depolarisation (19 mV) in the circular muscle. To test whether MCs were blocked by the depolarisation induced by TTX, graded membrane repolarisations were generated, in the continuing presence of TTX, using sodium nitroprusside (SNP, 10 nM–1 μM). Under these conditions, MC activity was not restored. The addition of SNP (1 μM) to control preparations, in normal Krebs' solution, hyperpolarised the membrane of the circular muscle cells, but did not inhibit ongoing MC activity. It is suggested that the underlying mechanisms involved in MC generation are unlikely to be dependent upon the level of membrane potential in circular smooth muscle.