1,721,126 research outputs found
CEINMS: an OpenSim toolbox to investigate the influence of different neural solutions in predicting muscle excitations and joint moments during dynamic motor tasks
No full textCEINMS: A toolbox to investigate the influence of different neural control solutions on the prediction of muscle excitation and joint moments during dynamic motor tasks
Full text linkPersonalized neuromusculoskeletal (NMS) models can represent the neurological, physiological, and anatomical characteristics of an individual and can be used to estimate the forces generated inside the human body. Currently, publicly available software to calculate muscle forces are restricted to static and dynamic optimisation methods, or limited to isometric tasks only. We have created and made freely available for the research community the Calibrated EMG-Informed NMS Modelling Toolbox (CEINMS), an OpenSim plug-in that enables investigators to predict different neural control solutions for the same musculoskeletal geometry and measured movements. CEINMS comprises EMG-driven and EMG-informed algorithms that have been previously published and tested. It operates on dynamic skeletal models possessing any number of degrees of freedom and musculotendon units and can be calibrated to the individual to predict measured joint moments and EMG patterns. In this paper we describe the components of CEINMS and its integration with OpenSim. We then analyse how EMG-driven, EMG-assisted, and static optimisation neural control solutions affect the estimated joint moments, muscle forces, and muscle excitations, including muscle co-contraction
MRI-based parallel mechanisms to model subject-specific joint kinematics
No full textSubject-specific musculoskeletal (MSK) computer models can estimate muscle and joint articular forces, enabling the identification of mechanical factors causing joint injury or disease. MSK models include models of skeletal anatomy and joint kinematics that can be created in OpenSim [1]. These models are generic and undergo simple linear scaling to a subject using markers from motion analysis. However, generic scaled MSK models produce less accurate estimates of measured knee articular forces compared to those that are subject-specific [2]. Thus, methods are needed to readily create subject-specific models. Passive tibiofemoral (TFJ), patellofemoral (PFJ) and talocrural (TAJ) kinematics measured in cadavers are well predicted using 3D parallel mechanisms [3,4]. These models integrate the cadaver’s measured bone, ligament and tendon geometries to constrain the joints’ degrees of freedom (DOFs). Using TFJ flexion angle as input, TFJ and PFJ models estimate the tibia’s and patella’s remaining Flex-Extension (FE), Abd-Adduction (AA), Int-External (IE) rotations and Ant-Posterior (AP), Prox-Distal (PD) and Med-Lateral (ML) translations. Similarly, TAJ models use talus flexion angle to estimate kinematics from the other TAJ DOFs. However, these models have only been used in cadavers where the kinematics were accurately measured and used to tune the models geometrical parameters. We aimed to use MRI images of in vivo lower limb bones, cartilages and ligaments to create models’ geometrical parameters. Using previously described mechanisms [3,4] we estimated subject-specific kinematics for use in OpenSim. However, without measured kinematics to tune the model, we created specialized algorithms to solve the mechanisms and then compared the results with those from cadaveric studies
Multiscale musculoskeletal modelling, data–model fusion and electromyography-informed modeling
No full textThis paper proposes methods and technologies that advance the state of the art for modelling the musculoskeletal system across the spatial and temporal scales; and storing these using efficient ontologies and tools. We present population-based modelling as an efficient method to rapidly generate individual morphology from only a few measurements and to learn from the ever-increasing supply of imaging data available. We present multiscale methods for continuum muscle and bone models; and efficient mechanostatistical methods, both continuum and particle-based, to bridge the scales. Finally, we examine both the importance that muscles play in bone remodelling stimuli and the latest muscle force prediction methods that use electromyography-assisted modelling techniques to compute musculoskeletal forces that best reflect the underlying neuromuscular activity. Our proposal is that, in order to have a clinically relevant virtual physiological human, (i) bone and muscle mechanics must be considered together; (ii) models should be trained on population data to permit rapid generation and use underlying principal modes that describe both muscle patterns and morphology; and (iii) these tools need to be available in an open-source repository so that the scientific community may use, personalize and contribute to the database of models.No Full Tex
Shape Modelling of Muscle Differences in Paediatric Cerebral Palsy
No full textCerebral palsy (CP) is a neuromusculoskeletal disorder caused by a onetime neural lesion of the central nervous system which occurs in utero or soon after birth. Although the brain lesion is static, the associated musculoskeletal deficits progressively worsen as people with the disorder age. The
progressive musculoskeletal deficits of CP makes it the most common cause
of disability in paediatric populations and impacts the lives of children
throughout their lifespans. However, the clinical interventions in CP are
centred around childhood because it is an accelerated period of growth and
development. CP is heterogeneous and its symptoms may vary from person
to person. The emerging focus of personalised medicine has the potential
to solve this issue, as techniques will increasingly be able to diagnose and
treat CP using bespoke interventions. With this in mind, there is a need to
characterise the person-specific characteristics of the bones and muscles of
young people with CP.
Statistical shape modelling (SSM) is a computational method that takes
advantage of medical imaging such as MRI and Diffusion Tensor Imaging
(DTI); and is based on dimensionality reduction methods such as principal
component analysis (PCA). SSM is capable of identifying dominant variations
of morphology (size and shape) and 3D muscle fibre orientations. The
dominant variations of morphology and 3D fibre orientations are presented
as independent modes of variations (or principal components), with variance
across a population. SSM has been used to characterise variations of musculoskeletal structures within and between populations.
Parameters of lower limb muscles including muscle volume, cross-sectional
area, pennation angle and fascicle length have been extensively investigated
in children and adolescents with CP compared to typically developing children
and adolescents. Muscle volume was reported to be smaller resulting
in weaker muscle and impaired movements in children and adolescents with
CP. To my knowledge, SSM has never been used to quantify variations of
morphology and 3D fibre orientations and their progression over a specific
period of time in lower limb muscles of children and adolescents with CP,
encouraging a further investigation of this area of study.
In this thesis, SSM was used to construct a de novo workflow to investigate
the morphology of the soleus muscle, the morphology and 3D fibre
orientations of the medial gastrocnemius muscle, and the progression of
morphology and 3D fibre orientations of the medial gastrocnemius muscle
over twelve months in children and adolescents with CP compared to typically
developing children and adolescents. Children with CP displayed a
distinct morphology of the soleus muscle, compared to typically developed
(TD) children with a superior-inferior shift of the broad central region of the
muscle. The medial gastrocnemius presented similar morphological variation,
but it displayed localised variation of the 3D fibre orientations between
the CP and TD cohorts. The CP cohort also experienced a reduced growth
rate of the medial gastrocnemius muscle compared to the TD cohort. Fibre
architecture, defined using 3D pennation angles, was also different between
cohorts, with a greater rate of variations in pennation angle over twelve
months in the CP cohort. The variance of each principal component differs
within the CP cohort at different time points, indicating heterogeneous
growth across participants in the CP cohort.
In conclusion, SSM offered a means to quantify the variations of shape and
3D fibre orientations of the soleus and medial gastrocnemius muscles from
both cross-sectional and longitudinal perspectives in paediatric CP. A quantitative
description of muscle shape and 3D fibre orientations will lend an indepth
knowledge of muscle differences in CP and will contribute to patientspecific modelling and targeted therapeutic interventions for the condition
Accuracy and Outcomes of a Novel Robotic-Arm Assisted System for Total Knee Arthroplasty
No full textRobotic-assisted surgical systems for total knee arthroplasty (TKA) have gained momentum in the last decade to address the challenges faced by conventional instrumentation. The ROSA (Robotic Surgical Assistant) Knee System is a novel semi-active collaborative platform with a robotic arm that assists in the placement of a cutting guide to optimise bone resections and implant positioning. Despite early studies and emerging literature demonstrating the improved accuracy of this robotic system in TKA, further high quality studies are needed to explore the value of utilizing the ROSA robotic platform in the TKA workflow and its impact on clinical outcomes. The aims of this thesis were to investigate and explore the learning curve, accuracy of implant positioning, and early clinical outcomes associated with the ROSA Knee System in primary TKA. A systematic review summarised the current literature associated with the ROSA Knee System for TKA in terms of the accuracy of implant positioning and clinical outcomes. A retrospective study investigated the learning curve associated with the introduction of this robotic system. A prospective case series investigated the robotic system’s in vivo accuracy and precision in implant positioning and limb alignment throughout the TKA workflow. Finally, a prospective pragmatic cohort study explored the early functional outcomes between patients undergoing robotic-assisted TKA and conventional TKA using wearable sensor technology and patient-reported outcome measures. The ROSA robotic platform for primary TKA was associated with a short learning curve for operative time and increased accuracy and precision of implant positioning in most radiographic parameters with few outliers through the platform’s workflow. In addition, ROSA-assisted TKA was associated with earlier functional recovery and improved IMU-based outcomes in the first six weeks of surgery compared to conventional TKA, which were not detected by PROMs at various post-operative timepoints.
The longer term clinical significance of these findings and survivorship of TKA using the ROSA Knee System are yet to be determined, but the addition of this technology to assist in TKA procedures has been shown to have both patient and surgeon benefits
Assessing Regional Tibial Bone Strength in the Real-World Setting
No full textTibial fatigue fractures are common overuse injuries in sports and military training, caused by microdamage accumulation from repetitive loading. Assessing tibial bone strength can help monitor the progression of fatigue fractures, enabling coaches and rehabilitation specialists to develop scientifically informed training and rehabilitation plans, thereby reducing the risk of injury. However, traditional methods for evaluating bone strength, such as in vivo measurements or finite element analysis (FEA) are limited by their invasiveness or reliance on specialized equipment, making them difficult to apply in real sports environments. This study proposes a theoretical framework for predicting the time-series stress and strain distribution in the middle and distal tibia during various movements. The framework requires only a single ankle-mounted inertial measurement unit (IMU) and easily measurable tibial anthropometry. The proposed framework integrates machine learning (ML), statistical shape model (SSM), and FEA. First, IMU data are used as boundary conditions to predict the kinematics and kinetics of the ankle and knee joints through a random forest model. Second, a personalized tibial cortical bone model is generated using SSM, adjusting morphological features based on tibial length. Finally, a partial least squares regression (PLSR) model predicts regional stress and strain distributions in the tibia. This method provides a non-invasive and portable solution for monitoring tibial loading in real-world settings, offering potential for injury prevention and clinical applications
Different Strokes for Different Folks : Patient-Specific Gait Modelling and Post-Stroke Rehabilitation
No full textOver the course of a doctoral project, Duncan Bakke completed a body of work
on improving and validating gait modelling, and applying those principles to human
gait studies designed to further explore (and develop novel interventions to
help rehabilitate) post-stroke hemiparetic gait.
It was demonstrated that the MAPClient Lower Limb Scaling Workflow had
higher inter-researcher repeatability than commonly-used linear scaling techniques.
This workflow was used to create all gait models reported in this thesis.
It was demonstrated that the omission of marker clusters on the thigh during
gait capture and inverse kinematics do not alter the resultant joint angle estimations
by a larger magnitude than the uncertainty that the soft tissue artefact of the
thigh introduces. Thigh markers went unused (or unweighted) in all kinematic
analyses reported in this thesis.
A novel haptic biofeedback system based on real-time estimation of peak ani
kle moment was proved to be effective in increasing ankle push-off magnitude in
an able-bodied population.
The same feedback system was pilot-trialled with post-stroke participants,
with results suggesting its use is justified in augmenting post-stroke gait retraining.
Post-stroke gait parameters measured at both self-selected and fast speeds
were used to train unsupervised machine learning categorisation tools called selforganising
maps (SOMs), which correlated with various traditional measures in
a varying manner. One such SOM successfully categorised all post-stroke participants
who attempted high-magnitude ankle push-off feedback thresholds; this
SOM was based on the average knee moment of the non-paretic limb at self-selected
speed.
A novel method of generating synthetic gait data using principal component
analysis (PCA) and convex combination of n-dimensional samples was developed
to use alongside the machine learning approach above
Neural circuitry in motion: Computational modelling of motor cortical dynamics to understand human movement
No full textMovement of the human body arises from a complex interaction between the activity
of the nervous system and the biomechanics of the musculoskeletal system. How
neurons of the brain and spinal cord produce purposeful, coordinated movements is
a fundamental question in neuroscience and motor control research. Computational
models offer a valuable tool to test theories and mechanisms underlying the challenge
of motor control. The primary motor cortex is the area of the brain with the most
direct influence and connections to producing voluntary activity in limbs. However, a
consensus on the link between motor cortex activity and the production of muscle
activity remains elusive.
Spiking neural networks (SNNs) represent the biological mechanisms of communication
between neurons via action potentials. A SNN model containing over 38,000
neurons and 160 million synapses was developed to represent a 1 mm2 surface area
of the motor cortex. The neural network model used realistic, physiological parameters
and connectivity to replicate the spontaneous firing behaviour of populations of
neurons in the motor cortex. A model of transcranial magnetic stimulation (TMS)
was also applied to the motor cortex model, resulting in the generation of highfrequency
(I-waves) at the spinal cord level, matching experimental observations. In
addition, the model was coupled with a musculoskeletal model of the upper limb
to simulate muscle contraction and multi-body dynamics, via a simple spinal cord
circuit controlling extensor and flexor muscles, showing the feasibility of coupled
brain-body models.
The motor cortex model is presented within a larger framework of modelling the
connection from brain to muscle that incorporates feedback pathways including
muscle spindles and Golgi-tendon organs, in addition to detailed muscle models. The
framework of this computational modelling approach uses multi-scale, multi-modal
modelling fitted where possible to experimental data to enable the observation of
emergent patterns of behaviour within the motor system. Using interdisciplinary
computational models to understand the neuromusculoskeletal system is widely applicable
and can be developed in conjunction with experimental work and hypotheses
of motor control
EMG-Informed Estimation of Human Walking Dynamics for Assistive Robots
No full textThe demand for gait rehabilitation is increasing globally, and conventional rehabilitation practices
cannot cope with this increase. Robotic-based rehabilitation and assistive robots are alternative
solutions for gait rehabilitation, but challenges remain to bring this technology into the clinic.
Assistive robots would ideally provide a personalized level of assistance based on an individual’s
physical and neurological condition, biomechanics, and muscular fatigue. An assistive robot should
also produce a smooth movement based on the user's motion intention. Thus, a prediction of
motion intention and corresponding adjustment of the robot actuator forces are the fundamental
requirements for a controller of an assistive robot.
Electromyography (EMG) signals have been used widely for motion intention estimation.
However, most EMG-based models are subject or task-specific, requiring complex calibration.
Creating an accurate, EMG-based motion estimation model which is generalizable across
individuals and experimental conditions is a major challenge and was the goal of this thesis.
The chosen application was to predict the motion and moments of the ankle joint during a range
of different walking conditions. As such, a set of experiments was designed to collect motionrelated
data from 10 individuals during a wide range of activities. Initially, an artificial neural
network was designed to predict ankle moment during constant speed walking based on a list of
input time series, including the EMG signals of four muscles from each leg and ankle kinematics.
The results helped find the list of most important input time series and the length of information
required for ankle moment prediction at each step. Next, a machine learning approach was
explored, including feature extraction and selection from the input time series. The selected list of
features optimized the model training process and was generalizable across individuals to estimate
the ankle moment during constant-speed walking. Exploring the influence of the training dataset
on model predictions at various walking speeds was the focus of the next step. It was discovered
that training the model on acceleration data from 0.5 m/s to 2.5 m/s enabled the model to predict
ankle moment during walking at any speed in this range. Random forest, backpropagation neural
networks, and linear regression were compared as potential predictive models, with the random
forest having the best predictions across walking speeds.
In addition to making the model compatible with a range of activities, the desire was to update the
model parameters based on the error between the model output and target value regardless of the
training dataset. An adaptive model was developed and implemented to predict ankle angle during
walking at four different speeds and three inclines to achieve this. The base model was initially
trained on data from level ground walking on one participant at 1 m/s. The simplicity of the model structure made it possible to update the parameters whenever there was an error between the
predicted and actual ankle angle with less than a 30 ms time delay. The RMSE of the model for all
of the test conditions was less than 5 degrees across the cohort of ten individuals (including nine
unseen individuals). Continuous and accurate prediction of joint kinematics under different walking
conditions and multiple individuals promises a stable and reliable control for wearable assistive
robots, thus achieving the goal of the thesis
- …
