Estimating muscle activation from EMG using deep learning-based dynamical systems models

ObjectiveTo study the neural control of movement, it is often necessary to estimate how muscles are activated across a variety of behavioral conditions. However, estimating the latent command signal that underlies muscle activation is challenging due to its complex relation with recorded electromyographic (EMG) signals. Common approaches estimate muscle activation independently for each channel or require manual tuning of model hyperparameters to optimally preserve behaviorally-relevant features.ApproachHere, we adapted AutoLFADS, a large-scale, unsupervised deep learning approach originally designed to de-noise cortical spiking data, to estimate muscle activation from multi-muscle EMG signals. AutoLFADS uses recurrent neural networks (RNNs) to model the spatial and temporal regularities that underlie multi-muscle activation.Main ResultsWe first tested AutoLFADS on muscle activity from the rat hindlimb during locomotion, and found that it dynamically adjusts its frequency response characteristics across different phases of behavior. The model produced single-trial estimates of muscle activation that improved prediction of joint kinematics as compared to low-pass or Bayesian filtering. We also tested the generality of the approach by applying AutoLFADS to monkey forearm muscle activity from an isometric task. AutoLFADS uncovered previously uncharacterized high-frequency oscillations in the EMG that enhanced the correlation with measured force compared to low-pass or Bayesian filtering. The AutoLFADS-inferred estimates of muscle activation were also more closely correlated with simultaneously-recorded motor cortical activity than other tested approaches.SignificanceUltimately, this method leverages both dynamical systems modeling and artificial neural networks to provide estimates of muscle activation for multiple muscles that can be used for further studies of multi-muscle coordination and its control by upstream brain areas.

THE DIFFERENCE OF MUSCULOSKELETAL DISORDERS BEFORE AND AFTER WORKPLACE STRETCHING EXERCISE

Background: Workers who undertake occupations that require bending, climbing, crawling, reaching, twisting, excessive activity, or repeated movements may suffer from musculoskeletal disorders (MSDs). MSDs can be prevented by doing Workplace Stretching Exercise (WSE) which is useful for minimizing the risk of musculoskeletal injury, fatigue reduction, muscle balance, and posture improvement, and muscle coordination improvement. Purpose: To analyze the difference before and after giving WSE to the reduction of MSDs complaints. Method: This research takes a quantitative approach, utilizing a quasi-experimental design in one group before and after WSE administration. Total sampling is used for sampling. The respondents of this study were 34 workers in the production division of PT Crown Pratama in 2021. The T-paired test was utilized as a statistical test in this study. Result: The results of the univariate test mean MSDs complaints before and after WSE administration were 42.97 and 36.29. Conclusion: The bivariate test revealed differences in complaints of Musculoskeletal Disorders (MSDs) before and after workplace stretching exercise.

An Improved EMG-Driven Neuromusculoskeletal Model for Elbow Joint Muscle Torque Estimation

The accurate measurement of human joint torque is one of the research hotspots in the field of biomechanics. However, due to the complexity of human structure and muscle coordination in the process of movement, it is difficult to measure the torque of human joints in vivo directly. Based on the traditional elbow double-muscle musculoskeletal model, an improved elbow neuromusculoskeletal model is proposed to predict elbow muscle torque in this paper. The number of muscles in the improved model is more complete, and the geometric model is more in line with the physiological structure of the elbow. The simulation results show that the prediction results of the model are more accurate than those of the traditional double-muscle model. Compared with the elbow muscle torque simulated by OpenSim software, the Pearson correlation coefficient of the two shows a very strong correlation. One-way analysis of variance (ANOVA) showed no significant difference, indicating that the improved elbow neuromusculoskeletal model established in this paper can well predict elbow muscle torque.

Evaluation of Muscle Function by Means of a Muscle-Specific and a Global Index

Gait analysis applications in clinics are still uncommon, for three main reasons: (1) the considerable time needed to prepare the subject for the examination; (2) the lack of user-independent tools; (3) the large variability of muscle activation patterns observed in healthy and pathological subjects. Numerical indices quantifying the muscle coordination of a subject could enable clinicians to identify patterns that deviate from those of a reference population and to follow the progress of the subject after surgery or completing a rehabilitation program. In this work, we present two user-independent indices. First, a muscle-specific index (MFI) that quantifies the similarity of the activation pattern of a muscle of a specific subject with that of a reference population. Second, a global index (GFI) that provides a score of the overall activation of a muscle set. These two indices were tested on two groups of healthy and pathological children with encouraging results. Hence, the two indices will allow clinicians to assess the muscle activation, identifying muscles showing an abnormal activation pattern, and associate a functional score to every single muscle as well as to the entire muscle set. These opportunities could contribute to facilitating the diffusion of surface EMG analysis in clinics.

Home-based self-help telerehabilitation of the upper limb assisted by an electromyography-driven wrist/hand exoneuromusculoskeleton after stroke

Background Most stroke survivors have sustained upper limb impairment in their distal joints. An electromyography (EMG)-driven wrist/hand exoneuromusculoskeleton (WH-ENMS) was developed previously. The present study investigated the feasibility of a home-based self-help telerehabilitation program assisted by the aforementioned EMG-driven WH-ENMS and its rehabilitation effects after stroke. Methods Persons with chronic stroke (n = 11) were recruited in a single-group trial. The training progress, including the training frequency and duration, was telemonitored. The clinical outcomes were evaluated using the Fugl–Meyer Assessment (FMA), Action Research Arm Test (ARAT), Wolf Motor Function Test (WMFT), Motor Functional Independence Measure (FIM), and Modified Ashworth Scale (MAS). Improvement in muscle coordination was investigated in terms of the EMG activation level and the Co-contraction Index (CI) of the target muscles, including the abductor pollicis brevis (APB), flexor carpi radialis-flexor digitorum (FCR-FD), extensor carpi ulnaris-extensor digitorum (ECU-ED), biceps brachii (BIC), and triceps brachii (TRI). The movement smoothness and compensatory trunk movement were evaluated in terms of the following two kinematic parameters: number of movement units (NMUs) and maximal trunk displacement (MTD). The above evaluations were conducted before and after the training. Results All of the participants completed the home-based program with an intensity of 63.0 ± 1.90 (mean ± SD) min/session and 3.73 ± 0.75 (mean ± SD) sessions/week. After the training, motor improvements in the entire upper limb were found, as indicated by the significant improvements (P < 0.05) in the FMA, ARAT, WMFT, and MAS; significant decreases (P < 0.05) in the EMG activation levels of the APB and FCR-FD; significant decreases (P < 0.05) in the CI of the ECU–ED/FCR–FD, ECU–ED/BIC, FCR–FD/APB, FCR–FD/BIC, FCR–FD/TRI, APB/BIC and BIC/TRI muscle pairs; and significant reductions (P < 0.05) in the NMUs and MTD. Conclusions The results suggested that the home-based self-help telerehabilitation program assisted by EMG-driven WH-ENMS is feasible and effective for improving the motor function of the paretic upper limb after stroke. Trial registration ClinicalTrials.gov. NCT03752775; Date of registration: November 20, 2018.

An approximate stochastic optimal control framework to simulate nonlinear neuromusculoskeletal models in the presence of noise

Optimal control simulations have shown that both musculoskeletal dynamics and physiological noise are important determinants of movement. However, due to the limited efficiency of available computational tools, deterministic simulations of movement focus on accurately modelling the musculoskeletal system while neglecting physiological noise, and stochastic simulations account for noise while simplifying the dynamics. We took advantage of recent approaches where stochastic optimal control problems are approximated using deterministic optimal control problems, which can be solved efficiently using direct collocation. We were thus able to extend predictions of stochastic optimal control as a theory of motor coordination to include muscle coordination and movement patterns emerging from non-linear musculoskeletal dynamics. In stochastic optimal control simulations of human standing balance, we demonstrated that the inclusion of muscle dynamics can predict muscle co-contraction as minimal effort strategy that complements sensorimotor feedback control in the presence of sensory noise. In simulations of reaching, we demonstrated that nonlinear multi-segment musculoskeletal dynamics enables complex perturbed and unperturbed reach trajectories under a variety of task conditions to be predicted. In both behaviors, we demonstrated how interactions between task constraint, sensory noise, and the intrinsic properties of muscle influence optimal muscle coordination patterns, including muscle co-contraction, and the resulting movement trajectories. Our approach enables a true minimum effort solution to be identified as task constraints, such as movement accuracy, can be explicitly imposed, rather than being approximated using penalty terms in the cost function. Our approximate stochastic optimal control framework predicts complex features, not captured by previous simulation approaches, providing a generalizable and valuable tool to study how musculoskeletal dynamics and physiological noise may alter neural control of movement in both healthy and pathological movements.

Balance and motion coordination parameters can be improved in patients with type 2 diabetes with physical balance training: non-randomized controlled trial

Background Type 2 diabetes (T2D) is a cause of multiple complications, including retinopathy and peripheral neuropathy. These complications are well understood and believed to contribute to gait instability. Poor balance control and increased falling risk have also been reported in people with diabetic peripheral neuropathy (DPN). Patients with DPN have increased risk of falling due to decreased proprioceptive feedback. Effective balance training should improve postural control in patients with DPN. For this purpose further evaluation was conducted and balance training was designed. Methods The goal of our study was to determine values of proprioception, balance, muscle coordination and strength in patients with T2D and analyze whether biofeedback balance training with use of the Biodex Balance System could improve these parameters. To assess the fall risk the general stability index (GSI), the index of frontal-posterior (FPI) and medial–lateral (MLI) stability were evaluated. 37 patients with diagnosed type 2 diabetes mellitus were recruited to this study. Their results were compared with control group consisting of 41 healthy participants who were homogenic to the study group in terms of age and body mass index (BMI). Results There were statistically significant differences between patients with diabetes compared to healthy subjects in GSI (2.79 vs 1.1), FPI (1.66 vs 0.7), MLI (0.88 vs 0.52) and risk of falling (5.18 vs 2.72) p < 0.05. There were also statistically significant changes before and after training in all stability indices (GSI: 2.79 vs 1.26, FPI: 1.66 vs 0.77, MLI: 0.88 vs 0.54 accordingly) p < 0.05 and risk of falling (5.18 vs 3.87) p < 0.05 in the study group who had undergone training with biofeedback. Conclusions This study found that there is a decreased balance and motor coordination and an increased risk of falling in patients with type 2 diabetes. These parameters improved in patients who have undergone training programme with biofeedback. Furthermore, an age-dependent deprivation of static balance was observed along with an increased risk of falling as a result of increasing BMI.