MuscleNET: mapping electromyography to kinematic and dynamic biomechanical variables by machine learning

General Muscle Torque Generator Model for a Two Degree-of-Freedom Shoulder Joint

Muscle torque generators (MTGs) have been developed as an alternative to muscle-force models, reducing the muscle-force model complexity to a single torque at the joint. Current MTGs can only be applied to single Degree-of-freedom (DoF) joints, leading to complications in modeling joints with multiple-DoFs such as the shoulder. This study aimed to develop an MTG model that accounts for the coupling between 2-DoF at the shoulder joint: shoulder plane of elevation (horizontal abduction/adduction) and shoulder elevation (flexion/extension). Three different 2-DoF MTG equations were developed to model the coupling between these two movements. Net joint torques at the shoulder were determined for 20 participants (10 females and 10 males) in isometric, isokinetic, and passive tests. Curve and surface polynomial fitting were used to find the best general fit for the experimental data in terms of the different degrees of coupling. The models were validated against experimental isokinetic torque data. It was determined that implicit coupling that used interpolation between single-DoF MTGs resulted in the lowest root-mean-square percent error of 8.5%. The work demonstrated that general MTG models can predict torque results that are dependent on multiple-DoFs of the shoulder.

Mapping Method of Human Arm Motion Based on Surface Electromyography Signals

This paper investigates a method for precise mapping of human arm movements using sEMG signals. A multi-channel approach captures the sEMG signals, which, combined with the accurately calculated joint angles from an Inertial Measurement Unit, allows for action recognition and mapping through deep learning algorithms. Firstly, signal acquisition and processing were carried out, which involved acquiring data from various movements (hand gestures, single-degree-of-freedom joint movements, and continuous joint actions) and sensor placement. Then, interference signals were filtered out through filters, and the signals were preprocessed using normalization and moving averages to obtain sEMG signals with obvious features. Additionally, this paper constructs a hybrid network model, combining Convolutional Neural Networks and Artificial Neural Networks, and employs a multi-feature fusion algorithm to enhance the accuracy of gesture recognition. Furthermore, a nonlinear fitting between sEMG signals and joint angles was established based on a backpropagation neural network, incorporating momentum term and adaptive learning rate adjustments. Finally, based on the gesture recognition and joint angle prediction model, prosthetic arm control experiments were conducted, achieving highly accurate arm movement prediction and execution. This paper not only validates the potential application of sEMG signals in the precise control of robotic arms but also lays a solid foundation for the development of more intuitive and responsive prostheses and assistive devices.

Scalable musculoskeletal model for dynamic simulations of upper body movement

A musculoskeletal (MSK) model is a valuable tool for assessing complex biomechanical problems, estimating joint torques during motion, optimizing motion in sports, and designing exoskeletons and prostheses. This study proposes an open-source upper body MSK model that supports biomechanical analysis of human motion. The MSK model of the upper body consists of 8 body segments (torso, head, left/right upper arm, left/right forearm, and left/right hand). The model has 20 degrees of freedom (DoFs) and 40 muscle torque generators (MTGs), which are constructed using experimental data. The model is adjustable for different anthropometric measurements and subject body characteristics: sex, age, body mass, height, dominant side, and physical activity. Joint limits are modeled using experimental dynamometer data within the proposed multi-DoF MTG model. The model equations are verified by simulating the joint range of motion (ROM) and torque; all simulation results have a good agreement with previously published research.

Scalable musculoskeletal model for dynamic simulations of lower body movement

A musculoskeletal (MSK) model is an important tool for analysing human motions, calculating joint torques during movement, enhancing sports activity, and developing exoskeletons and prostheses. To enable biomechanical investigation of human motion, this work presents an open-source lower body MSK model. The MSK model of the lower body consists of 7 body segments (pelvis, left/right thigh, left/right leg, and left/right foot). The model has 20 degrees of freedom (DoFs) and 28 muscle torque generators (MTGs), which are developed from experimental data. The model can be modified for different anthropometric measurements and subject body characteristics, including sex, age, body mass, height, physical activity, and skin temperature. The model is validated by simulating the torque within the range of motion (ROM) of isolated movements; all simulation findings exhibit a good level of agreement with the literature.

A novel unsupervised dynamic feature domain adaptation strategy for cross-individual myoelectric gesture recognition

Objective.Surface electromyography pattern recognition (sEMG-PR) is considered as a promising control method for human-machine interaction systems. However, the performance of a trained classifier would greatly degrade for novel users since sEMG signals are user-dependent and largely affected by a number of individual factors such as the quantity of subcutaneous fat and the skin impedance.Approach.To solve this issue, we proposed a novel unsupervised cross-individual motion recognition method that aligned sEMG features from different individuals by self-adaptive dimensional dynamic distribution adaptation (SD-DDA) in this study. In the method, both the distances of marginal and conditional distributions between source and target features were minimized through automatically selecting the optimal feature domain dimension by using a small amount of unlabeled target data.Main results.The effectiveness of the proposed method was tested on four different feature sets, and results showed that the average classification accuracy was improved by above 10% on our collected dataset with the best accuracy reached 90.4%. Compared to six kinds of classic transfer learning methods, the proposed method showed an outstanding performance with improvements of 3.2%-13.8%. Additionally, the proposed method achieved an approximate 9% improvement on a publicly available dataset.Significance.These results suggested that the proposed SD-DDA method is feasible for cross-individual motion intention recognition, which would provide help for the application of sEMG-PR based system.

Experimental Study of Fully Passive, Fully Active, and Active-Passive Upper-Limb Exoskeleton Efficiency: An Assessment of Lifting Tasks

Recently, robotic exoskeletons are gaining attention for assisting industrial workers. The exoskeleton power source ranges from fully passive (FP) to fully active (FA), or a mixture of both. The objective of this experimental study was to assess the efficiency of a new active-passive (AP) shoulder exoskeleton using statistical analyses of 11 quantitative measures from surface electromyography (sEMG) and kinematic data and a user survey for weight lifting tasks. Two groups of females and males lifted heavy kettlebells, while a shoulder exoskeleton helped them in modes of fully passive (FP), fully active (FA), and active-passive (AP). The AP exoskeleton outperformed the FP and FA exoskeletons because the participants could hold the weighted object for nearly twice as long before fatigue occurred. Future developments should concentrate on developing sex-specific controllers as well as on better-fitting wearable devices for women.

A Review of Myoelectric Control for Prosthetic Hand Manipulation

Myoelectric control for prosthetic hands is an important topic in the field of rehabilitation. Intuitive and intelligent myoelectric control can help amputees to regain upper limb function. However, current research efforts are primarily focused on developing rich myoelectric classifiers and biomimetic control methods, limiting prosthetic hand manipulation to simple grasping and releasing tasks, while rarely exploring complex daily tasks. In this article, we conduct a systematic review of recent achievements in two areas, namely, intention recognition research and control strategy research. Specifically, we focus on advanced methods for motion intention types, discrete motion classification, continuous motion estimation, unidirectional control, feedback control, and shared control. In addition, based on the above review, we analyze the challenges and opportunities for research directions of functionality-augmented prosthetic hands and user burden reduction, which can help overcome the limitations of current myoelectric control research and provide development prospects for future research.

Evaluation of a machine-learning-driven active-passive upper-limb exoskeleton robot: Experimental human-in-the-loop study

Evaluating exoskeleton actuation methods and designing an effective controller for these exoskeletons are both challenging and time-consuming tasks. This is largely due to the complicated human-robot interactions, the selection of sensors and actuators, electrical/command connection issues, and communication delays. In this research, a test framework for evaluating a new active-passive shoulder exoskeleton was developed, and a surface electromyography (sEMG)-based human-robot cooperative control method was created to execute the wearer's movement intentions. The hierarchical control used sEMG-based intention estimation, mid-level strength regulation, and low-level actuator control. It was then applied to shoulder joint elevation experiments to verify the exoskeleton controller's effectiveness. The active-passive assistance was compared with fully passive and fully active exoskeleton control using the following criteria: (1) post-test survey, (2) load tolerance duration, and (3) computed human torque, power, and metabolic energy expenditure using sEMG signals and inverse dynamic simulation. The experimental outcomes showed that active-passive exoskeletons required less muscular activation torque (50%) from the user and reduced fatigue duration indicators by a factor of 3, compared to fully passive ones.

RBF Sliding Mode Control Method for an Upper Limb Rehabilitation Exoskeleton Based on Intent Recognition

Aiming at the lack of active willingness of patients to participate in the current upper limb exoskeleton rehabilitation training control methods, this study proposed a radial basis function (RBF) sliding mode impedance control method based on surface electromyography (sEMG) to identify the movement intention of upper limb rehabilitation. The proposed control method realizes the process of active and passive rehabilitation training according to the wearer’s movement intention. This study first established a joint angle prediction model based on sEMG for the problem of poor human–machine coupling and used the least-squares support vector machine method (LSSVM) to complete the upper limb joint angle prediction. In addition, in view of the problem of poor compliance in the rehabilitation training process, an adaptive sliding mode controller based on the RBF network approximation system model was proposed. In the process of active training, an impedance model was added based on the position loop control, which could dynamically adjust the motion trajectory according to the interaction force. The experiment results showed that the impedance control method based on the RBF could effectively reduce the interaction force between the human and machine to improve the compliance of the exoskeleton manipulator and achieve the purpose of stabilizing the impedance characteristics of the system.

An EMG-Based Constitutive Law for Force Generation in Skeletal Muscle - Part II: Model Validation On the Ankle Joint Complex

This study evaluates the predictive ability of the skeletal muscle force model derived previously within the ankle joint complex. The model is founded in dimensional analysis, using electromyography and the muscle force-length, force-velocity, and force-frequency curves as inputs. Seventeen subjects (8 males, 9 females) performed five different exercises that activated the primary muscles crossing the ankle joint. Motion capture, force plate, and electromyography data were collected during these exercises. A constant, Km, was calculated for each muscle of each subject using four of the five exercises. The fifth exercise was used to validate the results by treating the moments due to muscle forces as known and all other components in Euler's second law as unknown. While muscle forces cannot be directly validated in vivo, methods can be developed to test these values with reasonable confidence. This study compared moments about the ankle joint due to the calculated muscle forces to the sum of the moments due to all other sources and the kinematic terms in the second Newton-Euler equation of rigid body motion. Average percent errors for each subject ranged from 4.2% to 15.5% with an average percent error across all subjects of 8.2% while maximum percent errors for each subject ranged from 33.3% to 78.0% with an overall average maximum of 52.4%. Future work will examine sensitivity analyses to identify potential simplifications to the model and solution process and will validate the model on a more complex joint.

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

OBJECTIVE

To study the neural control of movement, it is often necessary to estimate how muscles are activated across a variety of behavioral conditions. One approach is to try extracting the underlying neural command signal to muscles by applying latent variable modeling methods to electromyographic (EMG) recordings. However, estimating the latent command signal that underlies muscle activation is challenging due to its complex relation with recorded EMG signals. Common approaches estimate each muscle activation independently or require manual tuning of model hyperparameters to preserve behaviorally-relevant features.

APPROACH

Here, 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 RESULTS

We 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 applied AutoLFADS to monkey forearm muscle activity recorded during an isometric wrist force task. AutoLFADS uncovered previously uncharacterized high-frequency oscillations in the EMG that enhanced the correlation with measured force. The AutoLFADS-inferred estimates of muscle activation were also more closely correlated with simultaneously-recorded motor cortical activity than were other tested approaches.

SIGNIFICANCE

This method leverages dynamical systems modeling and artificial neural networks to provide estimates of muscle activation for multiple muscles. Ultimately, the approach can be used for further studies of multi-muscle coordination and its control by upstream brain areas.

InverseMuscleNET: Alternative Machine Learning Solution to Static Optimization and Inverse Muscle Modeling

InverseMuscleNET, a machine learning model, is proposed as an alternative to static optimization for resolving the redundancy issue in inverse muscle models. A recurrent neural network (RNN) was optimally configured, trained, and tested to estimate the pattern of muscle activation signals. Five biomechanical variables (joint angle, joint velocity, joint acceleration, joint torque, and activation torque) were used as inputs to the RNN. A set of surface electromyography (EMG) signals, experimentally measured around the shoulder joint for flexion/extension, were used to train and validate the RNN model. The obtained machine learning model yields a normalized regression in the range of 88-91% between experimental data and estimated muscle activation. A sequential backward selection algorithm was used as a sensitivity analysis to discover the less dominant inputs. The order of most essential signals to least dominant ones was as follows: joint angle, activation torque, joint torque, joint velocity, and joint acceleration. The RNN model required 0.06 s of the previous biomechanical input signals and 0.01 s of the predicted feedback EMG signals, demonstrating the dynamic temporal relationships of the muscle activation profiles. The proposed approach permits a fast and direct estimation ability instead of iterative solutions for the inverse muscle model. It raises the possibility of integrating such a model in a real-time device for functional rehabilitation and sports evaluation devices with real-time estimation and tracking. This method provides clinicians with a means of estimating EMG activity without an invasive electrode setup.

Assistive Sliding Mode Control of a Rehabilitation Robot with Automatic Weight Adjustment

There are approximately 13 million new stroke cases worldwide each year. Research has shown that robotics can provide practical and efficient solutions for expediting post-stroke patient recovery. This simulation study aimed to design a sliding mode controller (SMC) for an end-effector-based rehabilitation robot. A genetic algorithm (GA) was designed for automatic controller weight adjustment. The optimal weights were obtained by minimizing a cost function comprising the end-effector position error, robot input, robot input-rate, and patient input. To promote safe tuner optimization, a model of the human arm was incorporated to generate the human joint torque. A computed-torque proportional derivative controller (CTPD) was designed for the human arm to approximate the central nervous system (CNS) motor control. This controller was adjusted to simulate rehabilitation effects and patient adaptation. The tuner was optimized for a trajectory tracking task with an assistive high-level control scheme. The simulation results showed lower cost compared to seven manual weight settings. The optimal weights provided good tracking performance and suitable robot inputs. This research provides a framework to conduct various simulations before testing our controller on human subjects. The preliminary results of this study will be used as the starting point for online adaptive controller tuning, which will be examined in our future research.