Journal of Applied Physiology | April 2025
https://doi.org/10.1152/japplphysiol.00863.2024

Motor unit discharge behavior in human muscles throughout force gradation: a systematic-review and meta-analysis with meta-regression

Inglis, J. G.*, Cabral, H. V.*, Cosentino, C., Bonardi, A., Negro, F.
*Co-first authors

Abstact: 

The analysis of motor unit (MU) discharge behavior provides an effective way of assembling information about the generation and control of movement. In this systematic-review and meta-analysis we identified and summarized the literature investigating MU discharge rate and discharge rate variability (CoV-ISI) during voluntary isometric contractions at various force levels. Databases were searched up to January 7, 2025, and a total of 262 studies were included. The meta-means of MU discharge rate and CoV-ISI were estimated and compared across human muscles. The influence of contraction intensity on MU discharge behavior was assessed through linear meta-regressions. At low-to-moderate forces (<60% MVC), the first dorsal interosseous, biceps brachii (BB) and forearm extensors (FE) had the highest discharge rate, while the soleus had the lowest. At high force levels (>60% MVC), the tibialis anterior (TA) had the highest mean discharge rate compared to all other muscles, with the soleus maintaining the lowest. Regarding CoV-ISI results at low forces (<30% MVC), the TA had the lowest CoV-ISI values, except in comparison to the vastii. Additionally, the vastii had lower CoV-ISI values than the FE, gastrocnemius medialis, and soleus. Contraction intensity was positively associated with the mean discharge rates in all muscles investigated, although some muscles showed steeper slopes than others. Similar results were observed for CoV-ISI meta-regressions, with muscle-specific differences in slope. These findings suggest potential variations in neural control strategies across muscles during force gradation, such as differences in the relative contribution of rate coding to facilitate increasing force demands.

Journal of Neurophysiology | October 2024
https://doi.org/10.1152/jn.00391.2023

Achilles tendon morpho-mechanical parameters are related to triceps surae motor unit firing properties

Contreras-Hernandez, I., Arvanitidis, M., Falla, D., Negro, F., Martinez-Valdes, E.

Abstact: 

Recent studies combining high-density surface electromyography (HD-sEMG) and ultrasound imaging have yielded valuable insights into the relationship between motor unit activity and muscle contractile properties. However, limited evidence exists on the relationship between motor unit firing properties and tendon morpho-mechanical properties. This study aimed to determine the relationship between triceps surae motor unit firing properties and the morpho-mechanical properties of the Achilles tendon (AT). Motor unit firing properties [i.e. mean discharge rate (DR) and coefficient of variation of the interspike interval (COVisi)] and motor unit firing-torque relationships [cross-correlation between cumulative spike train (CST) and torque, and the delay between motor unit firing and torque production (neuromechanical delay)] of the medial gastrocnemius (MG), lateral gastrocnemius (LG), and soleus (SO) muscles were assessed using HD-sEMG during isometric plantarflexion contractions at 10% and 40% of maximal voluntary contraction (MVC). The morpho-mechanical properties of the AT (i.e. length, thickness, cross-sectional area, and resting stiffness) were determined using B-mode ultrasonography and shear-wave elastography. Multiple linear regression analysis showed that at 10% MVC, the DR of the triceps surae muscles explained 41.7% of the variance in resting AT stiffness. In addition, at 10% MVC, COVisi SO predicted 30.4% of the variance in AT length. At 40% MVC, COVisi MG and COVisi SO explained 48.7% of the variance in AT length. Motor unit-torque relationships were not associated with any morpho-mechanical parameter. This study provides novel evidence of a contraction intensity-dependent relationship between motor unit firing parameters of the triceps surae muscle and the morpho-mechanical properties of the AT.

eNeuro | June 2024
https://doi.org/10.1523/ENEURO.0043-24.2024

Neural filtering of physiological tremor oscillations to spinal motor neurons mediates short-term acquisition of a skill learning task

Cabral, H. V., Cudicio, A., Bonardi, A., Del Vecchio, A., Falciati, L., Orizio, C., Martinez-Valdes, E., Negro, F.

Abstact: 

The acquisition of a motor skill involves adaptations of spinal and supraspinal pathways to alpha motoneurons. In this study, we estimated the shared synaptic contributions of these pathways to understand the neural mechanisms underlying the short-term acquisition of a new force-matching task. High-density surface electromyography (HDsEMG) was acquired from the first dorsal interosseous (FDI; 7 males and 6 females) and tibialis anterior (TA; 7 males and 4 females) during 15 trials of an isometric force-matching task. For two selected trials (pre- and post-skill acquisition), we decomposed the HDsEMG into motor unit spike trains, tracked motor units between trials, and calculated the mean discharge rate and the coefficient of variation of inter-spike interval (CoVISI). We also quantified the post/pre ratio of motor units' coherence within delta, alpha, and beta bands. Force-matching improvements were accompanied by increased mean discharge rate and decreased CoVISI for both muscles. Moreover, the area under the curve within alpha band decreased by ~22% (TA) and ~13% (FDI), with no delta or beta bands changes. These reductions correlated significantly with increased coupling between force/neural drive and target oscillations. These results suggest that short-term force-matching skill acquisition is mediated by attenuation of tremor oscillations in the shared synaptic inputs. Supported by simulations, a plausible mechanism for alpha band reductions may involve spinal interneurons phase-cancelling descending oscillations. Therefore, during skill learning, the central nervous system acts as a matched filter, adjusting synaptic weights of shared inputs to suppress neural components unrelated to the specific task. 

The Journal of Physiology | May 2024
https://doi.org/10.1113/JP286078

Muscle contractile properties directly influence shared synaptic inputs to spinal motor neurons

Cabral, H. V., Inglis, J. G., Cudicio, A., Cogliati, M., Orizio, C., Yavuz, U. S., Negro, F.

Abstract: 

Alpha band oscillations in shared synaptic inputs to the alpha motor neuron pool can be considered an involuntary source of noise that hinders precise voluntary force production. This study investigated the impact of changing muscle length on the shared synaptic oscillations to spinal motor neurons, particularly in the physiological tremor band. Fourteen healthy individuals performed low-level dorsiflexion contractions at ankle joint angles of 90° and 130°, while high-density surface electromyography (HDsEMG) was recorded from the tibialis anterior (TA). We decomposed the HDsEMG into motor units spike trains and calculated the motor units’ coherence within the delta (1–5 Hz), alpha (5–15 Hz), and beta (15–35 Hz) bands. Additionally, force steadiness and force spectral power within the tremor band were quantified. Results showed no significant differences in force steadiness between 90° and 130°. In contrast, alpha band oscillations in both synaptic inputs and force output decreased as the length of the TA was moved from shorter (90°) to longer (130°), with no changes in delta and beta bands. In a second set of experiments (10 participants), evoked twitches were recorded with the ankle joint at 90° and 130°, revealing longer twitch durations in the longer TA muscle length condition compared to the shorter. These experimental results, supported by a simple computational simulation, suggest that increasing muscle length enhances the muscle's low-pass filtering properties, influencing the oscillations generated by the Ia afferent feedback loop. Therefore, this study provides valuable insights into the interplay between muscle biomechanics and neural oscillations.

Journal of Neural Engineering | March 2024
https://doi.org/10.1088/1741-2552/ad33b0

Adaptive HD-sEMG decomposition: towards robust real-time decoding of neural drive

Yeung, D., Negro, F., Vujaklija, I.

Abstact: 

Objective. Neural interfacing via decomposition of high-density surface electromyography (HD-sEMG) should be robust to signal non-stationarities incurred by changes in joint pose and contraction intensity. Approach. We present an adaptive real-time motor unit decoding algorithm and test it on HD-sEMG collected from the extensor carpi radialis brevis during isometric contractions over a range of wrist angles and contraction intensities. The performance of the algorithm was verified using high-confidence benchmark decompositions derived from concurrently recorded intramuscular electromyography. Main results. In trials where contraction conditions between the initialization and testing data differed, the adaptive decoding algorithm maintained significantly higher decoding accuracies when compared to static decoding methods. Significance. Using "gold standard" verification techniques, we demonstrate the limitations of filter re-use decoding methods and show the necessity of parameter adaptation to achieve robust neural decoding.

PLOS Computational Biology | January 2024
https://doi.org/10.1371/journal.pcbi.1011487

Postinhibitory excitation in motoneurons can be facilitated by hyperpolarization-activated inward currents: a simulation study

Schmid, L., Klotz, T., Röhrle, O., Powers, R., Negro, F., Yavuz, U.

Abstact: 

Postinhibitory excitation is a transient overshoot of a neuron’s baseline firing rate following an inhibitory stimulus and can be observed in vivo in human motoneurons. However, the biophysical origin of this phenomenon is still unknown and both reflex pathways and intrinsic motoneuron properties have been proposed. We hypothesized that postinhibitory excitation in motoneurons can be facilitated by hyperpolarization-activated inward currents (h-currents). Using an electrical circuit model, we investigated how h-currents can modulate the postinhibitory response of motoneurons. Further, we analyzed the spike trains of human motor units from the tibialis anterior muscle during reciprocal inhibition. The simulations revealed that the activation of h-currents by an inhibitory postsynaptic potential can cause a short-term increase in a motoneuron’s firing probability. This result suggests that the neuron can be excited by an inhibitory stimulus. In detail, the modulation of the firing probability depends on the time delay between the inhibitory stimulus and the previous action potential. Further, the postinhibitory excitation’s strength correlates with the inhibitory stimulus’s amplitude and is negatively correlated with the baseline firing rate as well as the level of input noise. Hallmarks of h-current activity, as identified from the modeling study, were found in 50% of the human motor units that showed postinhibitory excitation. This study suggests that h-currents can facilitate postinhibitory excitation and act as a modulatory system to increase motoneuron excitability after a strong inhibition.

IEEE Transactions on Neural Systems and Rehabilitation Engineering | 2023
https://doi.org/10.1109/TNSRE.2023.3326065

Optimal Motor Unit Subset Selection for Accurate Motor Intention Decoding: Towards Dexterous Real-Time Interfacing

Yeung, D., Negro, F., Vujaklija, I.

Abstract:

Objective: Motor unit (MU) discharge timings encode human motor intentions to the finest degree. Whilst tapping into such information can bring significant gains to a range of applications, current approaches to MU decoding from surface signals do not scale well with the demands of dexterous human-machine interfacing (HMI). To optimize the forward estimation accuracy and time-efficiency of such systems, we propose the inclusion of task-wise initialization and MU subset selection. Methods: Offline analyses were conducted on data recorded from 11 non-disabled subjects. Task-wise decomposition was applied to identify MUs from high-density surface electromyography (HD-sEMG) pertaining to 18 wrist/forearm motor tasks. The activities of a selected subset of MUs were extracted from test data and used for forward estimation of intended motor tasks and joint kinematics. To that end, various combinations of subset selection and estimation algorithms (both regression and classification-based) were tested for a range of subset sizes. Results: The mutual information-based minimum Redundancy Maximum Relevance (mRMR-MI) criterion retained MUs with the highest predicative power. When the portion of tracked MUs was reduced down to 25%, the regression performance decreased only by 3% (R2=0.79) while classification accuracy dropped by 2.7% (accuracy = 74%) when kernel-based estimators were considered. Conclusion and Significance: Careful selection of tracked MUs can optimize the efficiency of MU-driven interfacing. In particular, prioritization of MUs exhibiting strong nonlinear relationships with target motions is best leveraged by kernel-based estimators. Hence, this frees resources for more robust and adaptive MU decoding techniques to be implemented in future.

Journal of Neural Engineering | August 2023
https://doi.org/10.1088/1741-2552/ace7f7

High-density magnetomyography is superior to high-density surface electromyography for motor unit decomposition: a simulation study

Klotz, T., Lehmann, L., Negro, F., Röhrle, O

Abstact: 

Studying motor units (MUs) is essential for understanding motor control, the detection of neuromuscular disorders and the control of human-machine interfaces. Individual motor unit firings are currently identified in vivo by decomposing electromyographic (EMG) signals. Due to our body's electric properties, individual motor units can only be separated to a limited extent with surface EMG. Unlike electrical signals, magnetic fields pass through biological tissues without distortion. This physical property and emerging technology of quantum sensors make magnetomyography (MMG) a highly promising methodology. However, the full potential of MMG to study neuromuscular physiology has not yet been explored. In this work, we perform in silico trials that combine a biophysical model of EMG and MMG with state-of-the-art algorithms for the decomposition of motor units. This allows the prediction of an upper-bound for the motor unit decomposition accuracy. It is shown that non-invasive MMG is superior over surface EMG for the robust identification of the discharge patterns of individual motor units. Decomposing MMG instead of EMG increased the number of identifiable motor units by 71%. Notably, MMG exhibits a less pronounced bias to detect superficial motor units. The presented simulations provide insights into methods to study the neuromuscular system non-invasively and in vivo that would not be easily feasible by other means. Hence, this study provides guidance for the development of novel biomedical technologies.

bioRxiv | May 2025
https://doi.org/10.1101/2025.05.02.651931

Adaptations in common synaptic inputs to spinal motor neurons during grasping versus a less functional hand task

Cabral, H. V., Cosentino, C., Rizzardi, A., Inglis, J. G., Fuglevand, A., Negro, F.

Abstact: 

Previous evidence suggests that shared synaptic inputs across spinal motor neurons play a key role in coordinating multiple muscles during hand movements, reducing control complexity. In this study, we investigated how the nervous system modulates these common synaptic inputs during a functionally relevant grip (grasping) compared to less functionally relevant hand tasks. Seventeen participants performed three different tasks: simultaneous four-finger flexion without thumb involvement (four-finger flexion), thumb flexion, and simultaneous flexion of both fingers and thumb (grasping). For each task, subjects sustained isometric contractions at 5% and 15% of maximal voluntary contraction, while high-density surface electromyograms (HDsEMG) were recorded from the superficial extrinsic flexor muscles of the hand. Motor unit spike trains were decomposed from HDsEMG and tracked across tasks, and their mean discharge rate was calculated. Coherence between motor units was quantified within the delta, alpha, and beta bands to estimate common synaptic oscillations. At both force levels, the mean discharge rate decreased during grasping compared to four-finger flexion but increased during grasping compared to thumb flexion. Additionally, the area under the curve of coherence within the alpha band decreased by ~20% during grasping compared to the four-finger flexion task, with no significant delta or beta bands changes. These reductions in alpha band coherence were reflected in force oscillations, showing decreased force-neural drive coupling within the alpha band and increased force steadiness during grasping compared to four-finger flexion. Our findings suggest that a functionally relevant and frequently used grip involves distinct neural control mechanisms that ultimately enhance force control.

bioRxiv | April 2025
https://doi.org/10.1101/2025.04.23.650243

Differential changes in the effective neural drive following new motor skill acquisition between vastus lateralis and medialis

Cosentino, C., Cabral, H. V., Dos Santos, M. A., Pourreza, E., Inglis, J. G., Negro, F.

Abstact: 

Purpose: To investigate whether short-term learning of a new motor task is mediated by changes in common synaptic inputs to motor neurons within and between synergistic muscles. Methods: Seventeen healthy individuals performed 15 repetitions of a complex force-matching task at 10% of a maximal voluntary contraction. Two trials were selected for analysis, the one with the highest force-target error (pre-learning) and the one with the lowest (post-learning). High-density surface electromyograms recorded from vastus medialis (VM) and vastus lateralis (VL) were decomposed into their constituent motor unit spike trains, with individual motor units being tracked between trials. Motor unit discharge behavior and common synaptic oscillations across the delta, alpha, and beta bands were calculated and compared between pre- and post-learning. Results: Force-target matching improved across trials, accompanied by a significant decrease in the coefficient of variation of the inter-spike interval ( p < 0.01), while the mean discharge rate remained similar ( p > 0.85). The area under the curve within delta ( p < 0.003) and alpha ( p < 0.004) bands decreased between trials, with no significant changes in the beta band ( p > 0.05). Notably, reductions in the alpha band correlated significantly with performance improvements in VL (R = 0.81) but not in VM (R = 0.12). Conclusion: The acquisition of a new motor task is mediated by modulations in common synaptic inputs to motor units, leading to improved force control. Our findings further suggest that these changes in common synaptic inputs, particularly in the alpha band, differ between VM and VL.

bioRxiv | February 2025
https://doi.org/10.1101/2025.02.04.635938

A single low-dimensional neural component of motor unit activity explains force generation across repetitive isometric tasks

Cabral, H. V., Inglis, J. G., Pourreza, E., Dos Santos, M. A., Cosentino, C., O’Reilly, D., Delis, I., Negro, F.

Abstact: 

Previous studies suggest that low-dimensional control underlies motor unit activity, with low-frequency oscillations in common synaptic inputs serving as the primary determinant of muscle force production. In this study, we used principal component analysis (PCA) and factor analysis (FA) to investigate the relationship between low-dimensional motor unit components and force oscillations during repetitive isometric tasks with similar force profiles. We assessed the consistency of these components across trials in both individual (tibialis anterior; first dorsal interosseous) and synergistic muscles (vastus medialis, VM; vastus lateralis, VL). Participants performed 15 trials of a force-matching learning task. Three post-skill acquisition trials were selected for analysis to ensure high similarity in force profiles. Motor units were decomposed from high-density surface electromyograms, tracked across trials, and their smoothed discharge rates were decomposed into low-dimensional components using PCA and FA. Parallel analysis indicated that a single component could explain the smoothed discharge rates for the individual muscles and two components for VM-VL. Importantly, the first component explained most of the variance (∼70%) in smoothed discharge rates across all muscles. The first motor unit component also showed significantly higher correlations with force oscillations than the second component and remained highly consistent across trials. These findings were further supported by a non-linear framework combining network- and information-theoretic tools, which revealed high motor unit network density in the first component of all muscles. Collectively, these results suggest that, during isometric contractions, motor unit activity is primarily controlled by a single dominant shared synaptic input that closely mirrors force oscillations.

SRRN | October 2024
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5003494

Motor unit discharge behaviour in human muscles throughout force gradation: a systematic-review and meta-analysis with meta-regression

Inglis, J. G., Cabral, H. V., Cosentino, C., Bonardi, A., Negro, F.

Abstact: 

The analysis of motor unit (MU) discharge behaviour provides an effective way of assembling information about the generation and control of movement. In this systematic-review and meta-analysis we identified and summarized the literature investigating MU discharge rate and discharge rate variability (CoV-ISI) during voluntary isometric contractions at various force levels. Databases were searched up to May 9, 2024, and a total of 251 studies were included. The meta-means of MU discharge rate and CoV-ISI were estimated and compared across human muscles. The influence of contraction intensity on MU discharge behaviour was assessed through linear meta-regressions. At low-to-moderate forces (<60% MVC) the first dorsal interosseous, biceps brachii (BB) and forearm extensors (FE) had the highest discharge rate, while the soleus had the lowest. At high force levels (>60% MVC) the tibialis anterior (TA) had the highest mean discharge rate compared to all other muscles, with the soleus maintaining the lowest. Regarding CoV-ISI results at low forces (<30% MVC), the TA had the lowest CoV-ISI values, except in comparison to the BB and vastii. Additionally, the vastii had lower CoV-ISI values than the FE, gastrocnemius medialis and soleus. Contraction intensity was positively associated with the mean discharge rates in all muscles investigated, although some muscles showed steeper slopes than others. Similar results were observed for CoV-ISI meta-regressions, with muscle-specific differences in slope. These findings suggest potential variations in neural control strategies across muscles during force gradation, such as differences in the relative contribution of rate coding to facilitate increasing force demands.

medRxiv | August 2024
https://doi.org/10.1101/2024.08.27.24312381

Load and muscle dependent changes in triceps surae motor unit firing properties and motor unit firing-torque relationships in individuals with non-insertional Achilles tendinopathy

Contreras-Hernandez, I., Falla, D., Arvanitidis, M., Negro, F., Jimenez-Grande, D., Martinez-Valdes, E.

Abstact: 

Non-insertional Achilles tendinopathy (NIAT) induces morpho-mechanical changes to the Achilles tendon (AT). However, evidence on how triceps surae motor unit firing properties are influenced by altered tendon mechanics in NIAT is limited. This study investigated motor unit firing properties (mean discharge rate (DR), recruitment and de-recruitment thresholds, and discharge rate variability (COVisi)), motor unit firing-torque relationships (cross-correlation coefficient between cumulative spike train (CST) and torque, and neuromechanical delay), and neural drive distribution (connectivity strength and functional networks) of the medial gastrocnemius (MG), lateral gastrocnemius (LG), and soleus (SO) muscles during isometric plantarflexion contractions at 10%, 40%, and 70% maximal voluntary contraction (MVC) using high-density surface electromyography on 26 individuals with NIAT and 25 healthy controls. Furthermore, AT’s morpho-mechanical properties (thickness, cross-sectional area, length and stiffness) were assessed via ultrasound imaging. NIAT individuals showed reduced tendon stiffness and increased thickness (p<0.01). Motor unit properties changed in a load and muscle-dependent manner. LG DR increased (p=0.002) and de-recruitment threshold decreased (p=0.039) at 70%MVC in the NIAT group compared to controls. The CST-torque cross-correlation coefficient of the LG decreased at 10%MVC (p<0.0001) and increased at 70%MVC (p=0.013) in the NIAT group. Connectivity strength for the 0-5 Hz and 5-15 Hz frequency bands decreased (p<0.01) in the NIAT group at 10%MVC. This study shows that individuals with NIAT exhibit load-dependent changes in motor unit firing properties, motor unit-torque relationships, and neural drive distribution to the triceps surae. These alterations may be due to muscle-specific compensations for the modified mechanical properties of the AT.

bioRxiv | January 2024
https://doi.org/10.1101/2023.07.20.549840

Neural filtering of physiological tremor oscillations to spinal motor neurons mediates short-term acquisition of a skill learning task

Cabral, H. V., Cudicio, A., Bonardi, A., Del Vecchio, A., Falciati, L., Orizio, C., Martinez-Valdes, E., Negro, F.

Abstact: 

The acquisition of a motor skill involves adaptations of spinal and supraspinal pathways to alpha motoneurons. In this study, we estimated the shared synaptic contributions of these pathways to understand the neural mechanisms underlying the short-term acquisition of a new force-matching task. High-density surface electromyography (HDsEMG) was acquired from the first dorsal interosseous (FDI; 7 males and 6 females) and tibialis anterior (TA; 7 males and 4 females) during 15 trials of an isometric force-matching task. For two selected trials (pre- and post-skill acquisition), we decomposed the HDsEMG into motor unit spike trains, tracked motor units between trials, and calculated the mean discharge rate and the coefficient of variation of inter-spike interval (CoVISI). We also quantified the post/pre ratio of motor units’ coherence within delta, alpha, and beta bands. Improvements in force-matching were accompanied by a significant increase in the mean discharge rate and a decrease in CoVISI for both muscles. Moreover, the area under the curve within alpha band decreased by ∼22% and ∼13% for the TA and FDI muscles, respectively, with no changes in the delta or beta bands. These reductions correlated significantly with increased coupling between force/neural drive and target oscillations. These results suggest that the short-term acquisition of a new force-matching skill is mediated by the attenuation of tremor oscillations in the shared synaptic inputs. In other words, the central nervous system acts as a matched filter to modulate the synaptic weights of shared inputs and suppress neural components unrelated to the specific task. Supported by simulations, a plausible mechanism behind these alpha band reductions may involve spinal interneurons’ phase-cancelling descending oscillations.

bioRxiv | November 2023
https://doi.org/10.1101/2023.11.30.569389

Muscle contractile properties directly influence shared synaptic inputs to spinal motor neurons

Cabral, H. V., Inglis, J. G., Cudicio, A., Cogliati, M., Orizio, C., Yavuz, U. S., Negro, F.

Abstract: 

Alpha band oscillations in shared synaptic inputs to the alpha motor neuron pool can be considered an involuntary source of noise that hinders precise voluntary force production. This study investigated the impact of altering muscle length on the shared synaptic oscillations to spinal motor neurons, particularly in the physiological tremor band. Fourteen healthy individuals performed low-level dorsiflexion contractions at ankle joint angles of 90° and 130°, while high-density surface electromyography (HD-sEMG) was recorded from the tibialis anterior (TA). We decomposed the HDsEMG into motor units spike trains and calculated the motor units’ coherence within the delta (1-5 Hz), alpha (5-15 Hz) and beta (15-35 Hz) bands. Additionally, torque steadiness and torque spectral power within the tremor band was quantified. Results showed no significant differences in torque steadiness between 90° and 130°. In contrast, alpha band oscillations in both synaptic inputs and force output decreased as the length of the TA was moved from shorter (90°) to longer (130°), with no changes in delta and beta bands. In a second set of experiments, evoked twitches were recorded with the ankle joint at 70° and 130°, revealing longer twitch durations in the longer muscle lengthen condition compared to the shorter. These experimental results, supported by a simple computational simulation, suggest that increasing muscle length enhances the muscle’s low-pass filtering properties, influencing the oscillations generated by the Ia afferent feedback loop. Therefore, this study provides valuable insights into the interplay between muscle biomechanics and neural oscillations.

bioRxiv  | September 2023
https://doi.org/10.1101/2023.09.18.558259

Adaptive HD-sEMG decomposition: towards robust real-time decoding of neural drive

Yeung, D., Negro, F., Vujaklija, I.

Abstact: 

Neural interfacing via decomposition of high-density surface electromyography (HD-sEMG) should be robust to signal non-stationarities incurred by changes in joint pose and contraction intensity. We present an adaptive real-time motor unit (MU) decoding algorithm and test it on HD-sEMG collected from the extensor carpi radialis brevis during isometric contractions over a range of wrist angles and contraction intensities. The performance of the algorithm was verified using high-confidence benchmark decompositions derived from concurrently recorded intramuscular electromyography (iEMG). In trials where contraction conditions between the initialization and testing data differed, the adaptive decoding algorithm maintained significantly higher decoding accuracies when compared to static decoding methods. Using ‘gold standard’ verification techniques, we demonstrate the limitations of filter re-use decoding methods and show the necessity of parameter adaptation to achieve robust neural decoding.

bioRxiv | September 2023
https://doi.org/10.1101/2023.09.06.556472

Postinhibitory excitation in motoneurons can be facilitated by hyperpolarization-activated inward currents: a simulation study

Schmid, L., Klotz, T., Röhrle, O., Powers, R., Negro, F., Yavuz, U.

Abstact: 

Postinhibitory excitation is a transient overshoot of a neuron’s baseline firing rate following an inhibitory stimulus and can be observed in vivo in human motoneurons. However, the biophysical origin of this phenomenon is still unknown and both reflex pathways and intrinsic motoneuron properties have been proposed. We hypothesized that postinhibitory excitation in motoneurons can be facilitated by hyperpolarization-activated inward currents (h-currents). Using an electrical circuit model we investigated how h-currents can modulate the postinhibitory response of motoneurons. Further, we analyzed the spike trains of human motor units from the tibialis anterior muscle during reciprocal inhibition. The simulations revealed that the activation of h-currents by an inhibitory postsynaptic potential can cause a short-term increase in a motoneuron’s firing probability. This result suggests that the neuron can be excited by an inhibitory stimulus. In detail, the modulation of the firing probability depends on the time delay between the inhibitory stimulus and the previous action potential. Further, the strength of the postinhibitory excitation correlates with the amplitude of the inhibitory stimulus and is negatively correlated with the baseline firing rate as well as the level of input noise. Hallmarks of h-current activity, as identified from the modeling study, were found in 50 % of the human motor units that showed postinhibitory excitation. This study suggests that h-currents can facilitate postinhibitory excitation and act as a modulatory system to increase motoneuron excitability after a strong inhibition.

arXiv | June 2023
https://doi.org/10.48550/arXiv.2301.09494

High-density magnetomyography is superior to high-density surface electromyography for motor unit decomposition: a simulation study

Klotz, T., Lehmann, L., Negro, F., Röhrle, O

Abstact: 

Objective: Studying motor units (MUs) is essential for understanding motor control, the detection of neuromuscular disorders and the control of human-machine interfaces. Individual motor unit firings are currently identified in vivo by decomposing electromyographic (EMG) signals. Due to our body's properties and anatomy, individual motor units can only be separated to a limited extent with surface EMG. Unlike electrical signals, magnetic fields do not interact with human tissues. This physical property and the emerging technology of quantum sensors make magnetomyography (MMG) a highly promising methodology. However, the full potential of MMG to study neuromuscular physiology has not yet been explored. Approach: In this work, we perform in silico trials that combine a biophysical model of EMG and MMG with state-of-the-art algorithms for the decomposition of motor units. This allows the prediction of an upper-bound for the motor unit decomposition accuracy. Main results: It is shown that non-invasive high-density MMG data is superior over comparable high-density surface EMG data for the robust identification of the discharge patterns of individual motor units. Decomposing MMG instead of EMG increased the number of identifiable motor units by 76%. Notably, MMG exhibits a less pronounced bias to detect superficial motor units. Significance: The presented simulations provide insights into methods to study the neuromuscular system non-invasively and in vivo that would not be easily feasible by other means. Hence, this study provides guidance for the development of novel biomedical technologies.

2024 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering | October 2024
https://doi.org/10.1109/MetroXRAINE62247.2024.10795886

An innovative binary quadratic programming approach for the accurate identification of discharge timings of motor units from high-density surface EMG signals

Zanotti, R., Negro, F.

Abstact: 

Decoding motor unit (MU) activity in electromyographic (EMG) signals traditionally relies on spike sorting techniques, which require time-consuming expert intervention at some stage, limiting its use in neural engineering applications. We introduce a novel method using binary quadratic programming to identify MU discharge timings. Our approach, validated on simulated signals, demonstrates improved accuracy compared to conventional methods, typically consisting of unsupervised learning algorithms. In its optimal setup, the new method achieves a superior accuracy (94%±6), considerably higher than traditional methods (74%±21). Our approach offers a streamlined pathway for MU identification in high-density surface EMG (HDsEMG) recordings by automating the discharge identification process and reducing the need for expert oversight. This advancement holds significant promise for expanding the application of HDsEMG analysis in research and clinical settings within the field of neural engineering.

2023 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering | October 2023
https://doi.org/10.1109/MetroXRAINE58569.2023.10405698

High-density surface electromyography allows for longitudinal assessment of the neural drive to muscle in individuals with acute stroke

Benedini, M., Cabral, H. V., Cogliati, M., Falciati, L., Bissoloti, L., Orizio, C., McPherson, L., Negro, F.

Abstact: 

Previous work on neuromuscular impairments following stroke has mainly focused on the chronic phase of recovery, and relatively little is known regarding the acute phase. Studies demonstrating impairments in muscle activation have typically used single bipolar surface electromyography (sEMG) recordings, which may lead to a mischaracterization of muscle excitation. In this study, we assessed neuromuscular function of patients undergoing rehabilitation therapy in the acute phase post-stroke, combining high-density sEMG (HDsEMG) decomposition with isometric force recording to quantify changes in force production and motor unit discharge rates in comparison with global amplitude of a single bipolar sEMG. Seven patients with acute hemiparetic stroke were tested, beginning when a detectable dorsi- and plantarflexion movement could be observed (T0) and then again 15 and 30 days later (T15 and T30). The isometric maximal voluntary contraction (MVC) in dorsi- and plantarflexion were measured at these time points. HDsEMG signals recorded from tibialis anterior, gastrocnemius lateralis and medialis, and soleus muscles during isometric contractions at 10% and 30% MVC were decomposed into motor unit discharge offline. Our main results revealed significant impairments in maximal force production at T0, which improved over the 30 days of inpatient rehabilitation therapy. There were also increases in mean motor unit discharge rate for TA and SOL muscles at 10% MVC. These neuromuscular changes could not be captured by using the classical, bipolar sEMG approach. Our results suggest that the combination of force recordings with HDsEMG analysis may provide useful information in the acute phase of stroke and, longitudinally, during inpatient rehabilitation therapy.

45th Annual International Conference of the IEEE Engineering in Medicine & Biology Society | December 2023
https://doi.org/10.1109/EMBC40787.2023.10340187

Semi-Automated Identification of Motor Units Concurrently Recorded in High-Density Surface and Intramuscular Electromyography

Yeung, D., Negro, F., Vujaklija, I.

Abstact: 

An increasing focus on extending automated surface electromyography (EMG) decomposition algorithms to operate under non-stationary conditions requires rigorous and robust validation. However, relevant benchmarks derived manually from iEMG are laborsome to obtain and this is further exacerbated by the need to consider multiple contraction conditions. This work demonstrates a semi-automatic technique for extracting motor units (MUs) whose activities are present in concurrently recorded high-density surface EMG (HD-sEMG) and intramuscular EMG (iEMG) during isometric contractions. We leverage existing automatic surface decomposition algorithms for initial identification of active MUs. Resulting spike times are then used to identify (trigger) the sources that are concurrently detectable in iEMG. We demonstrate this technique on recordings targeting the extensor carpi radialis brevis in five human subjects. This dataset consists of 117 trials across different force levels and wrist angles, from which the presented method yielded a set of 367 high-confidence decompositions. Thus, our approach effectively alleviates the overhead of manual decomposition as it efficiently produces reliable benchmarks under different conditions.Clinical Relevance- We present an efficient method for obtaining high-quality in-vivo decomposition particularly useful in the verification of new surface decomposition approaches.

6th International Conference on NeuroRehabilitation (2024)
La Granja, Spain

Modulation of shared synaptic inputs to spinal motor neurons during short-term acquisition of a skill learning task

Cabral, H. V., dos Santos, M., Rizzardi, A., Benedini, M., Desmons, M., Pourreza, E., Inglis, J. G., Rizzetti, M. C., Padovani, A., Pilotto, A., Negro, F.

Abstract:

Repetitive transcranial magnetic stimulation (rTMS) has been extensively used to modulate cortical excitability and treat neurological disorders. However, the extent to which rTMS is effectively transmitted to spinal motor neuron output remains unexplored. In this study, we investigated the corticospinal transmission of various rTMS frequencies to motor unit activity. Eight participants performed low-level isometric contractions of thumb flexion while rTMS pulses were delivered to the primary motor cortex at 70% of the resting motor threshold. Stimulation frequencies of 5, 10, 20, 30 and 50 Hz were used. High-density surface electromyograms recorded from the thenar muscles were decomposed into motor unit spike trains, and the z-coherence between the rTMS stimuli (input) and the cumulative spike train (output) was calculated. Significant input-output coupling was observed only at frequencies of 20 Hz and higher. These preliminary findings suggest that higher rTMS frequencies are more efficiently transmitted to the neural drive to the muscle.

6th International Conference on NeuroRehabilitation (2024)
La Granja, Spain

Modulation of shared synaptic inputs to spinal motor neurons during short-term acquisition of a skill learning task

Cabral, H. V., Cosentino, C., dos Santos, M., Pourreza, E., Inglis, J. G., Negro, F.

Abstract:

The learning of a new motor task entails neural changes in spinal pathways. This study aimed to investigate alterations in the common synaptic inputs to spinal motor neurons during the short-term acquisition of a force-matching skill. 11, 13 and 17 participants had the first dorsal interosseus, tibialis anterior, and vastii muscles tested, respectively. They performed 15 isometric force-matching contractions at 5 or 10% of the maximal voluntary contraction. High-density surface electromyograms were decomposed for two selected trials: pre- and post-learning. The motor units were matched between trials and the coherence between cumulative spike trains was calculated within alpha (5-15 Hz) and beta (15-35 Hz) bands. In both individual and synergistic muscles, improvements in force-matching were followed by reductions in the alpha band, but not beta band. These findings suggest that during the learning of a new task, the central nervous system filters tremor noise oscillations unrelated to the required task.

XV Congresso Nazionale della Società Italiana delle Scienze Motorie (2024)
Chieti, Italy

Changes in the neural control of motor units of synergistic muscles during the acquisition of a new motor skill task

Cosentino, C., dos Santos, M., Cabral, H. V., Pourreza, E., Inglis, J. G., Negro, F.

Abstract:

Purpose: The acquisition of a new motor skill entails the activation and integration of different mechanisms at the cortical, muscular, and behavioral levels. Although supraspinal adaptations underlying motor skill acquisition are well investigated in the literature, the effects of skill learning on spinal synaptic connectivity are still poorly explored. Therefore, the aim of this study was to analyze potential alterations induced by short-term learning of a new motor skill on motor unit discharge patterns and common synaptic oscillations within and between synergistic muscles. Methods: Twelve participants initially performed maximal voluntary contractions (MVC) of isometric knee extension. Subsequently, they perform 15 trials of a challenging isometric force-matching task at 10% MVC. During the learning task, high-density surface electromyography (HDsEMG) was recorded from vastus medialis (VM) and lateralis (VL) muscles. Two trials were selected from the 15, the ones with highest error (pre-learning trial) and lowest error (post-learning trial) between force and target signals. HDsEMG from these two trials were decomposed into motor unit spike trains, through a convolutive blind source separation algorithm, and motor units were matched between trials. Mean discharge rate (MDR) and coefficient of variation of the inter-spike interval (COVISI) were calculated for the matched motor units. Furthermore, coherence analysis was conducted across different frequency bands (delta, alpha, and beta bands) to estimate common synaptic oscillations. Results: Significant improvements in force-matching were observed between pre- and post-learning, without significant changes in MDR in either muscle (p > 0.13). Contrarily, COVISI values in the VL significantly decreased from 45.42 ± 33.75% to 30.56 ± 26.10% between trials, but not in the VM (p = 0.07) For the coherence analysis, significant reductions of ~ 40% in common oscillations within alpha band were observed in the VL (p < 0.002). In addition, reductions of ~ 25% for alpha (p < 0.03) and 35% for delta (p < 0.008) bands were observed in the between-muscles coherence. Conclusions: Our results indicate that the short-term learning process of a new force-matching skill involves a reduction in alpha (VL and between-muscle) and delta (between-muscle) band oscillations. These changes might suggest the implementation of specific neural strategies to modulate and enhance the precision of the force output.

Society for Neuroscience Annual Meeting (2024)
Chicago, USA

Corticospinal transmission to spinal motor neuron output using repeated transcranial magnetic stimulation

dos Santos, M. A., Cabral, H. V.,  Rizzardi, A., Benedini, M., Desmons, M., Pourreza, E., Inglis, J. G.,  Pilloto, A., Negro, F.

Abstact: 

Repetitive transcranial magnetic stimulation (rTMS) has been extensively used to modulate cortical excitability in psychiatric and neurological disorders. However, the extent to which different rTMS frequencies are effectively transmitted to spinal motor neurons remains unexplored. Here, we investigated the corticospinal transmission of different rTMS-induced oscillations in the primary motor cortex (M1) to the neural drive to muscles. Eight participants performed 10% MVC isometric thumb flexions, while high-density surface electromyography (HDsEMG) was recorded from the thenar muscles. During the contractions, rTMS pulses were delivered on M1 contralaterally to the tested hand at 70% of the resting motor threshold. Stimulation frequencies of 5, 10, 20, 30, and 50 Hz were randomly applied. HDsEMG signals were decomposed into motor unit (MU) spike trains using a convolutive blind source separation algorithm. Z-coherence between the stimulation trigger (i.e., input) and the cumulative spike trains (CSTs) of decomposed MUs (i.e., output) was calculated and normalized by the number of MUs for each participant. Moreover, in a subgroup analysis (n = 4), we matched MUs between pre and during rTMS to calculate within muscle z-coherence. For both analyses, we quantified the area under the curve at each stimulus frequency and subsequent harmonics up to 100 Hz (windows of 2 Hz around each frequency). We applied statistical parametric mapping (t-test) to assess statistical significance and verify whether the areas under the z-coherence values differed from 0. On average, the number of MUs on each frequency were 11 (5Hz), 10 (10Hz), 8 (20Hz), 9 (30Hz) and 8 (50Hz). Main results showed that at low-frequency rTMS (5 and 10 Hz), significant coupling between the input and output occurred only at harmonic frequencies >= 20 and 30 Hz, respectively (P < 0.001 for both). Conversely, for rTMS at 20, 30, and 50 Hz, significant coupling was observed at the stimulus frequency and subsequent harmonics (P < 0.009). The results were confirmed in the subgroup analysis. These preliminary findings suggest that the transfer function of the corticospinal pathway to the alpha motoneuron output induced by rTMS has a marked bandpass behavior. In other words, high-frequency (> 20 Hz) cortical rTMS-induced oscillations appear to be more efficiently transmitted into the neural drive to the muscle. Therefore, this study provides insights into the corticospinal transmission of rTMS during isometric contractions and the effects of neuromodulation in pathological conditions. Additionally, it opens new perspectives on identifying robust neural biomarkers of corticospinal connectivity in humans. 

Society for Neuroscience Annual Meeting (2024)
Chicago, USA

Motor unit discharge behaviour in human muscles: a systematic-review and meta-analysis

Inglis, J. G., Cabral, H. V., Cosentino, C., Bonardi, A., Negro, F.

Abstact: 

Since the seminal work of Sherrington, the motoneuron has drawn much interest as the final common pathway of the neuromuscular system. The study of motoneurons in humans has focused on the activity of the motor unit (MU) during various tasks. Increased interest and technological advancements have led researchers to attempted to characterize MU behaviour throughout force gradation. However, because of the wide range of methods used across studies in relatively small sample sizes, it has been challenging to draw broad conclusions based on data from single studies. The primary aim of this systematic-review was to identify and summarize the findings of studies investigating MU discharge behaviour in various human muscles during isometric voluntary contractions. The secondary aim was to determine the influence of force output on MU discharge behaviour. Studies were searched from 1950 to 2024 in PubMed, Medline, and Web of Science databases. Included studies were primary surface or intramuscular electromyography (EMG) MU behaviour investigations using voluntary isometric contractions on baseline or control human subjects ≥18 years and who were free of neuromuscular impairment. Searches resulted in 14,759 papers identified, 8,931 remained after removing duplicates. 561 screened papers were identified for full text retrieval (33 not retrieved). Of the 528 remaining papers, after screening, 250 were included for data extraction. In addition, 11 were included through hand-searching. A data table was constructed from retrieved data and analysed to address the above research questions. Meta-means of MU discharge rate (MUDR) were calculated using multi-level models with random intercepts for study-level and were weighted using inverse-variance method. Subgroup analyses splitting the force outputs into HIGH (> 60%MVC) and LOW (≤ 20%MVC) were performed for the tibialis anterior (TA), vastus lateralis (VL), vastus medialis (VM), biceps brachii (BB), and first dorsal interosseous (FDI). There were significant increases in MUDR from the LOW to HIGH force output for all muscles (All p < 0.001; TA: 10.15 [9.66 10.64] pps; VL: 3.82 [3.37 4.27] pps; VM: 3.36 [2.74 3.99] pps; BB: 11 [9.81 12.18] pps; FDI: 8.92 [7.93 10.04] pps). Although all muscles MUDR increased from LOW to HIGH force output, the greatest differences were seen in the TA, BB and FDI compared to the VL and VM which may be related to differences in central and/or peripheral mechanisms. Overall, in the preliminary data, there is a significant increase in MUDR from LOW to HIGH force output suggesting the reliance on rate coding in the modulation of force. However, the magnitude of increase may be muscle dependant.

46th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (2024)
Orlando, USA

A quadratic programming approach for the exact resolution of superpositions of action potentials

Zanotti, R., Negro, F.

Abstact: 

Accurate decoding of neural activity is crucial for understanding the central nervous system. Current spike sorting methods face challenges when overlapping neural waveforms occur, impacting the accurate identification of the discharge times of each neuron. We introduce a computationally efficient quadratic programming approach to resolve waveform overlaps, offering flexibility and scalability. We tested the innovative method on single and multichannel intramuscular motor unit action potentials, and we compared it against existing state-of-the-art methods. The results show that our approach outperforms classical methods in both accuracy and speed. In particular, our approach was able to resolve complex superpositions of tens of action potentials, making it a valuable tool for the precise estimation of the neural code to muscles.

International Society for Electrophysiology and Kinesiology XXIV (2024)
Nagoya, Japan

Biomechanical changes in muscle length directly influence shared synaptic inputs to spinal motor neurons

Inglis, J. G., Cabral, H. V., Cudicio, A., Cogliati, M., Orizio, C., Yavuz, Utku S., Negro, F.

Abstact: 

BACKGROUND AND AIM: Alpha band oscillations in the shared synaptic input to the alpha motor neuron pool can be viewed as a source of common noise interfering with optimal force output control. Recent evidence has suggested that various factors can modulate this tremor frequency range in the common synaptic input to motor neurons [1-2]. This study investigated the impact of altering muscle length on the shared synaptic oscillations to spinal motor neurons and force steadiness. We also conducted a second set of experiments to explore how changes in muscle length affect the low-pass characteristics of the muscle twitches. METHODS: Fourteen participants performed trapezoidal isometric dorsiflexion contractions at 10% of maximal voluntary contraction while the ankle joint was placed at 90° (shortened) and 130° (lengthened). For both conditions, high-density surface electromyography (HDsEMG) recorded from the tibialis anterior (TA) was decomposed into motor unit spike trains [3], and motor unit coherence within the delta (1-5 Hz), alpha (5-15 Hz) and beta (15-35 Hz) bands was calculated. Torque steadiness and torque spectral power within the tremor band were quantified. In a second set of experiments, evoked torque twitches were recorded in five participants with the ankle joint placed at 70° and 130°. Wilcoxon signed-rank tests were used to compare the torque steadiness and power between conditions. Linear mixed models were applied to compare motor unit coherence between conditions. RESULTS: There were no significant differences in torque steadiness between ankle joint positions (P = 0.715). In contrast, torque power within alpha band significantly decreased between 90° and 130° (P = 0.009). Similarly, z-coherence within alpha band significantly decreased as TA length changed from 90° to 130° (from 1.20 [0.95, 1.45] to 0.99 [0.73, 1.24]; P < 0.001), with no changes for delta or beta bands (P > 0.117 for both). Comparison of the evoked potentials between 70° and 130° revealed an 8.3-14.1 ms increase in twitch duration. These changes in twitch duration induced alterations in the TA low-pass filtering characteristics, as the cut-off frequency was found to be lower when the muscle was lengthened than shortened. A simple computational simulation supported these experimental results, demonstrating that similar excitatory drives arriving at motor neuron ensembles with different twitch durations resulted in distinct alpha band coherence between motor unit spike trains. CONCLUSIONS: Our experimental results, supported by a simplified computational simulation, suggested that the increase of motor unit twitch duration resulting from increased muscle length directly influences the translation of the alpha band (physiological tremor) oscillations to force output. Therefore, this study provides valuable insights into the interplay between muscle biomechanics and neural adjustments. REFERENCES: [1] Laine et al., 2014. [2] Yavuz et al., 2015. [3] Negro et al., 2016.

International Society for Electrophysiology and Kinesiology XXIV (2024)
Nagoya, Japan

Exploring motor unit modes in repetitive isometric tasks

Cabral, H. V., Inglis, J. G., Pourreza, E., Cosentino, C., dos Santos, M. A., Negro, F.

Abstact: 

BACKGROUND AND AIM: The covariation in discharge rates can be used to estimate the motor unit (MU) modes (or common synaptic input components) underlying the activity of neuronal ensembles. Previous evidence suggested a principal component explaining fluctuations in MU discharge times within individual muscles [1], highly correlated with force output. Other findings proposed the existence of more than one factor, particularly in synergistic muscles [2]. This study used principal component analysis (PCA) and factorization analysis (FA) to investigate the consistency of MU modes during repetitive isometric tasks involving the same force output oscillations. METHODS: Two muscles, tibialis anterior (TA; 11 participants) and first dorsal interosseous (FDI; 7 participants), were assessed while participants completed 15 trials of an isometric force-matching task. Each trial involved following oscillations of a random signal for 30 s. The three consecutive trials with the smallest error between the force and target were selected. High-density surface electromyograms were decomposed into MU spike trains and MUs were tracked between trials. The number of components to be retained was determined using PCA with parallel analysis [3]. Subsequently, we applied PCA and FA (unrotated) to the standardized MU smoothed discharge rates. To explore the correlation between extracted MU mode fluctuations and force output, cross-correlation between these signals was calculated. Additionally, the cross-correlation between trials was computed to assess the consistency of the extracted MU modes. Linear mixed models (LMM) were used to compare main and interaction effect of trials and extracted MU modes on cross-correlation values. RESULTS: Parallel analysis revealed that one MU mode explained most of the variance of MU discharge rate in 11 out of 18 participants. In the other participants, two MU modes were extracted. We then opted to use two modes for subsequent analyses. For both muscles, the first MU mode was significantly more correlated with force output (~0.6 ± 0.1) than the second MU mode (~0.2 ± 0.1), regardless of the trial (LMM, P < 0.001 for all). Moreover, the first MU mode presented significantly greater correlation across trials (~0.5 ± 0.1) than the second MU mode (~0.1 ± 0.2; LMM, P < 0.001 for all). These results were consistent for both PCA and FA. CONCLUSIONS: Our main results demonstrated that the first MU mode showed high consistency across repetitive isometric tasks with similar force output, whereas the second mode did not. These findings suggest that motor units in individual muscles are mainly controlled by one low-frequency control synaptic input highly resembling the output force oscillations. In upcoming research, we aim to extend our methodology to synergistic muscles and examine how changes in the parametrization of PCA and FA might impact the results. REFERENCES: [1] Negro et al., 2009 [2] Del Vecchio et al., 2023 [3] Hayton et al., 2004

International Motoneuron Meeting (2024)
Bordeaux, France

Changes in common synaptic oscillations of synergistic muscles during the acquisition of a new motor skill task

Cabral, H. V., Cosentino, C., dos Santos, M. A., Pourreza, E., Inglis, J. G., Negro, F.

Abstact: 

The acquisition of a new motor skill involves the activation and integration of different mechanisms at the supraspinal and spinal levels. Although these supraspinal adaptations have been well characterized in the literature, the effects of motor skill learning on spinal synaptic connectivity remains relatively less explored. This study investigated whether short-term learning of a new motor skill modulates the common synaptic oscillations within and between synergistic muscles. Twelve participants performed maximal isometric voluntary (MVC) knee extension contractions followed by 15 trials of a challenging isometric force-matching task at 10% MVC. During this learning task, high-density surface electromyograms (HDsEMG) were recorded from the vastus medialis (VM) and lateralis (VL) muscles. Two trials were selected, one with the highest (pre-learning) and one with the lowest (post-learning) error between the force output and target trace. HDsEMGs were decomposed into motor unit spike trains using a convolutive blind source separation algorithm. The motor units were then matched between trials and subsequently, mean discharge rate (MDR) and coefficient of variation of the inter-spike interval (COVISI) were calculated. Common synaptic oscillations, within and between muscles, were estimated using coherence analysis for delta (1-5 Hz), alpha (5-15 hz) and beta (15-35 Hz) bands. The results showed significant improvements in force-matching between pre- and post-learning, however no significant changes in MDR were observed in either muscle (p > 0.16 for both). For COVISI, a significant reduction between trials was observed for VL (p < 0.01) but not for VM (p = 0.12). Regarding estimated common synaptic oscillations within muscle, significant reductions between pre- and post-learning were observed for VL, specifically within the alpha band (p < 0.002). In addition, shared synaptic oscillations between VM and VL significantly decreased between trials for both delta (p < 0.008) and alpha (p < 0.03) bands. Our results indicate that the acquisition of a force-matching skill entails reductions in alpha (VL and between muscles) and delta (between muscles) band oscillations. These preliminary findings indicate that specific changes in shared synaptic oscillations within and between synergistic muscles occur between pre- and post-learning, aimed at enhancing force output precision.

International Motoneuron Meeting (2024)
Bordeaux, France

Changes of rectus femoris length influence motor unit discharge behaviour of vastii muscles 

dos Santos, M. A., Cosentino, C., Pourreza, E., Inglis, J. G., Cabral, H. V., Negro, F.

Abstact: 

The quadriceps muscle group works in a synergistic way performing knee extension. While vastus lateralis (VL), medialis (VM) and intermedius are monoarticular muscles crossing the knee joint, rectus femoris (RF) is a biarticular muscle involved in both knee and hip joint movements. Since the control of knee extension force output relies on the interaction between quadriceps mechanical properties and neural inputs to spinal motor neurons of these muscles, changes in length of one muscle within this synergistic group may directly influence the neural output of the other muscles. In this study, we investigated the effect of modifying the length of the RF muscle, by altering the hip joint position, on motor unit discharge rate of the synergistic VM and VL muscles. Four women and five men (31.6 ± 7.7 years, 70.3 ± 23.8 kg, and 1.74 ± 0.9 m) participated in this preliminary study. After participants were positioned on an isokinetic dynamometer, they were asked to perform three maximal voluntary contractions (MVC) of isometric knee extension with the hip joint positioned at 90°(seated) and at 180° (supine position). Subsequently, they performed two trapezoidal contractions for each hip joint position at 10% MVC. During submaximal contractions, high-density surface electromyography (HDsEMG) signals were recorded from VM, RF and VL muscles and decomposed into individual motor unit (MU) spike trains using a convolutive blind source separation algorithm. MUs were then tracked between hip joint positions. The mean discharge rate of matched MUs was calculated and compared using linear mixed models (LMM). Our results showed no statistical differences in peak MVC between hip joint positions (Wilcoxon-test; p=1.00). The number of decomposed MUs for VM were 5 ± 1, from which 3 ± 1 were matched. For VL, 10 ± 8 were decomposed, from which 5 ± 8 were matched. Changing the hip joint angle from 90° to 180° (i.e., lengthening RF muscle) induced an increase in mean discharge rate of both VM (from 6.7 [5.28, 8.13] pps to 7.82 [6.39, 9.24] pps; LMM; p=0.01) and VL (from 7.69 [6.36, 9.01] pps to 8.51 [7.18, 9.83] pps; LMM; p=0.01). Our preliminary results suggest that to maintain the same force output between 90° and 180° hip joint angles, the neuromuscular system relies on a higher discharge rate of the synergistic VM and VL muscles, subsequently coping with the decreased force generation capacity of RF. 

International Motoneuron Meeting (2024)
Bordeaux, France

Motor unit discharge behaviour in human muscles: a systematic-review and meta-analysis

Inglis, J. G., Cabral, H. V., Cosentino, C., Bonardi, A., Negro, F.

Abstact: 

Since the seminal work of Sherrington, the motoneuron has drawn much interest as the final common pathway of the neuromuscular system. The study of motoneurons in humans has focused on the activity of the motor unit (MU) during various tasks. Increased interest and technological advancements have led researchers to attempted to characterize MU behaviour throughout force gradation. This review aimed to summarize the current wealth of knowledge in the study of MU behaviour between 1950 and 2023 to answer critical questions and address current knowledge gaps in the literature as we strive to better understand human motor control. The primary research question addressed discharge behaviours of MUs in various human muscles during isometric voluntary contractions. The secondary questions addressed the influence of force output on MU behaviour in various muscles and differences in MU behaviour between the upper, trunk and lower limbs, proximal to distal and small to large. Included studies were primary surface or intramuscular electromyography (EMG) MU behaviour investigations using voluntary isometric contractions on baseline or control human subjects ≥18 years and who were free of neuromuscular impairment. PubMed, Medline, and Web of Science searches resulted in 14,759 papers identified, 8,931 remained after removing duplicates. 561 screened papers were identified for full text retrieval (34 not retrieved). Of the 527 remaining papers, after screening, 432 were included for data extraction. A data table was constructed from retrieved data and analysed to address the above research questions. From a subgroup of the first 100 papers (alphabetically), meta-means of MU discharge rate (MUDR) were calculated using a random-effect meta-analysis with an inverse-variance method. Subgroup analyses splitting the force outputs into HIGH (>60%MVC) and LOW (≤30%MVC) were performed for the tibialis anterior (TA), vastus lateralis (VL), biceps brachii (BB), and first dorsal interosseous (FDI). There were significant increases in MUDR from the LOW to HIGH force output for all muscles (All p < 0.001; TA: 63%; VL: 41%; BB: 58%; FDI: 54%). Although all muscles MUDR increased from LOW to HIGH force output, the greatest differences were seen in the TA, BB and FDI compared to the VL which may be related to differences in central and/or peripheral mechanisms. Overall, in the preliminary data, there is a significant increase in MUDR from LOW to HIGH force output suggesting the reliance on rate coding in the modulation of force, regardless of muscle size, function, or location.

Society for Neuroscience Annual Meeting (2023)
Washington D. C., USA

Central and peripheral mechanisms modulating alpha band oscillations for enhanced precise force generation: experimental evidence from human motor control

Cabral, H. V., Negro, F.

Abstact: 

Previous studies have demonstrated a precise correspondence between muscle force output and the low-frequency components of the synaptic inputs broadly shared across the alpha motor neuron pool. In voluntary tasks, these shared inputs consist of control components that determine the command for optimal force generation, and common noise oscillations such as alpha band oscillations or physiological tremor, which reduce task precision. Thus, achieving precise force generation requires minimizing these shared noise components to maximize the control components in the force output. In this study, we conducted two sets of experiments to investigate central and peripheral mechanisms that may influence alpha band oscillations in the neural drive to the muscle to enhance precise force generation. The first experiment involved acquiring and decomposing high-density surface electromyography (HDsEMG) from the first dorsal interosseous (FDI; 13 participants) and tibialis anterior (TA; 11 participants) during the short-term acquisition of a new force-matching skill. Participants performed 15 trials of an isometric, challenging force-matching task. We selected the trials with the highest and lowest errors between the force and target (pre- and post-skill acquisition). We then quantified the ratio (post/pre) of the area under the curve of motor units’ z-coherence within delta (1-5 Hz), alpha (5-15 Hz) and beta (15-35 Hz) bands. We found that improvements in force-matching were accompanied by significant reductions of ~22% and ~13% in the area under the curve within the alpha band for the TA and FDI muscles, respectively. No changes were observed in the delta or beta bands. In the second experiment, 12 participants performed dorsiflexion isometric contractions for two TA muscle length conditions: shortened length (ankle at 90o) and optimal length for force production (110o). HDsEMG acquired from TA were decomposed into motor unit spike trains. The averages of motor units’ z-coherence within delta, alpha and beta bands were calculated and compared between TA muscle lengths. We found a significant reduction of ~14% in the alpha z-coherence values when the muscle length changed from shortened to optimal length. No such reduction was observed in the delta or beta bands. Our results provide experimental evidence that acquiring a force-matching skill and force production at optimal muscle length are associated with reductions in physiological tremor. These findings suggest that central (experiment 1) and peripheral (experiment 2) mechanisms may modify the characteristics of the shared noise inputs to alpha motoneurons and ultimately achieve optimal force generation.

XIV Congresso Nazionale della Società Italiana delle Scienze Motorie (2023)
Naples, Italy

Motor unit discharge pattern of the hand extrinsic flexor muscles changes between fingers flexion and synergistic finger-thumb flexion tasks

Cosentino, C., Cabral, H. V., Rizzardi, A., Orizio, C., Negro, F.

Abstract:

Purpose: Movement of the fingers requires a highly coordinated interplay of the hand extrinsic and intrinsic muscles, demanding complex control by the central nervous system. For instance, while the fingers do not flex in complete isolation, the opposable thumb has a high level of individuation and control. This suggests that there are potential differences in neural control between flexion of the fingers (fingers flexion task) and synergistic finger-thumb flexion (grasp task). In this study, we aimed to investigate the mean discharge rate of motor units recorded from the hand extrinsic flexor muscles during these two tasks using high-density surface electromyography (HDsEMG) signals. Methods: HDsEMG signals were recorded from the distal and proximal parts of the extrinsic flexor muscles in 15 healthy subjects. Two different tasks were performed: fingers flexion task, involving simultaneous flexion of the four fingers, and grasp task, involving synergistic flexion of the four fingers and the thumb. Both tasks were performed at a submaximal isometric level of 5% of maximal voluntary contraction. HDsEMG signals were decomposed into motor unit spike trains using a convolutive blind-source separation algorithm, and motor units were tracked between tasks. The mean discharge rate (DR) of motor units was then calculated and compared between the two tasks. In addition, to assess force steadiness, the coefficient of variation (CoV) of force was computed. Results: The analysis revealed significant changes in both force and motor unit mean discharge rate between fingers flexion and grasp tasks. The grasp task showed a reduction in force CoV (p = 0.006), reflecting an enhanced force steadiness. This improvement was accompanied by a reduction in mean DR of motor units located at the distal region of the hand extrinsic flexor muscles (p = 0.001), while no significant differences were observed at the proximal region (p = 0.505). Conclusions: Our findings indicate distinct neural control patterns between fingers flexion and grasp contractions. Importantly, our results suggest region-specific alterations in mean DR of hand extrinsic flexors, as differences were observed at distal motor units but not proximal. Moreover, our study demonstrates the feasibility of tracking and analyzing the same motor units in these specific tasks, offering methodological insights for future research. Acknowledgments: European Research Council Consolidator Grant INcEPTION (no. 101045605).

XIV Congresso Nazionale della Società Italiana delle Scienze Motorie (2023)
Naples, Italy

Exploring neural mechanisms underlying short-term skill acquisition: insights from alpha band oscillations estimated using motor unit spike trains and bipolar surface EMG

Cabral, H. V., Cudicio, A., Bonardi, A., Orizio, C., Negro, F.

Abstract:

Purpose: It is well-established that low-frequency oscillations in the synaptic inputs shared across the motor neuron pool directly impact the force output modulation. Among these oscillations, the alpha band components (5-15 Hz) are particularly implicated in introducing noise components that can reduce force precision. Consequently, when learning a new motor task that requires precise force generation, it becomes paramount to minimize these noise components unrelated to the intended motor task. In this study, we sought to investigate changes in alpha band oscillations during the short-term acquisition of a skilled motor task using two different approaches: EMG-EMG coherence and coherence between cumulative spike trains (CSTs) of motor units. Methods: High-density surface EMG (HDsEMG) was obtained from the first dorsal interosseous (FDI; 13 participants) and the tibialis anterior (TA; 11 participants). Participants performed 15 trials of an isometric, challenging force-matching task. We selected the trials with the highest and lowest errors between the force and target (pre- and post-skill acquisition). We then simulated two bipolar EMG signals from the HDsEMG grid and calculated the coherence between these signals (approach 1). Moreover, we decomposed HDsEMG signals into motor unit spike trains and calculated the coherence between the CSTs obtained from these motor units (approach 2). For both approaches, we quantified the ratio (post/pre) of the area under the curve of coherence within alpha band. Results: In the TA, improvements in force-matching were accompanied by significant reductions in alpha band oscillations, regardless of the approach used. For the FDI, a significant decrease in alpha coherence was also observed when using approach 2, but no changes were observed when using approach 1. Conclusions: Our findings indicate that the short-term acquisition of a skilled motor task involves changes in alpha band oscillations, suggesting neural adaptations aimed at minimizing common synaptic components unrelated to the required force fluctuations. Importantly, we observed these alterations in both investigated muscles using the CSTs approach but not the EMG-EMG coherence approach. These results suggest that the CSTs may be a more suitable methodology for detecting modifications in shared synaptic input during the acquisition of skilled motor tasks. Acknowledgments: European Research Council Consolidator Grant INcEPTION (no. 101045605).

Progress in Motor Control XIV (2023)
Rome, Italy

Changes in the shared synaptic input to the motor neurons during short-term acquisition of a force-matching skill

Cabral, H. V., Cudicio, A., Bonardi, A. Falciati, L., Orizio, C., Martinez-Valdes, E., Negro, F.

Abstact: 

Background: Previous studies demonstrated that only the low-frequency components of shared synaptic oscillations to the alpha motor neuron pools are represented in the force output. In tasks requiring voluntary control of motor outputs, this shared synaptic input is comprised by components related to the control process (e.g., control input signals from supraspinal pathways) and involuntary components that act as noise oscillations reducing the precision of the task (e.g., physiological tremor from spinal/supraspinal pathways). Thus, learning a new motor task involving precise force generation should require to minimize these shared noise components in order to maximize the representation of the shared synaptic input related to the voluntary control into the force output. Aim: We investigated whether the short-term acquisition of a new force-matching skill is mediated by specific changes in the components of the shared synaptic input to the motor neurons. Methods: Two muscles were recorded, first dorsal interosseous (FDI; 13 participants) and tibialis anterior (TA; 11 participants). Participants performed 15 trials of an isometric force-matching task at 5% MVC (FDI) and 10% MVC (TA). In each trial, they followed oscillations of a randomly generated signal for 30 s. The two trials with the highest and smallest root-mean-square-error (RMSE) between the force and target were selected to represent the pre- and post-skill acquisition trials. For these two trials, high-density surface electromyograms acquired from the muscles were decomposed into motor unit spike trains, and motor units were tracked between trials. The mean discharge rate and the coefficient of variation of the inter-spike interval were then calculated. To assess changes in the shared synaptic input with the force-matching skill acquisition, the coherence analysis was performed between two equally sized cumulative spike trains comprising motor units from the same muscle. We specifically quantified the ratio (post/pre) of the area under the curve of coherence within delta (0-5 Hz), alpha (5-15 Hz) and beta (15-35 Hz) bands. Results: For both muscles, RMSE between force and target signals significantly decreased from pre- to post-skill acquisition trial. These improvements in force-matching were followed by a significant increase in the mean discharge rate and a decrease in the coefficient of variation of the inter-spike interval. For the TA muscle, there was a significant reduction of ~24% in the area under the curve within the alpha band with the force-matching skill acquisition, which was not observed for delta or beta bands. For the FDI muscle, the area under the curve within the alpha and beta bands significantly decreased by ~23% and ~28%, but did not change for delta band. Conclusion: Our results suggest that the short-term acquisition of a new force-matching skill is mediated by an attenuation of the shared synaptic oscillations unrelated to the required force fluctuations (i.e., alpha and beta bands oscillations). Therefore, during the acquisition of a force-matching task, the central nervous system may tune the synaptic weights of the spinal and supraspinal projections to the motor neuron pool to create a matched filter that suppresses the neural components unrelated to the specific task.