Spatial variation of compound muscle action potentials across human gastrocnemius medialis

High-density EMG reveals atypical spatial activation of the gastrocnemius during walking in adolescents with Cerebral Palsy

Children with Cerebral Palsy (CP) exhibit less-selective, simplified muscle activation during gait due to injury of the developing brain. Abnormal motor unit recruitment, altered excitation-inhibition balance, and muscle morphological changes all affect the CP electromyogram. High-density surface electromyography (HDsEMG) has potential to reveal novel manifestations of CP neuromuscular pathology and functional deficits by assessing spatiotemporal details of myoelectric activity. We used HDsEMG to investigate spatial-EMG distribution and temporal-EMG complexity of gastrocnemius medialis (GM) muscle during treadmill walking in 11 adolescents with CP and 11 typically developed (TD) adolescents. Our results reveal more-uniform spatial-EMG amplitude distribution across the GM in adolescents with CP, compared to distal emphasis in TD adolescents. More-uniform spatial-EMG was associated with stronger ankle co-contraction and spasticity. CP adolescents exhibited a non-significant trend towards elevated EMG-temporal complexity. Homogenous spatial distribution and disordered temporal evolution of myoelectric activity in CP suggests less-structured and desynchronized recruitment of GM motor units, in combination with muscle morphological changes. Using HDsEMG, we uncovered novel evidence of atypical spatiotemporal activation during gait in CP, opening paths towards deeper understanding of motor control deficits and better characterization of changes in muscular activation from interventions.

Corticospinal excitability across lower limb muscles in humans

Electrophysiological studies in nonhuman primates reported the existence of strong corticospinal output from the primary motor cortex to distal compared with proximal hindlimb muscles. The extent to which corticospinal output differs across muscles in the leg in humans remains poorly understood. Using transcranial magnetic stimulation over the leg representation of the primary motor cortex, we constructed motor evoked potential (MEP) recruitment curves to measure the resting motor threshold (RMT), maximum MEP amplitude (MEP-max), and slope in the biceps femoris, rectus femoris, tibialis anterior, soleus, and a foot muscle (i.e., abductor hallucis) in intact humans. We found that the RMT was lower and the MEP-max and slope were larger in the abductor hallucis compared with most other muscles tested. In contrast, the RMT was higher and the MEP-max and slope were lower in the biceps femoris compared to all other muscles tested. Corticospinal responses in the rectus femoris, tibialis anterior, and soleus were in between those obtained from other leg muscles, with the soleus having a higher RMT and lower MEP-max and slope than the rectus femoris and tibialis anterior. To examine the origin of increases in corticospinal excitability in the abductor hallucis, we compared short-interval intracortical inhibition (SICI) and F-waves between the abductor hallucis and tibialis anterior. SICI was similar across muscles while the F-wave amplitude was larger in the abductor hallucis compared with the tibialis anterior. These results support a nonuniform distribution of corticospinal output to leg muscles, highlighting that increases in corticospinal excitability in a foot muscle could be related to a spinal origin.NEW & NOTEWORTHY We provide evidence on how corticospinal output differs across muscles in the leg in intact humans. We found that corticospinal responses were larger in a distal intrinsic foot muscle and were smaller in the biceps femoris compared to all other muscles in the leg. Increases in corticospinal excitability to an intrinsic foot muscle could have a spinal origin.

Inter-Person Differences in Isometric Coactivations of Triceps Surae and Tibialis Anterior Decrease in Young, but Not in Older Adults After 14 Days of Bed Rest

We examined activation patterns of the gastrocnemius medialis (GM), gastrocnemius lateralis (GL), soleus (SO), and tibialis anterior (TA) muscles in eight older (58.4 ± 3.3 years) and seven young (23.1 ± 2.9 years) participants, before and after 14 days of horizontal bed rest. Visual feedback on the exerted muscle torque was provided to the participants. The discharge patterns of individual motor units (MUs) were studied in three repetitions of isometric plantar flexion at 30 and 60% of Maximum Voluntary Contraction (MVC), before, and 1 day after the 14-day bed rest, respectively. In the GL and GM muscles, the older participants demonstrated higher MU discharge rates than the young, regardless of the contraction level, both before and after the bed rest. In the TA and SO muscles, the differences between the older and young participants were less consistent. Detailed analysis revealed person-specific changes in the MU discharge rates after the bed rest. To quantify the coactivation patterns we calculated the correlation coefficients between the cumulative spike trains of identified MUs from each muscle, and measured the root mean square difference of the correlation coefficients between the trials of the same session (intra-session variability) and between different sessions (inter-session variability) in each participant (intra-person comparison) and across participants (inter-person comparison). In the intra-person comparison, the inter-session variability was higher than the intra-session variability, either before or after the bed rest. At 60% MVC torque, the young demonstrated higher inter-person variability of coactivation than the older participants, but this variability decreased significantly after the bed rest. In older participants, inter-person variability was consistently lower at 60% than at 30% MVC torque. In young participants, inter-person variability became lower at 60% than at 30% MVC torque only after the bed rest. Precaution is required when analyzing the MU discharge and coactivation patterns, as individual persons demonstrate individual adaptations to aging or bed rest.

Endurance Trained Athletes Do Not per se Have Higher Hoffmann Reflexes Than Recreationally Active Controls

The impact of endurance training on spinal neural circuitries remains largely unknown. Some studies have reported higher H-reflexes in endurance trained athletes and therefore, adaptations within the Ia afferent pathways after long term endurance training have been suggested. In the present study we tested the hypothesis that cyclists (n = 12) demonstrate higher Hoffmann reflexes (H-reflexes) compared to recreationally active controls (n = 10). Notwithstanding, highly significant differences in endurance performance (VO_2peak: 60.6 for cyclists vs. 46.3 ml/min/kg for controls (p < 0.001) there was no difference in the size of the SOL H-reflex between cyclists and controls (H_max/M_max ratio 61.3 vs. 60.0%, respectively (p = 0.840). Further analyses of the H and M recruitment curves for SOL revealed a significant steeper slope of the M recruitment curve in the group of cyclists (76.2 ± 3.8° vs. 72.0 ± 4.4°, p = 0.046) without a difference in the H-recruitment curve (84.6 ± 3.0° vs. 85.0 ± 2.8°, p = 0.784) compared to the control group. Cycling is classified as an endurance sport and thus the findings of the present study do not further support the assumption that long-term aerobic training leads to a general increase of the H-reflex. Amongst methodological differences in assessing the H-reflex, the training-specific sensorimotor control of the endurance sport itself might differently affect the responsiveness of spinal motoneurons on Ia-afferent inputs.

Cross-correlated fractal components of H-wave amplitude fluctuations in medial gastrocnemius and soleus muscles

The time series of the H-wave amplitude in soleus muscle (SOL) shows fractal (long-range) correlation, which is attributed to input from supraspinal centers. However, whether such long-range power-law correlated input also contributes to the synergistic muscles remains unclear. The purpose of this study was therefore to examine the correlation in the fractal components of H-wave amplitude fluctuations between the synergistic muscles used for plantar flexion, i.e., the medial head of the gastrocnemius muscle (MG) and SOL. In eight young male participants, consecutive H-reflexes were recorded almost simultaneously from the MG and SOL at a stimulation frequency of 0.5 Hz for 30 min. We performed detrending moving-average cross-correlation analysis (DMCA) for each of the H- and M-wave amplitude time series between MG and SOL to assess the existence of a common noise input contributing to these long-range correlations. The cross-correlation coefficient ρDMCA (-1 to 1) was calculated to quantify the strength of the correlation between two different time series. The results indicated a significant long-range power-law correlation between H-wave amplitudes in MG and SOL (ρDMCA: 0.50 (0.22) and 0.22 (0.17), mean (standard deviation) for the original and randomly shuffled surrogate data, respectively, P < 0.05). This was not the case for M-wave amplitudes (ρDMCA: 0.29 (0.23) and 0.20 (0.15), P > 0.05). We conclude that there is a common noise input governing these synergistic muscles, possibly due to supraspinal origin, causing long-range power-law correlations in monosynaptic reflexes.

Novel Insights Into Biarticular Muscle Actions Gained From High-Density Electromyogram

Biarticular muscles have traditionally been considered to exhibit homogeneous neuromuscular activation. The regional activation of biarticular muscles, as revealed from high-density surface electromyograms, seems however to discredit this notion. We thus hypothesize the regional activation of biarticular muscles may contribute to different actions about the joints they span. We then discuss the mechanistic basis and methodological implications underpinning our hypothesis.

Association between Oxygen Consumption and Surface Electromyographic Amplitude and Its Variation within Individual Calf Muscles during Walking at Various Speeds

Surface electromyography (EMG) has been used to estimate muscle work and physiological burden of the whole body during human movements. However, there are spatial variations in surface EMG responses within individual muscles. The aim of this study was to investigate the relation between oxygen consumption and surface EMG responses of lower leg muscles during walking at various speeds and to quantify its spatial variation within an individual muscle. Nine young males walked on a treadmill at four speeds: preferred minus 1 km/h, preferred, preferred plus 1 km/h, and preferred plus 2 km/h, and the metabolic response was measured based on the expired gas. High-density surface EMG of the tibialis anterior (TA), medial gastrocnemius (MG), lateral gastrocnemius, and soleus muscles was performed using 64 two-dimensional electrode grids. Correlation coefficients between oxygen consumption and the surface EMG amplitude were calculated across the gait speeds for each channel in the electrode grid and for individual muscles. Mean correlation coefficients across electrodes were 0.69–0.87 for the four individual muscles, and the spatial variation of correlation between the surface EMG amplitude and oxygen consumption within an electrode grid was significantly greater in MG muscle than in TA muscle (Quartile deviations: 0.24 for MG and 0.02 for TA, p < 0.05). These results suggest that the physiological burden of the whole body during gait at various speeds can be estimated from the surface EMG amplitude of calf muscles, but we need to note its spatial distribution within the MG muscle.

The Accurate Assessment of Muscle Excitation Requires the Detection of Multiple Surface Electromyograms

When sampling electromyograms (EMGs) with one pair of electrodes, it seems implicitly assumed the detected signal reflects the net muscle excitation. However, this assumption is discredited by observations of local muscle excitation. Therefore, we hypothesize that the accurate assessment of muscle excitation requires multiple EMG detection and consideration of electrode-fiber alignment. We advise prudence when drawing inferences from individually collected EMGs.

Contraction level, but not force direction or wrist position, affects the spatial distribution of motor unit recruitment in the biceps brachii muscle

PURPOSE

Different motor units (MUs) in the biceps brachii (BB) muscle have been shown to be preferentially recruited during either elbow flexion or supination. Whether these different units reside within different regions is an open issue. In this study, we tested wheter MUs recruited during submaximal isometric tasks of elbow flexion and supination for two contraction levels and with the wrist fixed at two different angles are spatially localized in different BB portions.

METHODS

The MUs' firing instants were extracted by decomposing high-density surface electromyograms (EMG), detected from the BB muscle of 12 subjects with a grid of electrodes (4 rows along the BB longitudinal axis, 16 columns medio-laterally). The firing instants were then used to trigger and average single-differential EMGs. The average rectified value was computed separately for each signal and the maximal value along each column in the grid was retained. The center of mass, defined as the weighted mean of the maximal, average rectified value across columns, was then consdiered to assess the medio-lateral changes in the MU surface representation between conditions.

RESULTS

Contraction level, but neither wrist position nor force direction (flexion vs. supination), affected the spatial distribution of BB MUs. In particular, higher forces were associated with the recruitment of BB MUs whose action potentials were represented more medially.

CONCLUSION

Although the action potentials of BB MUs were represented locally across the muscle medio-lateral region, dicrimination between elbow flexion or supination seems unlikely from the surface representation of MUs action potentials.

Regional modulation of the ankle plantarflexor muscles associated with standing external perturbations across different directions

Maintenance of upright standing posture has often been explained using the inverted pendulum model. This model considers the ankle plantarflexors to act as a single synergistic group. There are differences in muscle properties among the medial and lateral gastrocnemius (MG and LG, respectively) and the soleus that may affect their activation. Twelve volunteers participated in an investigation to determine whether the activation of the ankle plantarflexor muscles was modulated according to perturbation direction during unilateral standing perturbations of 1% body mass. High-density surface electromyography (HDS-EMG) was used to determine the amplitude and barycenter of the muscle activation and kinematic analysis was used to evaluate ankle, knee, and hip joint movement. The HDS-EMG amplitude and barycenter of MG and LG were modulated with the perturbation direction (MG p < 0.05; LG p < 0.01; one-way repeated-measures ANOVA). In soleus, the HDS-EMG barycenter modulated across the perturbation direction (p < 0.01 for X&Y coordinates), but the HDS-EMG amplitude did not change. A repeated-measures correlation was used to interpret the HDS-EMG pattern in the context of the kinematics. The relative contribution of MG activation compared to the total gastrocnemii activation was significantly associated with ankle dorsi/plantarflexion (r_rm = 0.620), knee flexion/extension and abduction/adduction (r_rm = 0.622 and r_rm = 0.547, respectively), and hip flexion/extension and abduction/adduction (r_rm = 0.653 and r_rm = 0.432, respectively). The findings suggest that the central nervous system activates motor units within different regions of MG, LG and SOL in response to standing perturbations in different directions.

Potentiation of the first and second phases of the M wave after maximal voluntary contractions in the biceps brachii muscle

The study was undertaken to examine separately the potentiation of the first and second phases of the M wave in biceps brachii after conditioning maximal voluntary contractions (MVCs) of different durations. M waves were evoked in the biceps brachii muscle before and after isometric MVCs of 1, 3, 6, 10, 30, and 60 s. The amplitude, duration, and area of the first and second phases of monopolar M waves were measured during the 10-min period following each contraction. Our results indicated that the amplitude and area of the M-wave first phase increased after MVCs of long (≥ 30 s) duration (P < 0.05), while it decreased after MVCs of short (≤ 10 s) duration (P < 0.05). The enlargement after the long MVCs persisted for 5 min, whereas the depression after the short contractions lasted only for 15 s. The amplitude of the second phase increased immediately (1 s) after all MVCs tested (P < 0.05), regardless of their duration, and then returned rapidly (10 s) to control levels. Unexpectedly, the amplitude of the second phase decreased below control values between 15 s and 1 min after the MVCs lasting ≥ 6 s (P < 0.05). Our results reinforce the idea that the presence of fatigue is a necessary condition to induce an enlargement of the M-wave first phase and that this enlargement would be greater (and occur sooner) in muscles with a predominance of type II fibers (quadriceps and biceps brachii) compared to type-I predominant muscles (tibialis anterior). The unique findings observed for the M-wave second phase indicate that changes in this phase are highly muscle dependent. Graphical abstract Left panel-Representative examples of M waves recorded in one participant before (control) and at various times after conditioning maximal voluntary contractions (MVCs) of short (a1) and long (a2) duration. Left panel-Time course of recovery of the amplitude of the first (b1) and second (b2) phases of the M wave after conditioning MVCs of different durations.

Effect of TMS coil orientation on the spatial distribution of motor evoked potentials in an intrinsic hand muscle

Previous reports on the relationship between coil orientation and amplitude of motor evoked potential (MEP) in transcranial magnetic stimulation (TMS) did not consider the effect of electrode arrangement. Here we explore this open issue by investigating whether TMS coil orientation affects the amplitude distribution of MEPs recorded from the abductor pollicis brevis (APB) muscle with a bi-dimensional grid of 61 electrodes. Moreover, we test whether conventional mono- and bipolar montages provide representative MEPs compared to those from the grid of electrodes. Our results show that MEPs with the greatest amplitudes were elicited for 45° and 90° coil orientations, i.e. perpendicular to the central sulcus, for all electrode montages. Stimulation with the coil oriented at 135° and 315°, i.e. parallel to the central sulcus, elicited the smallest MEP amplitudes. Additionally, changes in coil orientation did not affect the spatial distribution of MEPs over the muscle extent. It has been shown that conventional electrodes with detection volume encompassing the APB belly may detect representative MEPs for optimal coil orientations. In turn, non-optimal orientations were identified only with the grid of electrodes. High-density electromyography may therefore provide new insights into the effect of coil orientation on MEPs from the APB muscle.

Characterization of torque generating properties of ankle plantar flexor muscles in ambulant adults with cerebral palsy

PURPOSE

Weakness of plantar flexor muscles is related to reduced push-off and forward propulsion during gait in persons with cerebral palsy (CP). It has not been clarified to what an extent altered muscle contractile properties contribute to this muscle weakness. Here, we investigated the torque generating capacity and muscle fascicle length in the triceps surae muscle throughout ankle range of motion (ROM) in adults with CP using maximal single muscle twitches elicited by electrical nerve stimulation and ultrasonography.

METHODS

Fourteen adults with CP (age 36, SD 10.6, GMFCS I-III) and 17 neurological intact (NI) adults (age 36, SD 4.5) participated. Plantar flexor torque during supramaximal stimulation of the tibial nerve was recorded in a dynamometer at 8 ankle angles throughout ROM. Medial gastrocnemius (MG) fascicle length was tracked using ultrasonography.

RESULTS

Adults with CP showed reduced plantar flexor torque and fascicle shortening during supramaximal stimulation throughout ROM. The largest torque generation was observed at the ankle joint position where the largest shortening of MG fascicles was observed in both groups. This was at a more plantarflexed position in the CP group.

CONCLUSION

Reduced torque and fascicle shortening during supramaximal stimulation of the tibial nerve indicate impaired contractile properties of plantar flexor muscles in adults with CP. Maximal torque was observed at a more plantarflexed position in adults with CP indicating an altered torque-fascicle length/ankle angle relation. The findings suggest that gait rehabilitation in adults with CP may require special focus on improvement of muscle contractility.

Does normalization of voluntary EMG amplitude to MMAX account for the influence of electrode location and adiposity?

Voluntary surface electromyography (sEMG) amplitude is known to be influenced by both electrode position and subcutaneous adipose tissue thickness, and these factors likely compromise both between- and within-individual comparisons. Normalization of voluntary sEMG amplitude to evoked maximum M-wave parameters (M_MAX peak-to-peak [P-P] and Area) may remove the influence of electrode position and subcutaneous tissue thickness. The purpose of this study was to: (a) assess the influence of electrode position on voluntary, evoked (M_MAX P-P and Area), and normalized sEMG measurements across the surface of the vastus lateralis (VL; experiment 1: n = 10); and (b) investigate if M_MAX normalization removes the confounding influence of subcutaneous tissue thickness [muscle-electrode distance (MED) from ultrasound imaging] on sEMG amplitude (experiment 2; n = 41). Healthy young men performed maximum voluntary contractions (MVCs) and evoked twitch contractions during both experiments. Experiment 1: voluntary sEMG during MVCs was influenced by electrode location (P ≤ 0.046, ES≥1.49 "large"), but when normalized to M_MAX P-P showed no differences between VL sites (P = 0.929) which was not the case when normalized to M_MAX Area (P < 0.004). Experiment 2: voluntary sEMG amplitude was related to MED, which explained 31%-38% of the variance. Normalization of voluntary sEMG amplitude to M_MAX P-P or M_MAX Area reduced but did not consistently remove the influence of MED which still explained up to 16% (M_MAX P-P) and 23% (M_MAX Area) of the variance. In conclusion, M_MAX P-P was the better normalization parameter for removing the influence of electrode location and substantially reduced but did not consistently remove the influence of subcutaneous adiposity.

Identification of regional activation by factorization of high-density surface EMG signals: A comparison of Principal Component Analysis and Non-negative Matrix factorization

In this study, we investigated whether principal component analysis (PCA) and non-negative matrix factorization (NMF) perform similarly for the identification of regional activation within the human vastus medialis. EMG signals from 64 locations over the VM were collected from twelve participants while performing a low-force isometric knee extension. The envelope of the EMG signal of each channel was calculated by low-pass filtering (8 Hz) the monopolar EMG signal after rectification. The data matrix was factorized using PCA and NMF, and up to 5 factors were considered for each algorithm. Association between explained variance, spatial weights and temporal scores between the two algorithms were compared using Pearson correlation. For both PCA and NMF, a single factor explained approximately 70% of the variance of the signal, while two and three factors explained just over 85% or 90%. The variance explained by PCA and NMF was highly comparable (R > 0.99). Spatial weights and temporal scores extracted with non-negative reconstruction of PCA and NMF were highly associated (all p < 0.001, mean R > 0.97). Regional VM activation can be identified using high-density surface EMG and factorization algorithms. Regional activation explains up to 30% of the variance of the signal, as identified through both PCA and NMF.

Changes in tibialis anterior architecture affect the amplitude of surface electromyograms

Background

Variations in the amplitude of surface electromyograms (EMGs) are typically considered to advance inferences on the timing and degree of muscle activation in different circumstances. Surface EMGs are however affected by factors other than the muscle neural drive. In this study, we use electrical stimulation to investigate whether architectural changes in tibialis anterior (TA), a key muscle for balance and gait, affect the amplitude of surface EMGs.

Methods

Current pulses (500 μs; 2 pps) were applied to the fibular nerve of ten participants, with the ankle at neutral, full dorsi and full plantar flexion positions. Ultrasound images were collected to quantify changes in TA architecture with changes in foot position. The peak-to-peak amplitude of differential M waves, detected with a grid of surface electrodes (16 × 4 electrodes; 10 mm inter-electrode distance), was considered to assess the effect of changes in TA architecture on the surface recordings.

Results

On average, both TA pennation angle and width increased by respectively 7 deg. and 9 mm when the foot moved from plantar to dorsiflexion (P < 0.02). M-wave amplitudes changed significantly with ankle position. M waves elicited in dorsiflexion and neutral positions were ~25% greater than those obtained during plantar flexion, regardless of where they were detected in the grid (P < 0.001). This figure increased to ~50% when considering bipolar M waves.

Conclusions

Findings reported here indicate the changes in EMG amplitude observed during dynamic contractions, especially when changes in TA architecture are expected (e.g., during gait), may not be exclusively conceived as variations in TA activation.

Optimization of surface electrodes location for H-reflex recordings in soleus muscle

The Hoffmann reflex (H reflex) is extensively used to investigate the spinal motor neuron excitability in healthy and pathological subjects. Obtaining a representative and robust amplitude estimation of the H reflex is of marked relevance in clinical as well as in research applications. As for the motor responses, this issue is strictly related to the electrode positioning, especially for large, pinnate muscles such as the triceps surae. In this study we investigated the effect of electrode position on soleus H-reflex amplitude. A grid of 96 electrodes was used to identify maximal H reflexes (Hmax) across the whole soleus region available for surface recording. The spatial distribution of Hmax amplitude detected in monopolar and single-differential derivations was used to determine where greatest reflex responses were detected from soleus. For both derivations and for all participants, largest Hmax were detected consistently over the central soleus region, in correspondence of the muscle superficial aponeurosis. Indeed, the amplitude of Hmax provided by conventional electrodes (1cm² area, 2cm apart) located centrally was significantly greater (median: 35% for monopolar and 79% for single-differential derivations) than that obtained medially, where surface electromyograms are typically recorded from soleus. Computer simulations, used to assist in the interpretation of results, suggest the soleus pinnate architecture was the key determinant of the medio-lateral variability observed for the experimental Hmax. The presented results provide a clear indication for electrode positioning, of crucial relevance in applied studies aimed at eliciting H reflexes.