Using Hill-Type Muscle Models and EMG Data in a Forward Dynamic Analysis of Joint Moment

2003 ◽  
Vol 03 (02) ◽  
pp. 169-186 ◽  
Author(s):  
Richard Heine ◽  
Kurt Manal ◽  
Thomas S. Buchanan
Keyword(s):  
Muscle Activation ◽  
Joint Moment ◽  
Model Parameters ◽  
Time Varying ◽  
Muscle Forces ◽  
Joint Moments ◽  
Delay Term ◽  
Activation Dynamics ◽  
Moment Models ◽  
Muscle Models

There has been considerable interest in estimating muscle forces and joint moments from EMG signals, but most approaches have not been very successful. This is largely because robust models of muscle activation dynamics, Hill-type muscle contraction dynamics, and musculoskeletal geometry are generally not included. Here we present a model which includes these sub-models and we determine which model parameters are most important. The models abilities to predict joint moments about the human elbow during time-varying isometric tasks were examined. Inputs to the models were EMGs from eight muscles. Joint moment was the output, which was compared with the measured moment. Models varied in complexity, having up to 59 adjustable parameters. It was found that a seven adjustable parameter model could adequately estimate time-varying joint moments without substantial sacrifice in performance. The key parameters that were fit for each subject were two global gain factors, a time delay term, a non-linear EMG-force term, two muscle activation terms, and a term for skewing the length-tension curve with muscle activation. This approach offers advantages over optimization-based methods for estimating individual muscle forces. Most importantly, it accounts for the way muscles are activated, which makes it potentially powerful to evaluate patients with pathologies.

scholarly journals Neuromusculoskeletal Modeling: Estimation of Muscle Forces and Joint Moments and Movements from Measurements of Neural Command

2004 ◽  
Vol 20 (4) ◽  
pp. 367-395 ◽  
Author(s):  
Thomas S. Buchanan ◽  
David G. Lloyd ◽  
Kurt Manal ◽  
Thor F. Besier
Keyword(s):  
Muscle Activation ◽  
Equations Of Motion ◽  
Muscle Forces ◽  
Joint Moments ◽  
Neural Signals ◽  
Knee Kinetics ◽  
Time Varying Parameter ◽  
Activation Dynamics ◽  
Muscle Activations ◽  
Neuromusculoskeletal Modeling

This paper provides an overview of forward dynamic neuromusculoskeletal modeling. The aim of such models is to estimate or predict muscle forces, joint moments, and/or joint kinematics from neural signals. This is a four-step process. In the first step,muscle activation dynamicsgovern the transformation from the neural signal to a measure of muscle activation—a time varying parameter between 0 and 1. In the second step,muscle contraction dynamicscharacterize how muscle activations are transformed into muscle forces. The third step requires a model of themusculoskeletal geometryto transform muscle forces to joint moments. Finally, theequations of motionallow joint moments to be transformed into joint movements. Each step involves complex nonlinear relationships. The focus of this paper is on the details involved in the first two steps, since these are the most challenging to the biomechanician. The global process is then explained through applications to the study of predicting isometric elbow moments and dynamic knee kinetics.

Estimation of Muscle Forces About the Ankle During Gait in Healthy and Neurologically Impaired Subjects

2011 ◽  
pp. 320-347
Author(s):  
Daniel N. Bassett ◽  
Joseph D. Gardinier ◽  
Kurt T. Manal ◽  
Thomas S. Buchanan
Keyword(s):  
Muscle Activation ◽  
Inverse Dynamics ◽  
Biomechanical Model ◽  
Forward Dynamics ◽  
Muscle Forces ◽  
Dynamics Model ◽  
Joint Moments ◽  
Neurologically Impaired ◽  
Model Predictions ◽  
Contraction Dynamics

This chapter describes a biomechanical model of the forces about the ankle joint applicable to both unimpaired and neurologically impaired subjects. EMGs and joint kinematics are used as inputs and muscle forces are the outputs. A hybrid modeling approach that uses both forward and inverse dynamics is employed and physiological parameters for the model are tuned for each subject using optimization procedures. The forward dynamics part of the model takes muscle activation and uses Hill-type models of muscle contraction dynamics to estimate muscle forces and the corresponding joint moments. Inverse dynamics is used to calibrate the forward dynamics model predictions of joint moments. In this chapter we will describe how to implement an EMG-driven hybrid forward and inverse dynamics model of the ankle that can be used in healthy and neurologically impaired people.

Using Reinforcement Learning to Estimate Human Joint Moments via EMG Signals or Joint Kinematics: An Alternative Solution to Musculoskeletal-Based Biomechanics

2020 ◽  
Author(s):  
Wen Wu ◽  
Kate Saul ◽  
He (Helen) Huang
Keyword(s):  
Reinforcement Learning ◽  
Correlation Coefficients ◽  
Metacarpophalangeal Joint ◽  
Joint Moment ◽  
Biomechanical Analysis ◽  
Model Parameters ◽  
Optimization Approach ◽  
Joint Moments ◽  
Moment Estimation ◽  
Policy Optimization

Abstract Reinforcement learning (RL) has potential to provide innovative solutions to existing challenges in estimating joint moments in motion analysis, such as kinematic or electromyography (EMG) noise and unknown model parameters. Here we explore feasibility of RL to assist joint moment estimation for biomechanical applications. Forearm and hand kinematics and forearm EMGs from 4 muscles during free finger and wrist movement were collected from six healthy subjects. Using the Proximal Policy Optimization approach, we trained and tested two types of RL agents that estimated joint moment based on measured kinematics or measured EMGs, respectively. To quantify the performance of RL agents, the estimated joint moment was used to drive a forward dynamic model for estimating kinematics, which were then compared with measured kinematics. The results demonstrated that both RL agents can accurately reproduce wrist and metacarpophalangeal joint motion. The correlation coefficients between estimated and measured kinematics, derived from the kinematics-driven agent and subject-specific EMG-driven agents, were 0.98±0.01 and 0.94±0.03 for the wrist, respectively, and were 0.95±0.02 and 0.84±0.06 for the metacarpophalangeal joint, respectively. In addition, a biomechanically reasonable joint moment-angle-EMG relationship (i.e. dependence of joint moment on joint angle and EMG) was predicted using only 15 seconds of collected data. In conclusion, this study serves as a proof of concept that an RL approach can assist in biomechanical analysis and human-machine interface applications by deriving joint moments from kinematic or EMG data.

A New Muscle Activation Dynamics Model, That Simulates the Calcium Kinetics and Incorporates the Role of Store-Operated Calcium Entry Channels, to Enhance the EMG-Driven Hill-type Models

2021 ◽  
Author(s):  
Moemen Hussein ◽  
Said Shebl ◽  
Rehab Elnemr ◽  
Hesham Elkaranshawy
Keyword(s):  
Muscle Activation ◽  
Computational Cost ◽  
Calcium Entry ◽  
Type Model ◽  
Model Parameters ◽  
Dynamics Model ◽  
Force Relation ◽  
Activation Dynamics ◽  
Store Operated Calcium Entry

Abstract Hill-type models are frequently used in biomechanical simulations. They are attractive for their low computational cost and close relation to commonly measured musculotendon parameters. Still, more attention is needed to improve the activation dynamics of the model specifically because of the nonlinearity observed in the EMG-Force relation. Moreover, one of the important and practical questions regarding the assessment of the model's performance is how adequately can the model simulate any fundamental type of human movement without modifying model parameters for different tasks? This paper tries to answer this question by proposing a simple physiologically based activation dynamics model. The model describes the ?kinetics of the calcium dynamics while activating and deactivating the muscle contraction process. Hence, it allowed simulating the recently discovered role of store-operated calcium entry (SOCE) channels as immediate counter-flux to calcium loss across the tubular system during excitation-contraction coupling. By comparing the ability to fit experimental data without readjusting the parameters, the proposed model has proven to have more steady performance than phenomenologically based models through different submaximal isometric contraction levels. This model indicates that more physiological insights is key for improving Hill-type model performance.

A phase-reduced neuro-mechanical model for insect locomotion: feed-forward stability and proprioceptive feedback

2010 ◽  
Vol 368 (1930) ◽  
pp. 5087-5104 ◽  
Author(s):  
J. Proctor ◽  
R. P. Kukillaya ◽  
P. Holmes
Keyword(s):  
Degrees Of Freedom ◽  
Muscle Activation ◽  
Phase Oscillators ◽  
Joint Torques ◽  
Insect Locomotion ◽  
Phase Reduction ◽  
Activation Dynamics ◽  
Order Of Magnitude ◽  
Sensory Pathway ◽  
Muscle Models

In earlier work, we have developed an integrated model for insect locomotion that includes a central pattern generator (CPG), nonlinear muscles, hexapedal geometry and a representative proprioceptive sensory pathway. Here, we employ phase reduction and averaging theory to replace 264 ordinary differential equations (ODEs), describing bursting neurons in the CPG, their synaptic connections to motoneurons, muscle activation dynamics and sensory neurons, with 24 one-dimensional phase oscillators that describe motoneuronal activation of agonist–antagonist muscle pairs driving the jointed legs. Reflexive feedback is represented by stereotypical spike trains with rates proportional to joint torques, which change phase relationships among the motoneuronal oscillators. Restriction to the horizontal plane, neglect of leg mass and use of Hill-type muscle models yield a biomechanical body–limb system with only three degrees of freedom, and the resulting hybrid dynamical system involves 30 ODEs: reduction by an order of magnitude. We show that this reduced model captures the dynamics of unperturbed gaits and the effects of an impulsive perturbation as accurately as the original one. Moreover, the phase response and coupling functions provide an improved understanding of reflexive feedback mechanisms.

A Model of Load Sharing Between Muscles and Soft Tissues at the Human Knee During Static Tasks

1996 ◽  
Vol 118 (3) ◽  
pp. 367-376 ◽  
Author(s):  
D. G. Lloyd ◽  
T. S. Buchanan
Keyword(s):  
Soft Tissue ◽  
External Load ◽  
Muscle Activation ◽  
Soft Tissues ◽  
Nonlinear Least Squares ◽  
Joint Moment ◽  
Muscle Forces ◽  
Least Squares Fit ◽  
Average Maximum ◽  
Flexion And Extension

In this study, we had subjects voluntarily generate various forces in a transverse plane just above their ankles. The contributions of their muscles and soft tissues to the support of the total external knee joint moment were determined by analyzing the experimental data using a biomechanical model of the knee. In this model, muscle forces were estimated using the recorded EMGs. To account for subject variability, various muscle parameters were adjusted using a nonlinear least-squares fit of the model’s estimated flexion and extension joint moments to those recorded externally. Using the estimated muscle forces, the contributions from the muscles and other soft tissues to the total joint moment were obtained. The results showed that muscles were primarily used to support flexion and extension loads at the knee, but in so doing, were able to support some part of the varus or valgus loads. However, soft tissue loading was still required. Soft tissues supported up to an average maximum of 83 percent of the external load in pure varus and valgus. Soft tissue loading in pure varus and valgus was less than 100 percent of the external load as the muscles, on average, were able to support 17 percent of the external load. This muscle support was by virtue of muscle cocontraction and/or specific muscle activation.

scholarly journals Comparative Sensitivity Analysis of Muscle Activation Dynamics

2015 ◽  
Vol 2015 ◽  
pp. 1-16 ◽  
Author(s):  
Robert Rockenfeller ◽  
Michael Günther ◽  
Syn Schmitt ◽  
Thomas Götz
Keyword(s):  
Sensitivity Analysis ◽  
Muscle Activation ◽  
Striated Muscle ◽  
Initial Conditions ◽  
Muscle Length ◽  
Model Parameters ◽  
Biomechanical Models ◽  
Activation Dynamics ◽  
Broad Variety ◽  
Parameter Values

We mathematically compared two models of mammalian striated muscle activation dynamics proposed by Hatze and Zajac. Both models are representative for a broad variety of biomechanical models formulated as ordinary differential equations (ODEs). These models incorporate parameters that directly represent known physiological properties. Other parameters have been introduced to reproduce empirical observations. We used sensitivity analysis to investigate the influence of model parameters on the ODE solutions. In addition, we expanded an existing approach to treating initial conditions as parameters and to calculating second-order sensitivities. Furthermore, we used a global sensitivity analysis approach to include finite ranges of parameter values. Hence, a theoretician striving for model reduction could use the method for identifying particularly low sensitivities to detect superfluous parameters. An experimenter could use it for identifying particularly high sensitivities to improve parameter estimation. Hatze’s nonlinear model incorporates some parameters to which activation dynamics is clearly more sensitive than to any parameter in Zajac’s linear model. Other than Zajac’s model, Hatze’s model can, however, reproduce measured shifts in optimal muscle length with varied muscle activity. Accordingly we extracted a specific parameter set for Hatze’s model that combines best with a particular muscle force-length relation.

Role of Trunk Muscles in Generating Follower Load in the Lumbar Spine of Neutral Standing Posture

2008 ◽  
Vol 130 (4) ◽  
Author(s):  
Kyungsoo Kim ◽  
Yoon Hyuk Kim
Keyword(s):  
Lumbar Spine ◽  
Shear Force ◽  
Muscle Activation ◽  
Optimization Problem ◽  
Load Condition ◽  
Joint Moment ◽  
Trunk Muscles ◽  
Follower Load ◽  
Muscle Forces

Recently, experimental results have demonstrated that the load carrying capacity of the human spine substantially increases under the follower load condition. Thus, it is essential to prove that a follower load can be generated in vivo by activating the appropriate muscles in order to demonstrate the possibility that the stability of the spinal column could be maintained through a follower load mechanism. The aim of this study was to analyze the coordination of the trunk muscles in order to understand the role of the muscles in generating the follower load. A three-dimensional finite element model of the lumbar spine was developed from T12 to S1 and 117 pairs of trunk muscles (58 pairs of superficial muscles and 59 pairs of deep muscles) were considered. The follower load concept was mathematically represented as an optimization problem. The muscle forces required to generate the follower load were predicted by solving the optimization problem. The corresponding displacements and rotations at all nodes were estimated along with the follower forces, shear forces, and joint moments acting on those nodes. In addition, the muscle forces and the corresponding responses were investigated when the activations of the deep muscles or the superficial muscles were restricted to 75% of the maximum activation, respectively. Significantly larger numbers of deep muscles were involved in the generation of the follower load than the number of superficial muscles, regardless of the restriction on muscle activation. The shear force and the resultant joint moment are more influenced by the change in muscle activation in the superficial muscles. A larger number of deep trunk muscles were activated in order to maintain the spinal posture in the lumbar spine. In addition, the deep muscles have a larger capability to reduce the shear force and the resultant joint moment with respect to the perturbation of the external load or muscle fatigue compared to the superficial muscles.

The Comparison of Muscle Forces Derived From Different Muscle Models

2005 ◽  
Author(s):  
Miloslav Vilimek
Keyword(s):  
Elbow Joint ◽  
Inverse Dynamics ◽  
Optimization Techniques ◽  
Muscle Forces ◽  
Bone Implant ◽  
Optimization Criteria ◽  
Functional Activities ◽  
Flexion Extension ◽  
Activation Dynamics ◽  
Muscle Models

This study, investigated the accuracy, practicality, and sensitivity of several different methods of calculating muscle forces during functional activities in humans. The upper extremity dynamic system was chosen, where the movement flexion / extension elbow joint was studied. The redundant mechanisms were solved using optimization criteria with and without models of individual muscles according to their active and passive properties. Exploration of the control problem for the redundant elbow system was performed using muscle models with and without tendon and activation dynamics. Comparisons with known movements solved by inverse dynamics approach and optimization techniques, provided similar results across to all optimization criteria. Moreover, if muscle models with active and passive properties are included in these analyses, it is relatively easy to calculate muscle forces of both agonists and antagonists. These approaches may be used to provide input data for dynamic FEM stress analysis of bones and bone-implant systems.

Adaptive control of time-varying non-linear plants by speed-gradient algorithms

2019 ◽  
pp. 37-44 ◽  
Author(s):  
O. P. Tomchina ◽  
D. N. Polyakhov ◽  
O. I. Tokareva ◽  
A. L. Fradkov
Keyword(s):  
Adaptive Control ◽  
Adaptive Systems ◽  
Reference Model ◽  
Maximum Deviation ◽  
Model Parameters ◽  
Time Varying ◽  
System Solution ◽  
Non Linear ◽  
Initial Uncertainty ◽  
Speed Gradient

Introduction: The motion of many real world systems is described by essentially non-linear and non-stationary models. A number of approaches to the control of such plants are based on constructing an internal model of non-stationarity. However, the non-stationarity model parameters can vary widely, leading to more errors. It is only assumed in this paper that the change rate of the object parameters is limited, while the initial uncertainty can be quite large.Purpose: Analysis of adaptive control algorithms for non-linear and time-varying systems with an explicit reference model, synthesized by the speed gradient method.Results: An estimate was obtained for the maximum deviation of a closed-loop system solution from the reference model solution. It is shown that with sufficiently slow changes in the parameters and a small initial uncertainty, the limit error in the system can be made arbitrarily small. Systems designed by the direct approach and systems based on the identification approach are both considered. The procedures for the synthesis of an adaptive regulator and analysis of the synthesized system are illustrated by an example.Practical relevance: The obtained results allow us to build and analyze a broad class of adaptive systems with reference models under non-stationary conditions.

Sign in / Sign up

Export Citation Format

Share Document