See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/261563590

Biomechanics of Musculoskeletal System and

Its Biomimetic Implications: A Review
Article in Journal of Bionic Engineering April 2014
Impact Factor: 1.63 DOI: 10.1016/S1672-6529(14)60033-0

CITATIONS

READS

424

3 authors, including:
Lei Ren

Zhihui Qian

The University of Manchester

Jilin University

94 PUBLICATIONS 394 CITATIONS

12 PUBLICATIONS 42 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,

letting you access and read them immediately.

SEE PROFILE

Available from: Zhihui Qian

Retrieved on: 24 June 2016

Journal of Bionic Engineering 11 (2014) 159175

Biomechanics of Musculoskeletal System and Its Biomimetic

Implications: A Review
Lei Ren1,2, Zhihui Qian2, Luquan Ren2
1. School of Mechanical, Aerospace and Civil Engineering, University of Manchester, Manchester M13 9PL, UK
2. Key Laboratory of Bionic Engineering (Ministry of Education, China), Jilin University, Changchun 130022, P. R. China

Abstract
Biological musculoskeletal system (MSK), composed of numerous bones, cartilages, skeletal muscles, tendons, ligaments
etc., provides form, support, movement and stability for human or animal body. As the result of million years of selection and
evolution, the biological MSK evolves to be a nearly perfect mechanical mechanism to support and transport the human or
animal body, and would provide enormously rich resources to inspire engineers to innovate new technology and methodology to
develop robots and mechanisms as effective and economical as the biological systems. This paper provides a general review of
the current status of musculoskeletal biomechanics studies using both experimental and computational methods. This includes
the use of the latest three-dimensional motion analysis systems, various medical imaging modalities, and also the advanced
rigid-body and continuum mechanics musculoskeletal modelling techniques. Afterwards, several representative biomimetic
studies based on ideas and concepts inspired from the structures and biomechanical functions of the biological MSK are discussed. Finally, the major challenges and also the future research directions in musculoskeletal biomechanics and its biomimetic
studies are proposed.
Keywords: musculoskeletal system, biomechanics, multi-scale, biomimetics, biologically inspired robots and mechanisms
Copyright 2014, Jilin University. Published by Elsevier Limited and Science Press. All rights reserved.
doi: 10.1016/S1672-6529(14)60033-0

1 Introduction
A musculoskeletal system (MSK) is a biological
system composed of bones, cartilages, skeletal muscles,
tendons, ligaments and other connective tissues (see
Fig. 1). The major function of the MSK is to provide
form, support, movement and stability for the human or
animal body. The MSK can be roughly considered to
have two constituent sub-systems: the skeletal system
and the muscular system.
The skeletal system consists of all the bones in the
body and also the connecting tissues, e.g. cartilages and
ligaments. The skeletal system provides the fundamental
framework for body shape and load bearing, and also
protects internal organs, e.g. brain, heart, lungs and liver
etc., from external impacts. In the skeletal system, bones
are connected to each other by joints, which provide
articulations in MSK. The most common type of joint is
synovial joint, which consists of fibrous connective
tissue capsule (ligaments) and the periosteum of the
Corresponding author: Lei Ren
E-mail: lei.ren@manchester.ac.uk

connecting bones lubricated by synovial fluid inside of

the joint.
The muscular system is the prime mover of human
or animal body. For humans, there are approximately

Fig. 1 The musculoskeletal system of human body[1].

160

Journal of Bionic Engineering (2014) Vol.11 No.2

over 430 skeletal muscle groups accounting for up to

40% of body weight[2]. The skeletal muscles are made up
of hundreds, or even thousands of muscle fibers, which
range in thickness from approximately 10 m to 100 m
and in length from about 1 cm to 30 cm[2]. Skeletal
muscles are also arranged in layers over the bones, and
they are normally attached to bones through tendons so
that the forces generated by the contractile elements of
the muscle fibers can drive body motions.
As the result of million years of selection and evolution, the biological MSK evolves to be a highly efficient and economic mechanical mechanism to support
and transport the human or animal body with many extraordinary characteristics compared to their man-made
counterparts. For example, cheetah is known as the
fastest living land quadrupedal animal, which can reach
to a speed of 29 ms1[3]. Some special features of cheetahs MSK have been reported probably contributing to
attaining such a high speed. For example, their divergent
talar ridges may help to avoid limb interference in the
aerial phase of galloping. Their long hindlimb bones
may potentially assist them in making large stride length.
Moreover, their particularly large psoas muscles may
help them to rapidly protract the hindlimbs and also to
resist pitching moments around the hip during accelerating[4]. Another example is horses, a representative
athletic ungulate land animal with excellent locomotor
capacity. Their third metacarpus bone has a small hole
where blood vessels enter the bone[5]. As common
knowledge from mechanical engineering, holes normally weaken structures by increasing dramatically the
stresses near the holes as a result of stress concentration[5]
(see Fig. 2). However, the holes at the third metacarpus

bones of horses do not appear to cause bone fractures

even for racing horses. This is probably due to an increased compliance near the foramen[5]. The sharp discontinuity in geometry due to the hole is softened by an
embedded compliant region[5]. A reinforcing ring with
increased stiffness, together with the ring of lamellar
bone along the foramens inner edges, might help to
reduce the possibility of cracking[5]. The special configuration of the hole helps to move the highest stresses
away from the foramen to regions with higher material
strength[5] (see Fig. 2).
The human foot complex is another good example
of highly efficient mechanical mechanism from the
biological world. The human foot is a complicated
structure comprising numerous bones, muscles, tendons,
ligaments, synovial joints and other tissues. It has been
found recently that such a small body component delivers multiple critical biomechanical functions in attenuating ground impact, supporting body against gravity, maintaining locomotor stability, generating and
transmitting propulsive power during locomotion[68].
The fascinating structure and characteristics of the
biological MSK, which has been optimally selected after
million years evolution, would provide enormously rich
resources to inspire engineers to innovate new technology and methodology to develop mechanisms and machineries as effective and economical as the biological
systems. For example, legged locomotion, as a biological transportation solution over rough terrain, has been
attracting intensive researches from mechanical engineering and robotics field[913]. This may greatly facilitate the development of legged robots with high agility,
stability and energy efficiency.

= 1.5 MPa

Fig. 2 The foramen in horse third metacarpus bone and its stress analysis[5].

Ren et al.: Biomechanics of Musculoskeletal System and Its Biomimetic Implications: A Review

This paper starts with a brief introduction of the

MSK and the biomechanics of its constituent components. Afterwards, a general review of the current status
of the MSK biomechanics, including both experimental
and computational studies, is provided. Several representative biomimetic examples based on inspired ideas
from MSK biomechanics are discussed. Finally, the
major challenges and also the future research directions
in MSK biomechanics and its biomimetic studies are
proposed.

2 Experimental studies of MSK biomechanics

2.1 Motion capture
In the past decades, motion capture technique has
been increasingly used in recording the two-dimensional
(2D) or three-dimensional (3D) human/animal motion,
which is characterized by the time histories of segmental
or joint angles. Generally, optoelectronic motion analysis systems are employed by using infrared camera arrays to track the positions of active or passive markers
placed on the body segments of interest[1422] (see Fig. 3).
Although the optoelectronic motion capture technique
has now been widely used in human/animal movement
studies, the collected data normally suffer from some
technical problems, e.g. skin artifact (due to the relative
movement of skin mounted markers with respect to the
underlying bones)[16,2325], light reflections and marker
occlusions[26]. Markerless motion tracking method based
on computer vision provides a new promising motion
capture technique[27]. Currently, the markerless systems
can work with large, obvious movements, while the
measure of more subtle movements still remains chalHead

Torso

Humerus
Forearm+hand

Thigh
Pelvis
Shank

X
K
Fgl

161

lenging.
To complement the motion capture, many biomechanics labs are also equipped with force sensing devices to record the simultaneous kinematic and kinetic
data associated with human/animal motions. 3D force
platforms are normally used to measure the ground reaction forces and moments induced by human/animal
motions. Integrated with the simultaneous motion data,
joint kinetic analysis can normally be conducted by
using the inverse dynamics method[17,2830]. Additionally,
pressure plates are often employed to record the foot
pressure distribution during human/animal motions[3134].
In addition to multi-camera systems, portable motion
sensors are also used to capture human/animal body
motions, e.g. inertia sensors (i.e. accelerometers, gyroscopes etc.) and magnetometers etc.[3538]. This offers an
alternative way to measure the human/animal motions
outdoors without the constraints of the indoor equipments[39,40].
2.2 Surface electromyography
Human or animal electromyography signals at different motor activities can be recorded using invasive or
non-invasive methods. Surface electromyography
(sEMG) is a non-invasive technique widely used for
detecting and recording muscle electrical activity that
occurs during muscle contraction and relaxation cycles
by using surface electrodes. The sEMG signal is normally used in musculoskeletal biomechanics studies as
an indicator of the initiation of muscle activation, as an
estimator of the force produced by a contracting muscle,
or as an index of the fatigue occurring within a muscle[28,41,42]. In other words, sEMG signals may contain
information about whether a muscle is active or not, if a
muscle is more or less active, when it is on and/or off,
and also if it fatigues[43]. However, raw sEMG data may
contain mixed electrical signals from multiple muscles
nearby the electrodes and/or noisy signals due to
movement artifact, so dedicated signal processing is
normally needed before useful information can be obtained to interpret the muscle functions[28,44,45].

Foot
K
M gl

K
M gr

K
Fgr

Fig. 3 The three-dimensional whole body model with 13 segments and 12 connecting joints. A specially designed marker
cluster system mounted on plastic plates was used to capture the
segmental motions[17].

2.3 Medical imaging

Medical imaging domains, e.g. Computed Tomography (CT), Magnetic Resonance Imaging (MRI), ultrasound, are very useful techniques to examine the
anatomy, geometry and structure of a MSK. Among

162

Journal of Bionic Engineering (2014) Vol.11 No.2

them, the ultrasound provides an economic noninvasive

imaging technique by transmitting high-frequency
sound waves through a body part. As a handy diagnostic
tool, ultrasound has been widely employed in musculoskeletal biomechanics studies to investigate the in vivo
muscle structure, muscle fiber movement, neuromuscular disorders etc.[4649]. Although superficial muscles
may be easily detected by ultrasound probes, it is difficult to identify individual small muscles when multiple
muscle groups are involved[50]. Additionally, due to the
reflection or absorption of sound by superficial tissue
layers, it is still challenging for ultrasound to detect
deeper muscles especially in the pelvic region or around
the trunk. Since ultrasound has difficulty in penetrating
bones, to examine the structure of bones, or a particular
musculoskeletal complex, e.g. certain joints or body
segments, other imaging modalities, such as CT and/or
MRI, are generally used.

3 Biomechanical MSK modelling

3.1 Rigid body modelling of MSK
Rigid body dynamics is typically used to simulate
human/animal motions by considering the body segments of interest as rigid bodies without any deformations. Over the past decades, numerous rigid body
models have been developed to simulate human/animal
movements or motions of particular body parts with
varying complexity from simple one-segment model
with 1 DOF to complex model with 10 segments and 23
DOFs[17,5166] (see Fig. 4). In those models, the con-

Fig. 4 A large-scale musculoskeletal model with 23 generalized

DoFs and 54 musculotendon actuators[61].

necting joints are typically defined as frictionless

ball-and-socket joints, hinge joints or universal joints.
Whilst the mechanical behaviour of muscles was normally represented using Hill-type models[56,66,67], where
the normalized muscle force of the contractile element is
the product of three independent experimentally measured factors describing the force-length property, the
force-velocity property and the dynamics of neural activation[56,57]. Model simulations can be applied in two
different ways, which are usually referred to as the
forward (direct) dynamics method, and the inverse dynamics method. In the forward dynamics method, the
motions of the segments are determined by integrating
the equations of motion based on predefined joint moment or muscle force/activation data[53,61,65,66,68,69]. In the
inverse dynamics method, the joint forces and moments
are determined based on the measured joint or segment
motion data[17,2830,66].
Biomechanical analysis based on rigid body models
can be applied to a wide range of problems, such as the
assessment of the effect of tendon transfer surgeries[6971],
the investigation of multi-segment interaction[72,73], the
musculoskeletal performance during walking and
jumping[65,68,7476], the development of neural prostheses[77], the neural control principles of movement[7880],
the effect of musculotendon loss or damage on the
overall joint moment capacity[81], and the effect of load
carriage design on walking performance[82]. Recently,
several software packages have been developed for rigid
body musculoskeletal model construction, simulation
and analysis, e.g. SIMM[69], OpenSim[83], AnyBody[84],
MSMS[77] etc., which normally provide graphical user
interface for general users.
Despite the recent great progress and success, there
are still some major unsolved problems in rigid body
musculoskeletal modelling. Firstly, the anatomical joints
are typically simplified as ideal ball-and-socket or hinge
joints to reduce computational load, which are not realistic representations of most biological joints. In addition, there are still a large amount of works needed to
improve the in vivo representation of musculotendon
mechanics and neural dynamics, especially on a subject-specific basis. Finally, the development of predictive musculoskeletal models, which are capable of predicting kinematic and kinetic variables during motions
with minimal measurement inputs, still remains as a very
challenging task[61,66,85].

Ren et al.: Biomechanics of Musculoskeletal System and Its Biomimetic Implications: A Review

3.2 Continuum mechanics modelling of MSK

Over the past decades, methods based on continuum mechanics, e.g. Finite Element (FE) method,
boundary element method etc., have been increasingly
used to investigate the mechanical behaviour of musculoskeletal structures by considering constituent components as deformable bodies. The FE method is a very
useful numerical tool to handle biological structures
normally with highly non-linear material properties,
irregular geometries and complicated boundary conditions. Recently, the FE method has been widely used in
biomechanical studies of MSK, covering a broad range
of topics, e.g. functions of musculoskeletal complexes,
mechanics of joints and skeletal muscle mechanics etc.
3.2.1 FE modelling of musculoskeletal complexes
The FE method has been particularly useful to investigate the mechanical behaviour of specific musculoskeletal complexes of the human body, normally
comprising numerous bones, joints and soft tissues, e.g.
the spinal column or the ankle-foot complex. Substantial
FE studies have been conducted on the biomechanics of
cervical vertebrae, thoracic vertebrae and lumbar vertebrae. Some earliest works involves the modelling of the
cervical spine and the lumbar vertebral motions[8690].
Saito et al. constructed a spine model to analyze the
prevention of spinal column deformity with oversimplified representation of vertebral geometry and the inter-vertebral joints, which may lead to an unrealistic
assessment of load sharing and stress distributions[91].
The model may be suitable for the study of gross responses of the whole column rather than the local
changes at the individual vertebrae level. Kleinberger
presented a sophisticated model including head and
various spinal components for FE analysis[92]. However,
the model lacked sound representations of the anatomical structure of the spine probably because the major aim
of the study was for crash impact analyses rather than
medical applications. The three-segment lower cervical
spinal unit model constructed by Voo et al. had a reasonable representation of the cervical anatomy based on
CT scans and cryomicrotome anatomical sections[93,94],
which could be extended to the construction of the entire
cervical column. Most of the FE studies of spinal column are based on static analysis[9597]. Although internal
stresses, strains and other biomechanical responses under complex loading conditions could be predicted, they

163

provide very little information about the in vivo condition of the whole column during dynamic motions.
Whereas some FE models consisting of a series of
connected vertebrae could predict the dynamic responses of the spine to external loads[98,99]. Zhang et al.
constructed a detailed cervical spine model (C0-C7)[100].
The predicted biomechanical response of human neck
under physiological loadings, near vertex drop impact
and rear-end impact conditions, were analyzed and
compared with published measurement data, demonstrating potential for future biomedical and traumatic
studies.
The human foot is a very complex structure comprising numerous bones, joints and soft tissues, delivering a variety of biomechanical functions during human
motions. Over the past decades, a large number of studies
based on FE method have been conducted to investigate
the biomechanical functions of the foot complex. Lemmon et al. used a 2D FE model to study the effect of
insoles on therapeutic footwear based on quasi-static
simulations[101]. Patil et al.[102] conducted a stress distribution study on normal and neuropathic feet during gait
using a 2D model, which was constructed from a lateral
X-ray image. Wu[103] constructed a 2D FE model to study
the foot bone and muscle stresses resulting from plantar
fasciotomy and major plantar ligament injuries. Chu et
al.[104] conducted a static parametric analysis using an
asymmetric 3D FE foot model to investigate the ankle-foot orthosis effects by considering the foot complex
as a single segment. Jacob et al.[105] developed a 3D FE
model with the purpose being to investigate the contributing factors to disintegration of tarsal bones in Hansens
disease and diabetes. Gefen et al.[106] constructed a subject-specific 3D foot model based on realistic bone geometry to investigate the biomechanical foot function
during gait. The stress distribution analysis was conducted at six representative instants of time during gait.
Gefen[107,108] also developed a 2D FE model to investigate the foot biomechanics following surgical plantar
fascia release and also to evaluate the plantar stress distribution of a standing diabetic foot. Cheung et al.[109111]
developed a more complicated 3D foot model by using
realistic bone geometry and nonlinear material properties.
The model was used to investigate the effects of plantar
fascia stiffness and Achilles tendon loading, and also to
conduct the parametric design using different structural
and material properties of a foot orthosis. Recently, the

Journal of Bionic Engineering (2014) Vol.11 No.2

164

foot plantar fascia mechanics, stress concentration in

plantar soft tissues and also the load transfer mechanism
were analyzed using different 3D FE foot models[112116].
However, it is noteworthy that almost all these studies are
static or quasi-static in nature. So far, very few foot
biomechanics studies used dynamic FE analysis. Dai et
al.[117] used a 3D foot model to investigate the effect of
sock wearing on the plantar pressure under different
contact conditions from the foot-flat to the push-off
during the stance phase of gait based on dynamic FE
simulations. However, constant loads were assumed and
extra constraints were used to define the model, which
may lead to unrealistic motion of the foot complex. Very
recently, a fully dynamic foot model without any extra
constraints has been developed to simulate the dynamic
behaviour of the human foot structure during stance
phase of walking, which has demonstrated some advantages over the traditional static or quasi-static FE models[118] (see Fig. 5).
3.2.2 FE modelling of musculoskeletal joints
The FE method has also been extensively used to
investigate the joint mechanics, especially the contact
stress and strain responses of different joint components.
Brown and Digioia[119] used a 2D FE model to analyze

t = 0 (s)

t = 0.03 (s)

t = 0.19 (s)

t = 0.23 (s)

t = 0.40 (s)

t = 0.44 (s)

the articular contact at hip joint, by representing the

cartilage using non-linear contact elements. An
axis-symmetric FE model of meniscus was proposed
with non-linear material properties, and the simulation
results suggested high radial strains in the regions where
lesions were most often observed[120]. Eckstein[121] used
the FE model to analyze the stress distribution at elbow
joint, and found that incongruity generates advantageous
mechanical stimuli in the joint tissues. The FE method
was also applied to the shoulder mechanism by using
truss, hinge and surface elements to construct the
model[122]. Sophisticated FE models of the knee joint
were also reported to study the meniscus and the contact
condition between cartilage and meniscus[123125]. In
addition to cartilages, ligaments are also important
components of a musculoskeletal joint. Usually, 1D
element was employed to represent ligaments in FE
modelling of joints. The 1D representation requires only
few parameters to define the mechanical behaviour. This
approach was proved useful for predicting joint kinematics under the application of external loads[126]. But it
has some shortcomings: (1) non-uniform 3D stresses and
strains cannot be predicted; (2) multiple sets of parameters and initial tensions routinely produce nearly
identical predictions of joint kinematics[127]. Ligaments

t = 0.08 (s)

t = 0.27 (s)

t = 0.43 (s)

t = 0.45 (s)

Fig. 5 The von Mises stress distribution predicted by a dynamic finite element foot model at 10 representative instants of
time over the whole stance phase of human walking[118].

Ren et al.: Biomechanics of Musculoskeletal System and Its Biomimetic Implications: A Review

are subjected to highly non-uniform deformations in vivo

that result from a combination of tension, shear, bending,
and compression[128,129], and the regional contribution of
a ligament to joint stability changes with joint orientation[130,131]. Therefore, 3D FE modelling is desirable to
represent these mechanical characteristics[132134]. The
FE modelling of musculoskeletal joints offers a useful
tool to predict the spatial and temporal variations in joint
contact stresses and strains, which are difficult or impossible to obtain experimentally.
3.2.3 FE modelling of skeletal muscles
As the prime mover of the MSK, skeletal muscles
demonstrate very complicated mechanical properties
coupled with neural excitations and muscle fibre contractions[135,136]. Currently, most of the mathematical
representations of muscle contraction mechanics are
based on either Hills model or Huxleys theory[137140].
Hills model depicts the dynamics of a musculotendon
unit using a set of connected discrete mechanical elements. This could help stimulate some initial schemes
toward a simple FE model of skeletal muscles[141].
Some initial works on FE modelling of skeletal
muscles involved the representation of muscle-tendon
complex using a number of simple active 1D line elements, each of which is composed of motor and viscoelastic elements. Since the 1D line elements did not
have volumes and masses, the information about muscle
tissue stresses and inertia effects could not be obtained.
Moreover, the muscle moment arms were assumed to be
equivalent for all fibres within a muscle compartment.
This limits the ability of the model to accurately represent the actual paths of muscles with complex geometry
and also the stress response of the active part and passive
part individually. A 3D FE muscle model has the potential to represent the complex muscular structures and
improve our understanding of the musculotendon mechanics[58]. Since a skeletal muscle consists of contractile muscle fibres arranged within a passive matrix of
connective tissues[142], a number of FE models have been
developed to describe the active behaviours of skeletal
muscles. Tsui et al.[143] constructed a 3D active FE
muscle model by using a user-defined muscle behaviour.
The simulation results of isometric force-length relationship and force-shortening contraction demonstrated
the potential of the model for studying muscle damage
and fatigue. Tang et al.[140] developed a 3D FE model of

165

skeletal muscles by integrating a modified Hills muscle

model with a muscle fatigue formula, but neglected the
different fibre types inside of the muscle. Based on the
two-state Huxley model, Oomens et al.[144] constructed a
3D muscle model to estimate the inhomogeneous strain
distribution in a skeletal muscle. A good agreement
between the measurement data and the simulation result
showed the proposed model could be employed as a tool
for studies on damage and adaptation of skeletal muscles.
Different from the mechanical properties during active
conditions, the passive muscle behaviour was normally
represented by non-linear hyperelastic or viscoelastic
constitutive relationship in the FE modelling of the
skeletal muscle[140,145,146] (see Fig. 6).
Due to the highly anisotropic, nonlinear material
property, and also the complex geometry (multiple layers
of soft tissue) and boundary conditions, 3D FE modelling
of stress-strain behaviour of skeletal muscles is normally
complicated and computationally demanding. The traditional solutions for FE simulations based on standard

Extended

Flexed

Fig. 6 The reconstructed 3D muscle models of gluteus maximus,

gluteus medius, iliacus, psoas by segmentation of MR images[146].

166

Journal of Bionic Engineering (2014) Vol.11 No.2

non-linear FE formulations are time-consuming even for

one or two muscles[147]. Blemker et al.[148] developed a
quasi-static invertible FE algorithm to reduce the computational load, which provided two major improvements over the traditional methods: (1) elements are
allowed to invert by computing robust FE forces with a
invertible framework; (2) the stiffness matrix is positive
semi-definite.
Nowadays, mathematical models representing the
mechanical behaviours for muscle-tendon complexes
used for studying the dynamics of the human MSK are
dominated by various Hill-type models[149151]. The development of bio-realistic and computationally efficient
FE models of the muscular system for assessing the
dynamic functions of biological MSK is still at its very
early stage.

4 Biomimetic studies inspired by MSK biomechanics

4.1 Bioinspired quadrupedal robot - BigDog
BigDog is a dynamically stable quadrupedal robot
developed by Boston Dynamics[152], with aim being to
provide load carriage service to accompany soldiers in
harsh rough terrains, which are impossible for conventional vehicles with wheels or treads. The size of the
robot is of a large dog or a small mule, about 1.1 m long
and 1 m tall, and weighs 109 kg. The robot has four legs
that are articulated like a typical quadrupedal animal,
with compliant elements to absorb shock and store energy during moving. It is capable of performing a variety
of locomotion behaviours, such as walking, running,
climbing, jumping and carrying heavy loads in
rough-terrain conditions[152].
Ideas and concepts inspired from quadrupedal
animals have been used in the structure and actuation
design, sensor and motion control of the BigDog robot.
The single leg structure, in terms of joint configuration,
standing posture and actuator position, is very similar to
the leg of a typical quadrupedal animal (see Fig. 7).
Going distally from hip joint to metatarsal joint, the leg
actuation of a typical quadrupedal animal becomes less
strong, and contains more compliance. In addition, the
forelimb and hindlimb configuration of the BigDog uses
similar design principle of the four-legged configuration
in horses, dogs and goats[153,154]. As the compliant
components of animals MSK play important role during
dynamic moving, such as running, jumping and gallop-

Fig. 7 The bio-inspired back leg of the BigDog robot compared

to the hindlimb of a typical quadrupedal animal[152].

ing[155], spring components are integrated into the leg of

the BigDog to attenuate the ground impact during dynamic moving (see Fig. 7). When moving on the ground,
the joint position, joint force, ground contact, ground
load and external obstacles of the robot are monitored by
using onboard sensors to ensure its dynamic balance and
stability. The robot could also behave like quadrupedal
animals to adapt to the local terrain variation by adjusting its body height and attitude, and also foot placements.
Based on those bio-inspired ideas and concepts, the
BigDog exhibits excellent locomotion performance, e.g.
it can climb slopes up to 35 degrees, walk across rubbles,
climb muddy hiking trails, walk in snow and water etc.
So, the BigDog has been considered as one of the most
advanced quadrupedal robots moving on rough terrains[152].
4.2 Efficient human-like robot with compliant legs
Over the past decades, many bipedal robots
have been developed to mimic human locomotion[10,12,13,156163]. Most of those walkers used precise
control to regulate the angle values of each individual
joint at each instant of time during locomotion. This
requires actuators with high precision and frequency
response, a precise environment model and also high
energy cost[10,12,13]. Recently, the concept of passive
dynamic waking, which needs less actuators and active
control than mainstream robots, was proposed as a new
design and control paradigm[159162,164]. It has been
demonstrated that periodic stable walking can be
achieved with high energy efficiency and little control
by integrating simple actuations into passive dynamic
walkers[160].

Ren et al.: Biomechanics of Musculoskeletal System and Its Biomimetic Implications: A Review

A recently developed bipedal robot based on the

passive dynamic walking concept, used minimum control to drive two elastic legs inspired from the structure
of the human leg MSK[164]. The robot consists of seven
body segments, two servomotors at the hip joints, four
passive joints at knee and ankle joints, and totally eight
linear springs (see Fig. 8). The spring components were
used to mimic the passive mechanical function of the
major musculotendon units in human legs. A unique
feature of the robot is that it has six springs spanning
over two joints to mimic the major biarticular muscles in
the leg. The experimental study showed that this
bio-inspired bipedal robot could produce human-like
walking gait by using extremely simple control without
sensory feedback[165]. This is a good example for development of bipedal robots with higher energy efficiency and more natural walking pattern based on ideas
inspired from the biological structure of the MSK.
(a)

(b)
Hip motor
Springs

Passive joints

Rubber

(c)

Fig. 8 The human-like bipedal robot with compliant legs[165]. (a)

bipedal robot model, only one leg is shown; (b) schematic of the
robot design; (c) the physical robot platform.

167

5 Challenges and future directions

The last decade has seen great progress and advance in the human/animal movement studies, which
aim for understanding the biomechanical functions of
the biological MSK using both experimental and computational approaches. However, due to the great complexity of the biological MSK, there still remain many
unsolved problems and grand challenges in musculoskeletal biomechanics. Skin artefact has been the major barrier for 3D motion analysis systems to become a
useful clinical diagnostic tool. X-ray based video systems (e.g. fluoroscopic systems) may be helpful to reduce the skin artefact during motion capture[166168], but
the effect of radiation and the limited measuring volume
prevent it from being useful in general cases. Assessment of individual muscle forces in vivo during human/animal motions has been proven to be a grand
challenge in musculoskeletal biomechanics field[169171].
Due to the limitation of the current measuring techniques and ethical reasons, the direct measurement of the
in vivo musculotendon forces is almost impossible. So,
rigid body musculoskeletal modelling technique is
normally used together with optimisation algorithms to
estimate the muscle forces that can reproduce measured
joint motions. However, the determination of the subject-specific muscle parameters, musculoskeletal geometry and the rigorous experimental validation of the
calculated results still remain big challenges[172]. In addition, development of predictive musculoskeletal
models, which are capable of predicting body kinematics and kinetics during various movements with minimum measurement inputs[61,66,85], will be one of the
major future research directions due to their great potential in clinical diagnosis, rehabilitation engineering
and surgical planning.
In musculoskeletal biomechanics studies, computational FE modelling provides a unique tool to assess
the internal stress/strain conditions of the biological
MSK, which is normally not measurable in vivo. Properly conducted FE studies could help to investigate the
fundamental biomechanical mechanisms of the MSK, to
improve our understanding of the associated musculoskeletal disorders, and hence to provide sound scientific basis to facilitate clinical diagnosis and surgical
treatments. One of the challenging works in FE modelling of MSK is to provide accurate definitions of the in

168

Journal of Bionic Engineering (2014) Vol.11 No.2

vivo material properties of the soft tissues and hard tissues in MSK. For example, the definition of the constitutive equation of cancellous bones is still a subject of
debate, in particular those relating to post-elastic behaviour[173175], and the failure criteria[176,177]. Similarly,
there are also lacks of accurate definitions for the in vivo
material properties for the soft tissues (e.g. cartilages,
ligaments, tendons and muscles etc.).
Another challenging work in FE modelling of MSK
is to provide bio-realistic representations of the anatomy,
structure and function of the human MSK at different
levels/scales (e.g. organ level, tissue level and cell level).
As we know, mechanical loadings at macro level have
effect on behaviours at micro level, conversely mechanical properties at micro level influence system responses at macro level[178]. For example, diabetic foot
ulceration may have a biomechanical etiology[179]. For
patients with diabetes, some common daily activities,
e.g. walking, may be harmful because diabetes may
affect the biological functions of MSK at various levels.
Dysfunctions at different levels manifest themselves in
terms of loss of sensation[180], changes in control of
movement[181], and alteration of tissues[182] and also cell
properties[183]. It is unclear how do mechanical loads at
macro level (e.g. ground reaction forces) response to
cellular deformations that may cause cell damage or
even ulceration. Mechanical loadings at macro level (e.g.
increased foot contact pressures), redistribution of stress
due to changes in tissue composition (e.g. muscular
atrophy[184], cell distribution within tissues, increased
mechanical loading of cells or their decreased damage
resistance may all have contributions to the development
of ulceration. Therefore, a multi-scale modelling
framework is needed to identify the pathways to cell
damage from the mechanical loadings at organ level
through to the deformations at cell level.
Multi-scale modelling has been used in basic science and engineering areas e.g. mathematics, material
science, chemistry and fluid dynamics etc. for many
years. When applied to MSK biomechanics, the
multi-scale modelling approach is normally based on an
integrated hierarchical structure at multiple body levels,
where the mechanical outputs of macro level models are
transmitted to micro level models with detailed representations of MSK at tissue and cell level[185]. Normally,
rigid body dynamics is used to simulate the mechanical
behaviour of MSK at body level, and continuum me-

chanics is employed to represent the stress-strain interplay at organ level, whereas for simulations at tissue and
cell levels specialized algorithms and solvers are normally needed[178,185]. Therefore, multi-scale MSK
simulations are computationally intensive, and require
intricate representations and also effective simulation
strategies/approaches to describe the complex interactions among multiple levels.
After multi-scale MSK simulations are conducted
to address specific research problems or particular
clinical questions, the next challenging stage is to interpret and validate the simulation results. It is a very
daunting and time consuming task to interpret the complicated calculation outcomes obtained or to extract
clinically meaningful information from the huge amount
of database generated by the multi-scale simulations.
Moreover, the lack of in vivo subject-specific data (e.g.
muscle forces, mechanical properties of hard and soft
tissues etc.) and the complexity associated with experimental measurements make the validation of the
simulation results even more challenging[178]. Although
parameter sensitivity studies coupled with statistical
populations of in vivo and primarily in vitro data may
provide some initial verifications, the limitation of the
current measuring techniques make a thorough subject-specific in vivo validation impossible. Furthermore,
the highly demanding nature of clinical problems need
the future multi-scale MSK models to be easy-to-use,
robust and also with timely solutions.
It is evident that, for multi-scale modelling of human/animal MSK, from its solution formulation to experimental validation and clinical application, the inherent challenges are hard to be handled based on the
current capacity of experimental and computational
biomechanics. To tackle them effectively, some synergetic efforts are necessary not only by coordinating all
works involved in a coherent way, but also by increasing
and encouraging the level of resources sharing and exchange in biomechanics community, e.g. data and model
sharing (including those developed by commercial
software packages and self-coded models), format
standardization, and dissemination of solution databases
with model distribution.

Acknowledgements
This work was supported by the International
Cooperation Project of National Natural Science

Ren et al.: Biomechanics of Musculoskeletal System and Its Biomimetic Implications: A Review

Foundation of China (No. 50920105504), the UK Engineering and Physical Sciences Research Council Grant
(No. EP/I033602/1), the Project of National Natural
Science Foundation of China (No. 51105167) and the
scientific and technological development planning project of Jilin Province, China (No. 20130522187JH).

References
[1]

van De Graaff K M, Strete D, Creek C H. Human Anatomy,

6th ed, McGraw Hill Higher Education, USA, 2001.

[2]

[3]

[6]

[7]

Mathematical and Engineering Sciences, 2007, 365, 4964.

[14] Ren L, Butler M, Miller C, Schwerda D, Fischer MS,
Hutchinson JR. The limb segment and joint kinematics of
Asian (Elephas maximus) and African (Loxodonta africana) elephants. Journal of Experimental Biology, 2008,
211, 27352751.
[15] Ren L, Miller C, Lair R, Hutchinson J R. Integration of
limbs. Proceedings of the National Academy of Sciences of

Baltimore, USA, 2001.

the United States of America (PNAS), 2010, 107,

Sharp N. Timed running speed of a cheetah (Acinonyx

70787082.

Hudson P E, Corr S A, Payne-Davis R C, Clancy S N, Lane

[16] Cappozzo A, Della Croce U, Leardini A. Chiarid L. Human

movement analysis using stereophotogrammetry - Part 1:

E, Wilson A M. Functional anatomy of the cheetah (Acinoyx

Theoretical background. Gait and Posture, 2005, 21,

jubatus) hindlimb. Journal of Anatomy, 2011, 218,

186196.
[17] Ren L, Jones R K, Howard D. Whole body inverse dy-

Gtzen N, Cross A R, Ifju P G, Rapoff A J. Understanding

namics over a complete gait cycle based only on measured

stress concentration about a nutrient foramen. Journal of

kinematics.

Biomechanics, 2003, 36, 15111521.

27502759.

Journal

of

Biomechanics,

2008,

41,

Ker R F, Bennett M B, Bibby S R, Kester R C, Alexander R

[18] Arndt A, Wolf P, Liu A, Nester C, Stacoff A, Jones R,

M. The spring in the arch of the human foot. Nature, 1987,

Lundgren P, Lundberg A. Intrinsic foot kinematics meas-

325, 147149.

ured in vivo during the stance phase of slow running.

Carrier D R, Heglund N C, Earls K D. Variable gearing

Science, 1994, 265, 651653.
Ren L, Howard D, Ren L Q, Nester C, Tian L M. A phase

Journal of Biomechanics, 2007, 40, 26722678.

[19] Cappozzo A. Gait analysis methodology. Human Movement
Science, 1984, 3, 2754.
[20] Allard P, Blanchi J P, Aissaoui R. Base of three-dimensional

dependent Hypothesis for locomotor functions of human

reconstruction. In: Allard P, Stokes I A F, Blanchi J P (eds).

foot complex. Journal of Bionic Engineering, 2008, 5,

Three-Dimensional Analysis of Human Movement, Human

175180.
[9]

phical Transactions of the Royal Society A: Physical,

biomechanical compliance, leverage, and power in elephant

during locomotion in the human musculoskeletal system.

[8]

Waseda University: WL and WABIAN family. Philoso-

loskeletal System, 3rd ed, Lippincott Williams & Wilkins,

363374.
[5]

[13] Lim H O, Takanishi A. Biped walking robots created at

Nordin M, Frankel V H. Basic Biomechanics of Muscu-

jubatus). Journal of Zoology, 1997, 241, 493494.

[4]

169

Kinetics, Champaign, IL, USA, 1995, 1940.

Pfeiffer F, Inoue H. Walking: Technology and biology.

[21] Kadaba M P, Ramakrishnan H K, Wootten M E. Meas-

Philosophical Transactions of the Royal Society A: Physi-

urement of lower extremity kinematics during level walk-

cal, Mathematical and Engineering Sciences, 2007, 365,

ing. Journal of Orthopaedic Research, 1990, 8, 383392.

39.

[22] Pedotti A, Ferrigno G. Optoelectronic-based system. In:

[10] Dillmann R, Albiez J, Gassmann B, Kerscher T, Zllner M.

Allard P, Stokes I A F, Blanchi J P (eds). Three-Dimensional

Biologically inspired walking machines: Design, control

Analysis of Human Movement, Human Kinetics, Cham-

and perception. Philosophical Transactions of the Royal

paign, IL, USA, 1995, 5777.

Society A: Physical, Mathematical and Engineering Sciences, 2007, 365, 133151.

[11] Hirose M, Ogawa K. Honda humanoid robots development.
Philosophical Transactions of the Royal Society A: Physical, Mathematical and Engineering Science, 2007, 365,
1119.
[12] Kimura H, Fukuoka Y, Cohen A H. Biologically inspired
adaptive walking of a quadruped robot. Philosophical
Transactions of the Royal Society A: Physical, Mathematical and Engineering Sciences, 2007, 365, 153170.

[23] Cappozzo A, Catani F, Leardini M, Benedetti G, Della

Croce U. Position and orientation in space of bones during
movement: Experimental artifacts. Clinical Biomechanics,
1996, 11, 90100.
[24] Fulle J, Liu J, Murphy M C, Mann R W. A comparison of
lower-extremity skeletal kinematics measured using skinand pin-mounted markers. Human Movement Science,
1997, 16, 219242.
[25] Sati M, De Guise J A, Larouche S L, Drouin G. Quantitative
assessment of skin marker movement at the knee. The Knee,

170

Journal of Bionic Engineering (2014) Vol.11 No.2

1996, 3, 121138.

Physiological Measurement, 2004, 25, 120.

[26] Krigslund R, Dosen S, Popovski P, Dideriksen J L, Peder-

[39] Pfau T, Spence A, Starke S, Ferrari M, Wilson A. Modern

sen G F, Farina D. A novel technology for motion capture

riding style improves horse racing times. Science, 2009,

using passive UHF RFID tags. IEEE Transactions on

325, 289.

Biomedical Engineering, 2013, 60, 14531457.

[40] Ren L, Hutchinson J R. The three-dimensional locomotor

[27] Mndermann L, Corazza S, Andriacchi T P. The evolution

dynamics of African (Loxodonta africana) and Asian

of methods for the capture of human movement leading to

(Elephas maximus) elephants reveal a smooth gait transi-

markerless motion capture for biomechanical applications.

tion at moderate speed. Journal of the Royal Society: In-

Journal of NeuroEngineering and Rehabilitation, 2006, 15,

36.
[28] Nigg B M, Herzog W. Biomechanics of the Musculoskeletal System, 2nd ed, John Wiley and Sons Ltd, New
York, USA, 1999.

terface, 2008, 5, 195211.

[41] de Luca C J. The use of surface electromyography in biomechanics. Journal of Applied Biomechanics, 1997, 13,
135163.
[42] Graiela-Flavia D, Flavia R, Emilia G. Surface electromy-

[29] Siegler S, Liu W. Inverse dynamics in human locomotion.

ography in biomechanics: Application and signal analysis

In: Allard P, Cappozzo A, Lundberg A, Vaughan C (eds).

aspects. Journal of Physical Education and Sport, 2009, 25,

Three-dimensional Analysis of Human Locomotion, John

Wiley and Sons Ltd, New York, USA, 1997, 191209.
[30] Winter D A. Biomechanics and Motor Control of Human
Movement, 3rd ed, John Wiley and Sons Ltd, New York,
USA, 2004.

110.
[43] Konrad P. The ABC of EMG: A Practical Introduction to
Kinesiological Electromyography (Version 1.0), Noraxon
Inc, USA, 2005.
[44] de Luca C J, Gilmore L D, Kuznetsov M, Roy S H. Filtering

[31] Cavanagh P R, Rodgers M M, Iiboshi A. Pressure distribu-

the surface EMG signal: Movement artifact and baseline

tion under symptom-free feet during barefoot standing.

noise contamination. Journal of Biomechanics, 2010, 43,

Foot & Ankle, 1987, 7, 262276.

15731579.

[32] Groundy M, Blackburn T P A, McLeish R D, Smidt L. An

[45] von Tscharner V. Intensity analysis in time-frequency space

investigation of the centers of pressure under the foot while

of surface myoelectric signals by wavelets of specified

walking. Journal of Bone and Joint Surgery, 1975, 57,

resolution. Journal of Electromyography & Kinesiology,

98103.

2000, 10, 433445.

[33] Bates K T, Savage R, Pataky T C, Morse S A, Webster E,

[46] Fornage B D. The case for ultrasound of muscles and ten-

Falkingham P L, Ren L, Qian Z H, Collins D, Bennett M R,

dons. Seminars in Musculoskeletal Radiology, 2000, 4,

McClymont J, Crompton R H. Does footprint depth correlate with foot motion and pressure. Journal of the Royal
Society Interface, 2013, 10, 20130009.
[34] Meyring S, Diehl R R, Milani T L, Hennig E M, Berlit P.

375391.
[47] Fukunaga T, Kawakami Y, Kubo K, Kanehisa H. Muscle
and tendon interaction during human movements. Exercise
and Sport Sciences Reviews, 2002, 30, 106110.

Dynamic plantar pressure distribution measurement in

[48] Kawakami Y, Fukunaga T. New insights into in vivo human

hemiparetic patients. Clinical Biomechanics, 1997, 12,

skeletal muscle function. Exercise and Sport Sciences Re-

6065.

views, 2006, 34, 1621.

[35] Gebruers N, Vanroy C, Truijen S, Engelborghs S, de Deyn P

[49] Hashimoto B E, Kramer D J, Wiitala L. Application of

P. Monitoring of physical activity after stroke: A systematic

musculoskeletal sonography. Journal of Clinical Ultra-

review of accelerometry-based measures. Archives of

Physical Medicine and Rehabilitation, 2010, 91, 288297.
[36] Yang C C, Hsu Y L. A review of accelerometry-based
wearable motion detectors for physical activity monitoring.
Sensors, 2010; 10, 77727788.
[37] Kavanagh J J, Menz H B. Accelerometry: A technique for
quantifying movement patterns during walking. Gait and
Posture, 2008, 28, 115.

sound, 1999, 27, 293299.

[50] Pillen S, van Alfen N. Skeletal muscle ultrasound. Neurological Research, 2011, 33, 10161024.
[51] Miller D I. A Computer Simulation Model of the Airborne
Phase of Diving, Pennsylvania State University, Pennsylvania, USA, 1970.
[52] Passerello C E, Huston R L. Human attitude control.
Journal of Biomechanics, 1971, 4, 95102.

[38] Mathie M J, Coster A C, Lovell N H, Celler B G. Accel-

[53] Hatze H. A comprehensive model for human motion

erometry: Providing an integrated, practical method for

simulation and its application to the take-off phase of the

long-term, ambulatory monitoring of human movement.

long jump. Journal of Biomechanics, 1981, 14, 135142.

Ren et al.: Biomechanics of Musculoskeletal System and Its Biomimetic Implications: A Review
[54] Brand R A, Pedersen D R, Friederich J A. The sensitivity of

171

neering perspective. In: Winters J M, Woo S L-Y (eds).

muscle force predictions to changes in physiologic

Multiple Muscle Systems: Biomechanics and Movement

cross-sectional area. Journal of Biomechanics, 1986, 19,

Organization, Spring-Verlag, New York, USA, 1990,

589596.

6993.

[55] Friederich J A, Brand R A. Muscle fiber architecture in the

[68] Pandy M G, Zajac F E, Sim E, Levine W S. An optimal

human lower limb. Journal of Biomechanics, 1990, 23,

control model for maximum-height human jumping.

9195.

Journal of Biomechanics, 1990, 23, 11851198.

[56] Zajac F E. Muscle and tendon: Properties, models, scaling,

[69] Delp S, Loan J. A computational framework for simulating

and application to biomechanics and motor control. Critical

and analyzing human and animal movement. Computing in

Reviews in Biomedical Engineering, 1989, 17, 359410.

Science and Engineering, 2000, 2, 4655.

[57] Amankwah K, Triolo R, Kirsch R, Audu M. A model-based

[70] Herrmann

A M,

Delp

L.

Moment

arm

and

study of passive joint properties on muscle effort during

force-generating capacity of the extensor carpi ulnaris after

static stance. Journal of Biomechanics, 2006, 39,

transfer to the extensor carpi radialis brevis. Journal of

22532263.

Hand Surgery, 1999, 24, 10831090.

[58] Delp S L, Loan J P, Hoy M G, Zajac F E, Topp E L, Rosen J

[71] Loren G J, Shoemaker S D, Burkholder T J, Jacobson M D,

M. An interactive graphics-based model of the lower ex-

Friden J, Leiber R L. Human wrist motors: Biomechanical

tremity to study orthopaedic surgical procedures. IEEE

design and application to tendon transfers. Journal of

Transactions on Biomedical Engineering, 1990, 37,

Biomechanics, 1996, 29, 331342.

757767.
[59] Delp S L, Loan J P. A graphics-based software system to
develop and analyze models of musculoskeletal structures.
Computers in Biology and Medicine, 1995, 25, 2134.
[60] Audu M L, Kirsch R F, Triolo R J. A computational technique for determining the ground reaction forces in human
bipedal stance. Journal of Applied Biomechanics, 2003, 19,
361371.
[61] Anderson F C, Pandy M G. Dynamic optimisation of human
walking. Journal of Biomechanical Engineering, 2001,
123, 381390.

[72] Zajac F E, Gordon M E. Determining muscles force and

action in multi-articular movement. Exercise Sport Science
Review, 1989, 17, 187230.
[73] Hollerbach J M, Flash T. Dynamic interactions between
limb segments during planar arm movement. Biological
Cybernetics, 1982, 44, 6777.
[74] Zajac F E, Neptune R R, Kautz S A. Biomechanics and
muscle coordination of human walking, Part I: Introduction
to concepts, power transfer, dynamics and simulations. Gait
and Posture, 2003, 16, 215232.
[75] Zajac F E, Neptune R R, Kautz S A. Biomechanics and

[62] Jenkyn T R, Nicol A C. A multi-segment kinematic model

muscle coordination of human walking, Part II: Lessons

of the foot with a novel definition of forefoot motion for use

from dynamical simulations and clinical implications. Gait

in clinical gait analysis during walking. Journal of Biomechanics, 2007, 40, 32713278.
[63] Nester C J, Liu A M, Ward E, Howard D, Cocheba J, Derrick T, Paterson P. In vitro study of foot kinematics using a
dynamic walking cadaver model. Journal of Biomechanics,
2007, 40, 19271937.
[64] Ren L, Howard D, Ren L Q, Tian L M, Nester C J. A generic
analytical foot rollover model for predicting translational
ankle kinematics in gait simulation studies. Journal of
Biomechanics, 2010, 43, 194202.
[65] Neptune R R, Kautz S A, Zajac F E. Contributions of the

and Posture, 2003, 17, 117.

[76] Ren L, Howard D, Kenney L. Computational models to
synthesize human walking. Journal of Bionic Engineering,
2006, 3, 127138.
[77] Davoodi R, Urata C, Hauschild M, Khachani M, Loeb G E.
Model-based development of neural prostheses for movement. IEEE Transaction of Biomedical Engineering, 2007,
54, 19091918.
[78] McKay J L, Burkholder T J, Ting L H. Biomechanical
capabilities influence postural control strategies in the cat
hindlimb. Journal of Biomechanics, 2007, 40, 22542260.

individual ankle plantar flexors to support, forward pro-

[79] Berniker M, Jarc A, Bizzi E, Tresch M C. Simplified and

gression and swing initiation during normal walking.

effective motor control based on muscle synergies to ex-

Journal of Biomechanics, 2001, 34, 13871398.

ploit musculoskeletal dynamics. Proceedings of the Na-

[66] Pandy M G. Computer modeling and simulation of human

movement. Annual Review Biomedical Engineering, 2001,
3, 245273.
[67] Winters J M. Hill-based muscle models: A systems engi-

tional Academy of Sciences, USA, 2009, 106, 76017606.

[80] Pai D K. Muscle mass in musculoskeletal models. Journal
of Biomechanics, 2010, 43, 20932098.
[81] Ramsay J W, Hunter B V, Gonzalez R V. Muscle moment

172

Journal of Bionic Engineering (2014) Vol.11 No.2

arm and normalized moment contributions as reference data

[95] Goel V K, Clausen J D. Prediction of load sharing among

for musculoskeletal elbow and wrist joint models. Journal

spinal components of a C5-C6 motion segment using the

of Biomechanics, 2009, 42, 463473.

finite element approach. Spine, 1998, 23, 684691.

[82] Ren L, Jones R, Howard D. Dynamic analysis of load

[96] Teo E C, Ng H W. First cervical vertebra (atlas) fracture

carriage biomechanics during human level walking. Jour-

mechanism studies using finite element method. Journal of

nal of Biomechanics, 2005, 38, 853863.

Biomechanics, 2001, 34, 1321.

[83] Delp S L, Anderson F C, Arnold A S, Loan P, Habib A, John

[97] Puttlitz C M, Goel V K, Clark C R, Traynelis V C, Scifert J

C T, Guendelman E, Thelen D G. OpenSim: Open-source

L, Grosland N M. Biomechanical rationale for the pathol-

software to create and analyze dynamic simulations of

ogy of rheumatoid arthritis in the craniovertebral junction.

movement. IEEE Transactions of Biomedical Engineering,

Spine, 2000, 25, 16071616.

2007, 54, 19401950.

[98] Stemper B D, Kumaresan S, Yoganandan N, Pintar F A.

[84] Saraswat P, Andersen M S, Macwilliams B A. A muscu-

Head-neck finite element model for motor vehicle inertial

loskeletal foot model for clinical gait analysis. Journal of

impact: Material sensitivity analysis. Biomedical Sciences

Biomechanics, 2010, 43, 16451652.

Instrumentation, 2000, 36, 331335.

[85] Ren L, Jones R, Howard D. Predictive modelling of human

[99] Camacho D L, Nightingale R W, Myers B S. Surface fric-

walking over a complete gait cycle. Journal of Biome-

tion in near-vertex head and neck impact increases risk of

chanics, 2007, 40, 15671574.

injury. Journal of Biomechanics, 1999, 32, 293301.

[86] Moroney S P, Schultz A B, Miller J A, Andersson G B.

[100] Zhang Q H, Teo E C, Ng H W. Development and validation

Load-displacement properties of lower cervical spine mo-

of a C0-C7 FE complex for biomechanical study. Journal of

tion segments. Journal of Biomechanics, 1988, 21,

769779.

Biomechanical Engineering, 2005, 127, 729735.

[101] Lemmon D, Shiang T Y, Hashmi A, Ulbrecht J S, Cavanagh

[87] Shirazi-Adl A, M Ahmed A, Shrivastava S C. A finite ele-

P R. The effect of insoles in therapeutic footwear: A finite

ment study of a lumbar motion segment subjected to pure

element approach. Journal of Biomechanics, 1997, 30,

sagittal plane moments. Journal of Biomechanics, 1986, 19,

331350.

615620.
[102] Patil K M, Braak L H, Huson A. Analysis of stresses in

[88] Yoganandan N, Kumaresan S, Voo L, Pintar F A. Finite

two-dimensional models of normal and neuropathic feet.

element applications in human cervical spine modelling.

Medical & Biological Engineering & Computing, 1996, 34,

Spine, 1996, 21, 18241834.

280284.

[89] Bozic K J, Keyak J H, Skinner H B, Bueff H U, Bradford D

[103] Wu L J. Nonlinear finite element analysis for muscu-

S. Three-dimensional finite element modeling of a cervical

loskeletal biomechanics of medial and lateral plantar lon-

vertebra: An investigation of burst fracture mechanism.

gitudinal arch of virtual Chinese human after plantar liga-

Journal of Spinal Disorders, 1994, 7, 102110.

[90] Teo E C, Paul J P, Evans J H. Finite element stress analysis
of a cadaver second cervical vertebra. Medical & Biological
Engineering & Computing, 1994, 32, 236138.
[91] Saito T, Yamamuro T, Shikata J, Oka M, Tsutsumi S.
Analysis and prevention of spinal column deformity following cervical laminectomy, pathogenetic analysis of post
laminectomy deformities. Spine, 1991, 16, 494502.
[92] Kleinberger M. Application of finite element techniques to

mentous structure failures. Clinical Biomechanics, 2007,

22, 221229.
[104] Chu T M, Reddy N P, Padovan J. Three-dimensional finite
element stress analysis of the polypropylene, ankle-foot
orthosis: Static analysis. Medical Engineering and Physics,
1995, 17, 372379.
[105] Jacob S, Patil K M, Braak L H, Huson A. Stresses in a 3D
two arch model of a normal human foot. Mechanics Research Communications, 1996, 23, 387393.

the study of cervical spine mechanics. Proceedings of the

[106] Gefen A, Megido-Ravid M, Itzchak Y, Arcan M. Biome-

37th Stapp Car Crash Conference, San Antonio, Texas,

chanical analysis of the three-dimensional foot structure

USA, 1993, 78, 261272.

[93] Voo L, Denman J A, Yoganandan N, Pintar F, Cusick J F. A
3D FE model of cervical spine with CT-based geometry.
Advanced Bioengineering, 1995, 29, 323324.
[94] Voo L, Denman J, Kumaresan S, Yoganandan N, Pintar F A,

during gait: A basic tool for clinical applications. Journal of

Biomechanical Engineering, 2000, 122, 630639.
[107] Gefen A. Stress analysis of the standing foot following
surgical plantar fascia release. Journal of Biomechanics,
2000, 35, 629637.

Cusick J F. Development of 3D finite element model of the

[108] Gefen A. Plantar soft tissue loading under the medial

cervical spine. Advanced Bioengineering, 1995, 31, 1314.

metatarsals in the standing diabetic foot. Medical Engi-

Ren et al.: Biomechanics of Musculoskeletal System and Its Biomimetic Implications: A Review
neering and Physics, 2003, 25, 491499.

173

545552.

[109] Cheung J T, Zhang M. Parametric design of pressure- re-

[122] van der Helm F C T, Veeger H E J, Pronk G M, Van der

lieving foot orthosis using statistics-based finite element

Woude L H V, Rozendal R H . Geometry parameters for

method, Medical Engineering and Physics, 2008, 30,

musculoskeletal modeling of the shoulder system. Journal

269277.

of Biomechanics, 1992, 25, 129144.

[110] Cheung J K, Zhang M, An K N. Effects of plantar fascia

[123] Spilker R L, Donzelli P S, Mow V C. A transversely iso-

stiffness on the biomechanical responses of the ankle-foot

tropic biphasic finite element model of the meniscus.

complex. Clinical Biomechanics, 2004, 19, 839846.

Journal of Biomechanics, 1992, 25, 10271045.

[111] Cheung J T, Zhang M, Leung A K, Fan Y B.

[124] Donzelli P S, Spilker R L. A mixed penalty contact finite

Three-dimensional finite element analysis of the foot dur-

element formulation with applications to biphasic tissues of

ing standing: A material sensitivity study. Journal Biome-

the knee. ASME International Mechanical Engineering

chanics, 2005, 38, 10451054.

Congress and Exposition, Chicago, USA, 1994, 1314.

[112] Garcia-Aznar J M, Bayod J, Rosas A, Larrainzar R, Gar-

[125] Donahue T L, Hull M L, Rashid M M, Jacobs C R. A finite

ca-Bgalo R, Doblar M, Llanos L F. Load transfer

element model of the human knee joint for the study of ti-

mechanism for different metatarsal geometries: A finite

bio-femoral contact. Journal of Biomechanical Engineer-

element study. Journal of Biomechanical Engineering,

ing, 2002, 124, 273280.

2009, 131, 021011.

[113] Cheng H Y, Lin C L, Wang H W, Chou SW. Finite element
analysis of plantar fascia under stretch: The relative contribution of windlass mechanism and Achilles tendon force.
Journal of Biomechanics, 2008, 41, 19371944.
[114] Hsu Y C, Gung Y W, Shih S L, Feng C K, Wei S H, Yu C H,
Chen C S. Using an optimization approach to design an

[126] Li G, Kanamori A, Woo S L. A validated three-dimensional

computational model of a human knee joint. Journal of
Biomechanical Engineering, 1999, 121, 657662.
[127] Weiss J A, Gardiner J C, Ellis B J, Lujan T J, Phatak N S.
Three-dimensional finite element modeling of ligaments:
Technical aspects. Medical Engineering and Physics, 2005,
27, 845861.

insole for lowering plantar fascia stress: A finite element

[128] Giori N J, Beaupre G S, Carter D R. Cellular shape and

study. Annals of Biomedical Engineering, 2008, 36,

pressure may mediate mechanical control of tissue compo-

13451352.

sition in tendons. Journal of Orthopaedic Research, 1993,

[115] Chen W M, Lee T, Lee P V, Lee J W, Lee S J. Effects of

11, 581591.

internal stress concentrations in plantar soft tissue: A pre-

[129] Matyas J R, Anton M G, Shrive N G, Frank C B. Stress

liminary three-dimensional finite element analysis. Medical

governs tissue phenotype at the femoral insertion of

Engineering and Physics, 2010, 32, 324331.

[116] Chen W M, Park J, Park S B, Shim V P, Lee T. Role of

the rabbit MCL. Journal of Biomechanics, 1995, 28,

147157.

gastrocnemius-soleus muscle in forefoot force transmission

[130] Gardiner J C, Weiss J A, Rosenberg T D. Strain in the hu-

at heel rise: A 3D finite element analysis, Journal of Bio-

man medial collateral ligament during valgus loading of the

mechanics, 2012, 45, 17831789.

knee. Clinical Orthopaedics, 2001, 391, 266274.

[117] Dai X Q, Li Y, Zhang M, Cheung J T. Effect of sock on

[131] Kawada T, Abe T, Yamamoto K, Hirokawa S, Soejima T,

biomechanical responses of foot during walking. Clinical

Tanaka N, Inoue A. Analysis of strain distribution in the

Biomechanics, 2006, 21, 314321.

medial collateral ligament using a photoelastic coating

[118] Qian Z H, Ren L, Ding Y, Hutchinson J R, Ren L Q. A

dynamic finite element analysis of human foot complex in
the sagittal plane during level walking. PLoS ONE, 2013, 8,
e79424.

method. Medical Engineering and Physics, 1999, 21,

279291.
[132] Song Y, Debski R E, Musahl V, Thomas M, Woo S L. A
three-dimensional finite element model of the human ante-

[119] Brown T D, DiGioia A M. A contact-coupled finite element

rior cruciate ligament: A computational analysis with ex-

analysis of the natural adult hip. Journal of Biomechanics,

perimental validation. Journal of Biomechanics, 2004, 37,

1984, 17, 437448.

[120] Aspeden R M. A model for function and failure of the
meniscus, Engineering in Medicine, 1985, 14, 119122.

383390.
[133] Debski R E, Weiss J A, Newman W J, Moore S M,
McMahon P J. Stress and strain in the anterior band of the

[121] Eckstein F, Merz B, Schmid P, Putz R. The influence of

inferior glenohumeral ligament during a simulated clinical

geometry on the stress distribution in joints: A finite ele-

exam. Journal of Shoulder and Elbow Surgery, 2005, 14,

ment analysis, Anatomy and Embryology, 1994, 189,

S24S31.

Journal of Bionic Engineering (2014) Vol.11 No.2

174

[134] Limbert G, Taylor M, Middleton J. Three-dimensional finite

[147] Blemker S S, Pinsky P M, Delp S L. A 3D model of muscle

element modelling of the human ACL: Simulation of pas-

reveals the causes of nonuniform strains in the biceps

sive knee flexion with a stressed and stress-free ACL.

Journal of Biomechanics, 2004, 37, 17231731.

brachii. Journal of Biomechanics, 2005, 38, 657885.

[148] Blemker S S, Teran J, Sifakis E, Fedkiw R, Delp S. Fast 3D

[135] Huxley A F. The mechanism of muscular contraction. Science, 1969, 164, 13561365.

muscle simulations using a new quasi-static invertible finite

element algorithm. First MIT International Symposium on

[136] Huxley A F. Muscular contraction. Journal of Physiology,

1974, 243, 143.

Computer Simulation in Biomechanics, Cleveland, USA,

2005.

[137] Hatze H. Myocybernetic Control Models of Skeletal Mus-

[149] van den Bogert A J. Analysis and simulation of mechanical

cle: Characteristics and Applications, University of South

loads on the human musculoskeletal system. Exercise and

Africa, Pretoria, South Africa, 1981.

Sport Sciences Reviews, 1994, 22, 2351.

[138] Riek S, Chapman A E, Milner T A. Simulation of muscle

[150] Gerritsen K G M, van den Bogert A J, Nachbauer W.

force and internal kinematics of extensor carpi radialis

Computer simulation of loading movement in downhill

brevis during backhand tennis stroke: Implications for in-

skiing: Anterior cruciate ligament injuries. Journal of

jury. Clinical Biomechanics, 1999, 4, 7783.

Biomechanics, 1996, 29, 845854.

[139] Stojanovic B, Kojic M, Rosic M, Tsui C P, Tang C Y. An

[151] Neptune R R, Kautz S A. Knee joint loading in forward

extension of Hills three-component model to include dif-

versus backward pedaling: Implications for rehabilitation

ferent fiber types in finite element modeling of muscle.

strategies. Clinical Biomechanics, 2000, 15, 528535.

International Journal for Numerical Methods in Engi-

[152] Raibert M, Blankespoor K, Nelson G, Playter R. BigDog,

neering, 2007, 71, 801817.

the rough-terrain quadruped robot. Proceeding of the 17th

[140] Tang C Y, Tsui C P, Stojanovic B, Kojic M. Finite element

modelling of skeletal muscles coupled with fatigue. International Journal of Mechanical Sciences, 2007, 49,

World Congress, the International Federation of Automatic

Control, Seoul, Korea, 2008.
[153] Flammang B E, Porter M E. Bioinspiration: Applying
mechanical design to experimental biology. Integrative and

11791191.
[141] Johansson T, Meier P, Blickhan R. A finite-element model
for the mechanical analysis of skeletal muscles. Journal of

Comparative Biology, 2011, 51, 128132.

[154] Lee D V, Biewener A A. BigDog-inspired studies in the
locomotion of goats and dogs. Integrative and Comparative

Theoretical Biology, 2000, 206, 131149.

[142] Yucesoy C A, Koopman B H, Huijing P A, Grootenboer H
J. Three-dimensional finite element modeling of skeletal

Biology, 2011, 51, 190202.

[155] Alexander R M. Elastic Mechanisms in Animal Movement,

muscle using a two-domain approach: Linked fiber-matrix

Cambridge University Press, Cambridge, UK, 1988.

mesh model. Journal of Biomechanics, 2002, 35,

[156] Kato I, Ohteru S, Kobayashi H, Shirai K, Uchiyama A.

Information-power machine with senses and limbs. Pro-

12531262.
[143] Tsui C P, Tang C Y, Leung C P, Cheng K W, Ng Y F, Chow

ceedings of the First CISM-IFTOMM Symposium on The-

D H, Li C K. Active finite element analysis of skeletal

ory and Practice of Robots and Manipulators, Udine, Italy,

muscle-tendon complex during isometric, shortening and

lengthening contraction. Bio-Medical Materials and Engineering, 2004, 14, 271279.

R, Canudas-de-Wit C, Grizzle J W. A testbed for advanced

[144] Oomens C W, Maenhout M, van Oijen C H, Drost M R,

Baaijens F P. Finite element modelling of contracting
skeletal muscle, Philosophical Transactions of the Royal
Society B: Biological Sciences, 2003, 358, 14531460.
C V C, Drost M RBaaijens F P T. Passive transverse
mechanical properties of skeletal muscle under the in vivo
Journal

of

Biomechanics,

2001,

control theory. IEEE Control Systems Magazine, 2003, 23,

5779.
[158] Pfeiffer F, Loffler K, Gienger M. The concept of jogging
JOHNNIE. Proceedings of IEEE International Conference

[145] Bosboom E M H, Hesselink M K C, Oomens C W J, Bouten

compression.

1973.
[157] Chevallereau C, Abba G, Aoustin Y, Plestan F, Westervelt E

34,

13651368.

on Robotics and Automation, Washington DC, USA, 2002,

31293135.
[159] Coleman M J, Ruina A. An uncontrolled walking toy that
cannot stand still. Physical Review Letters, 1998, 80,
36583661.

[146] Blemker S S, Delp S L. Three-dimensional representation

[160] Collins S, Ruina A, Tedrake R, Wisse M. Efficient bipedal

of complex muscle architectures and geometries. Annals of

robots based on passive-dynamic walkers. Science, 2005,

Biomedical Engineering, 2005, 33, 661673.

307, 10821085.

Ren et al.: Biomechanics of Musculoskeletal System and Its Biomimetic Implications: A Review

175

[161] Collins S H, Wisse M, Ruina A. A 3-D passive-dynamic

[174] Keyak J H, Lee I Y, Nath D S, Skinner H B. Postfailure

walking robot with two legs and knees. International

compressive behavior of tibial trabecular bone in three

Journal of Robotics Research, 2001, 20, 607615.

anatomic directions. Journal of Biomedical Materials Re-

[162] Wisse M. Three additions to passive dynamic walking;

actuation, an upper body, and 3D stability. International
Journal of Humanoid Robotics, 2005, 2, 459478.
[163] Manoonpong P, Geng T, Kulvicius T, Porr B, Worgotter, F.
Adaptive, fast walking in a biped robot under neuronal
control and learning. PLoS Computational Biology, 2007,
3, e134.
[164] McGeer T. Passive dynamic walking. International Journal
of Robotics Research, 1990, 9, 6282.
[165] Iida F, Minekawa Y, Rummel J, Seyfarth A. Toward a human-like biped robot with compliant legs. Robotics and
Autonomous Systems, 2009, 57,139144.

search, 1996, 31, 373378.

[175] Viceconti M, Bellingeri L, Cristofolini L, Toni A. A comparative study on different methods of automatic mesh
generation of human femurs. Medical Engineering and
Physics, 1998, 20, 110.
[176] Morgan E F, Keaveny T M. Dependence of yield strain of
human trabecular bone on anatomic site. Journal of Biomechanics, 2001, 34, 569577.
[177] Morgan E F, Bayraktar H H, Keaveny T M. Trabecular bone
modulus-density relationships depend on anatomic site.
Journal of Biomechanics, 2003, 36, 897904.
[178] Tawhai M, Bischoff J, Einstein D, Erdemir A, Guess T,

[166] Kedgley A E, Birmingham T, Jenkyn T R. Comparative

Reinbolt J. Multiscale modeling in computational biome-

accuracy of radiostereometric and optical tracking systems.

chanics. Engineering in Medicine and Biology Magazine,

Journal of Biomechanics, 2009, 42, 13501354.

2009, 28, 4149.

[167] Shultz R, Kedgley A E, Jenkyn T R. Quantifying skin motion artifact error of the hindfoot and forefoot marker clus-

[179] International Working Group on the Diabetic Foot.

Apelqvist

J,

Bakker

K,

van

Houtum

H,

ters with the optical tracking of a multi-segment foot model

Nabuurs-Franssen M H, Schaper N C (eds). International

using single-plane fluoroscopy. Gait and Posture, 2011, 34,

Consensus on the Diabetic Foot, Maastricht, the Nether-

4448.

land, 1999.

[168] Akbarshahi M, Schache A G, Fernandez J W, Baker R,

[180] Yagihashi S, Yamagishi S, Wada R. Pathology and patho-

Banks S, Pandy M G. Non-invasive assessment of

genetic mechanisms of diabetic neuropathy: Correlation

soft-tissue artifact and its effect on knee joint kinematics

with clinical signs and symptoms. Diabetes Research and

during functional activity. Journal of Biomechanics, 2010,

Clinical Practice, 2007, 77, S184S189.

43, 12921301.
[169] Erdemir A, McLean S, Herzog W, van den Bogert AJ.

[181] Kwon O Y, Minor S D, Maluf K S, Mueller M J. Comparison of muscle activity during walking in subjects with

Model-based estimation of muscle forces exerted during

and without diabetic neuropathy. Gait and Posture, 2003,

movements. Clinical Biomechanics, 2007, 22, 131154.

18, 105113.

[170] Fregly B J, Besier T F, Lloyd D G, Delp S L, Banks S A,

[182] Loganathan R, Bilgen M, Al-Hafez B, Smirnova I V.

Pandy M G, DLima D D. Grand challenge competition to

Characterization of alterations in diabetic myocardial tissue

predict in vivo knee loads. Journal of Orthopedic Research,

using high resolution MRI. International Journal of Car-

2012, 30, 503513.

diovascular Imaging, 2006, 22, 8190.

[171] Kinney A L, Besier T F, DLima D D, Fregly B J. Update on

[183] Lorenzi M, Gerhardinger C. Early cellular and molecular

grand challenge competition to predict in vivo knee loads.

changes induced by diabetes in the retina. Diabetologia,

Journal of Biomechanical Engineering, 2013, 135, 021012.

2001, 44, 791804.

[172] Fregly B J, Boninger M L, Reinkensmeyer D J. Personal-

[184] Bus S A, Yang Q X, Wang J H, Smith M B, Wunderlich R,

ized neuromusculoskeletal modeling to improve treatment

Cavanagh P R. Intrinsic muscle atrophy and toe deformity

of mobility impairments: A perspective from European re-

in the diabetic neuropathic foot: A magnetic resonance

search sites. Journal of NeuroEngineering and Rehabilitation, 2012, 9, 18.

[173] Ford C M, Keaveny T M, Hayes W C. The effect of impact

imaging study. Diabetes Care, 2002, 25, 14441450.

[185] Agoram B, Barocas V H. Coupled macroscopic and microscopic scale modeling of fibrillar tissues and tissue

direction on the structural capacity of the proximal femur

equivalents. Journal of Biomechanical Engineering, 2001,

during falls. Journal of Bone and Mineral Research, 1996,

123, 362369.

11, 377383.