bioengineering

Perspective
The Digital Twin: A Potential Solution for the Personalized
Diagnosis and Treatment of Musculoskeletal System Diseases
Tianze Sun 1,2,† , Jinzuo Wang 1,2,† , Moran Suo 1,2 , Xin Liu 1,2 , Huagui Huang 1,2 , Jing Zhang 1,2 ,
Wentao Zhang 1,2 and Zhonghai Li 1,2, *

1 Department of Orthopedics, First Affiliated Hospital of Dalian Medical University, Dalian 116600, China;
[email protected] (M.S.)
2 Key Laboratory of Molecular Mechanism for Repair and Remodeling of Orthopedic Diseases,
Dalian 116000, China
* Correspondence: [email protected]
† These authors contributed equally to the manuscript and should be considered as co-first authors.

Abstract: Due to the high prevalence and rates of disability associated with musculoskeletal system
diseases, more thorough research into diagnosis, pathogenesis, and treatments is required. One of
the key contributors to the emergence of diseases of the musculoskeletal system is thought to be
changes in the biomechanics of the human musculoskeletal system. However, there are some defects
concerning personal analysis or dynamic responses in current biomechanical research methodologies.
Digital twin (DT) was initially an engineering concept that reflected the mirror image of a physical
entity. With the application of medical image analysis and artificial intelligence (AI), it entered our
lives and showed its potential to be further applied in the medical field. Consequently, we believe that
DT can take a step towards personalized healthcare by guiding the design of industrial personalized
healthcare systems. In this perspective article, we discuss the limitations of traditional biomechanical
methods and the initial exploration of DT in musculoskeletal system diseases. We provide a new
opinion that DT could be an effective solution for musculoskeletal system diseases in the future,
which will help us analyze the real-time biomechanical properties of the musculoskeletal system and
Citation: Sun, T.; Wang, J.; Suo, M.;
achieve personalized medicine.
Liu, X.; Huang, H.; Zhang, J.; Zhang,
W.; Li, Z. The Digital Twin: A
Potential Solution for the
Keywords: digital twin; AI; musculoskeletal system; biomechanics; personalized medicine
Personalized Diagnosis and
Treatment of Musculoskeletal System
Diseases. Bioengineering 2023, 10, 627.
https://doi.org/10.3390/ 1. Introduction
bioengineering10060627 With the aging of populations and increasing attention to non-communicable dis-
Academic Editors: Victor Hugo C. De eases, the global incidence of musculoskeletal system disease has been increasing annually,
Albuquerque and Andrea Cataldo which has been both a strong indication of the need for medical treatment and the main
cause of days lost from work [1]. It has become a globally common symptom and occurs
Received: 8 April 2023 in all ages. Musculoskeletal system diseases have high incidence and disability rates;
Revised: 12 May 2023
thus, it is necessary to conduct more in-depth research on diagnosis, pathogenesis, and
Accepted: 20 May 2023
treatments [2]. The occurrence of these disorders is often accompanied by biomechanical
Published: 23 May 2023
changes, and biomechanical factors are indispensable in the development and treatment
of these disorders [3]. Analyzing biomechanical changes can help us improve the under-
standing of disease pathogenesis and prevention and help physicians choose appropriate
Copyright: © 2023 by the authors.
treatment strategies. However, conventional biomechanical research methods have certain
Licensee MDPI, Basel, Switzerland. defects concerning personality analysis or dynamic responses. Digital twin (DT) technol-
This article is an open access article ogy may provide new ideas for further improving biomechanical analysis methods in the
distributed under the terms and medical field.
conditions of the Creative Commons Early DTs were introduced and employed in industrial and manufacturing fields, such
Attribution (CC BY) license (https:// as aerospace, DT workshops, etc. [4–6]. Currently, the goal of personalized medicine is
creativecommons.org/licenses/by/ to offer each patient a unique course of treatment. Artificial intelligence (AI) has been
4.0/). gradually applied in clinics, providing clinicians with a repeatable second opinion and

Bioengineering 2023, 10, 627. https://doi.org/10.3390/bioengineering10060627 https://www.mdpi.com/journal/bioengineering

Bioengineering 2023, 10, 627 2 of 11

laying the technical groundwork for the realization of DT. On the basis of multi-modal
data integration and fusion, it has gained several successful applications in the medical
field [7]. Referring to cases of DTs applied in industry and combining characteristics of
medicine itself, we used this brand-new technical method to create a DT of the lumbar
spine in the first instance [8]. The research was the first case of DT technology applied
in regard to the spine, using interactions between the physical and digital worlds to
predict the biomechanical parameters of intervertebral discs. In some areas, such as
real-time synchronization, faithful mapping, high fidelity, various data integration, and
full periodicity, DT has incomparable advantages over conventional methods [9]. It can
be further applied in the whole musculoskeletal system to counteract the shortcomings
of conventional methods and explain complex biomechanical problems. In the future,
high-resolution models will be established and DT will finally become a solution for
musculoskeletal system diseases to achieve real-time monitoring and/or precise diagnosis.
Herein, we discuss the limitations of conventional biomechanical methods and suggest a
potential solution for personalized diagnosis and the treatment of musculoskeletal system
diseases. Current applications of DT in musculoskeletal systems were reviewed and we
hope that there will be more related research on the musculoskeletal system, providing
valuable research data and solutions for the prevention, treatment, and monitoring of
diseases. At the same time, it will further help us analyze the real-time biomechanical
properties of the musculoskeletal system and realize personalized medicine.

2. Biomechanics of Musculoskeletal System

Biomechanics are ubiquitous in the musculoskeletal system and play an indispensable
role in the occurrence, development, and treatment of disorders, including spine-related
diseases such as lower back and neck pain, joint-related diseases such as osteoarthritis,
traumatic diseases such as fractures, and other musculoskeletal diseases such as ventilation
and rheumatoid arthritis [10]. The musculoskeletal system supports and protects other
parts of the body and promotes movement. This system expresses mechanical properties
through intricate interactions between bones and soft tissues [11,12]. Stresses on bones
include compressive, tensile, and fluid shear stress. Bone formation and reconstruction
under stress occur due to changes in the function of local osteoblasts and osteoclasts [13,14].
Intervertebral disc degeneration is one of the major causes of lower back pain; its
biology and mechanics have been implicated as the predominant inductive cause [15]. In
addition to the common loading mechanisms, chronic loading has a large destructive effect
on the human spine. Adverse outcomes include uneven stress distribution and reduced
mechanical stiffness of the vertebral body and appendages, disrupting the original stable
biomechanical structure of the spine [15,16]. The load-sharing relationship between the
discs and facet joints is very intricate, and much depends on the posture of our spine [17].
The degeneration of discs causes an increase in the transmission of force across the facet
joints [18]. When facet tropism (FT) occurs, the angles of the ipsilateral sagittal plane
of the facet joint are different, whereby the intervertebral disc stress on the larger angle
side increases, and the stress increases with increases in FT values [19]. Biomechanical
factors also play an important role in joint diseases. Taking the knee as an example, using
biomechanical methods can sufficiently analyze and verify the role of the anterolateral
soft tissue in its rotational stability and the change of the internal rotation angle after the
corresponding ligament injury [20,21].

3. Limitations of Conventional Biomechanical Methods

The anatomy of the human musculoskeletal system is complex and there are multiple
mechanisms of bone diseases. Conventional biomechanical methods have shortcomings
in some aspects, which inevitably lead to deviations in the human anatomy [22]. The
research methods of human biomechanics can be divided into in vivo and in vitro research.
In vivo research includes conventional technology based on morphology and imaging
systems and implantable sensor technology [23]. In vivo research can better understand
Bioengineering 2023, 10, 627 3 of 11

some complex structures that are difficult to accurately simulate by computer models so as
to better predict the progression of diseases and optimize treatment plans [24]. However,
several technical and safety bottlenecks such as the volume of sensor systems and the safety
of battery technology should be overcome [25].
In vitro experiments include cell or animal experiments, as well as cadaver studies.
However, it is hard to simulate the microenvironment with cell experiments, and results
from animal models and tissue cultures have low reproducibility. Additionally, human
cadavers are scarce and often subject to ethical constraints, making finite element anal-
ysis (FEA) an effective tool for studying orthopedic-related diseases [26,27]. Compared
with conventional methods, FEA is based on computer models and can be quantitatively
analyzed. It has the advantages of non-invasiveness and repeatability, and can also be
combined with big data and deep learning [28]. Medical image analysis based on machine
and deep learning is playing an increasingly important role in disease prevention and
control [29].
However, the finite element (FE) method for the biomechanical analysis of the muscu-
loskeletal system also has limitations. First, the accuracy of the model parameters has a
significant impact on the accuracy of this numerical method for simulation. Additionally,
current FE methods are based on the quasi-static force analysis of different attitudes, and
the dynamic response of the object is simulated by separating the actions, which makes the
dynamic response of human actions difficult to achieve in real-time [30]. The analysis of
individualized biomechanical properties is also difficult. There are great differences in the
biomechanical properties of bones in different individuals and genders [31,32]. Further-
more, the integration of models of the musculoskeletal system with clinical applications
is not well understood or demonstrated. There are many data sources and methods for
constructing FE models; for example, biomechanical CT is one of the state-of-the-art meth-
ods for testing skeletal biomechanics [33]. It can help monitor responses to treatments and
assess surgical risks. However, its effectiveness has only been demonstrated for oppor-
tunistic use, and there are individual differences in the effectiveness and robustness of
thresholds [34].

4. What Is Digital Twin?

Professor Michael Grieves first put forward the concept of DT to describe the life
management of productions [35]. In the early stages, it was initially defined by the Na-
tional Aeronautics and Space Administration as an integrated, multiscale, multi-physics
simulation of an as-built vehicle that mirrors the life of the corresponding flying twin in
the field of aerospace to predict the lifecycle of aircrafts [4]. Then, wider uses emerged in
the engineering and manufacturing fields, including simulation, validation, accreditation,
etc. [5]. In industry, it can build a virtual copy of a product or workshop through big
data, which can interact in real-time with the deep information and physical space that
cannot normally be observed. Tao et al. [6] put forward the concept of a DT workshop
driven by fused twin data to realize iterative operation, optimal management, and planning
and control of workshops. According to the latest research, DT mainly consists of five
dimensions, namely, physical entities, models in the virtual space, fused data, service,
and the connections [36]. Figure 1 shows the basic structures and new technologies of DT.
The new technologies include big data, AI, cloud computing, 5G, and virtual reality (VR),
which act as important interactions of the DT system. At the same time, the advancement
and popularization of novel technologies promote the progress and precision of DT. The
physical entity performs instructions and collects data via sensor devices. Virtual models
are twin models in digital space. DT data are muti-scaled and include the fusion among
them. The service and connections improve the diversity and fidelity of the DT model.
 and popularization of novel technologies promote the progress and precision of DT. The
physical entity performs instructions and collects data via sensor devices. Virtual models
Bioengineering 2023, 10, 627 4 of 11
are twin models in digital space. DT data are muti-scaled and include the fusion among
them. The service and connections improve the diversity and fidelity of the DT model.

Figure 1.
Figure The basic
1. The basic structure
structure and
and key
key technologies
technologies of
of digital
digital twin.
twin. IoT:
IoT: Internet
Internet of
of Things;
Things; VR:
VR: virtual
virtual
reality; AI: artificial intelligence.

In recent
In recent years,
years, the
theidea
ideaof
ofDT
DThas
hasattracted
attractedwidespread
widespreadfocus
focusfrom
frommedical
medicalresearchers,
research-
ers, and it is believed that the application of DT technology in the field of health has
and it is believed that the application of DT technology in the field of health has great
great
promise [7]. DT is data-centric and various complex substructures are related to
promise [7]. DT is data-centric and various complex substructures are related to each other each other
through technologies such as machine learning and AI to finally form a
through technologies such as machine learning and AI to finally form a DT system [37].DT system [37].
At present, the
At present, the DT
DT system has made
system has made certain progress in
certain progress in the
the prevention
prevention andand treatment of
treatment of
cardiovascular, musculoskeletal, and immune system diseases [38,39].
cardiovascular, musculoskeletal, and immune system diseases [38,39]. It has also madeIt has also made
some breakthroughs in imaging diagnosis and radiotherapy [40,41]. Musculoskeletal
some breakthroughs in imaging diagnosis and radiotherapy [40,41]. Musculoskeletal sys-
system diseases are greatly affected by biomechanical factors; thus, the application of the
tem diseases are greatly affected by biomechanical factors; thus, the application of the DT
DT system may be more promising in this area.
system may be more promising in this area.
5. Current Applications of Digital Twin in the Musculoskeletal System
5. Current Applications of Digital Twin in the Musculoskeletal System
The DT system offers new opportunities and possibilities for innovative development
in theThe DT system
medical field. offers new opportunities
High fidelity and real-time anddata
possibilities
collectionforareinnovative
the biggest development
advantages
in the medical field. High fidelity and real-time data collection
of this technology. As the backbone of DT, high fidelity has a significant impact are the biggest advantages
on the
of this technology.
quality of simulation As[42].
the The
backbone of DT,
healthcare high fidelity
applications of has a significant
digital twins focus impact on the
on personal
quality of simulationand
health management [42].precision
The healthcare
medicine, applications of digital
in which reliable twins focus
progress on personal
has been made in
health management
patient recovery [43],and precision
drug medicine,
development [44],indisease
which reliable
treatment progress hasHowever,
[45], etc. been made in
chal-
patient recovery
lenges remain in [43],
areasdrug
suchdevelopment
as standardized [44], disease treatment
modeling, [45],and
data security, etc. data
However,
fusionchal-
[46].
lenges
Real-timeremain
data in areas such
collection canas standardized
realize modeling,
the real-time data security,
monitoring of patients and
to data fusion
analyze [46].
possible
Real-time data collection can realize the real-time monitoring of
abnormal conditions and select individualized treatment plans according to a patient’spatients to analyze pos-
sible abnormal conditions and select individualized treatment plans
condition. It has already been employed for several applications, such as disease prediction according to a pa-
tient’s condition.
and improving It has already
diagnostic accuracy been employed
[47,48]. Therefor areseveral applications,
many categories such asinforma-
of medical disease
prediction and improving
tion, including human datadiagnostic accuracy
from different [47,48].
sources. DataThere
fromarethemany categories
biological body of med-
include
ical information,
computed including
tomography (CT),human data from different
MRI, electrocardiogram sources.
or other Data equipment
scanning from the biological
test data,
body include computed
and biochemical data suchtomography
as blood routine,(CT), urine
MRI, testing,
electrocardiogram
and biological or enzymes.
other scanning
There
equipment test data, and biochemical data such as blood routine, urine testing,
are also virtual simulation data such as disease prediction, surgical simulation, virtual trials, and bio-
logical
etc. The enzymes.
fusion of There
theseare alsotovirtual
data producesimulation
diagnosticdataresults
such asand
disease prediction,
treatment options surgical
must
simulation, virtual trials,
be further studied. However,etc. The fusion
the data of theseofdata
collection the to produce
human diagnostic results
musculoskeletal system andis
mainly about geometric and biomechanical data, giving the application of DT in the mus-
culoskeletal system great advantages. Table 1 compares the advantages and disadvantages
of traditional biomechanical methods and DT techniques in the musculoskeletal system.
Bioengineering 2023, 10, 627 5 of 11

Table 1. Comparison of traditional biomechanical methods and DT techniques in musculoskeletal system.

Method Advantage Disadvantage

More accurate structural features can be presented
Morphology Invasive, ethical and safety issues.
based on anatomical and imaging techniques.
Quantitative evaluation of biomechanical changes in
Sensers Sensor volume and safety issues.
human body parameters through digital simulations.
Low reproducibility of animal models and
Animal model Similar to the human body and avoids ethical barriers.
tissue cultures.
Low simulation accuracy and quasi-static
Finite element analysis Reproducible, quantifiable, and non-invasive.
analysis.
Reproducible, quantitative, personalized, and Lack of standardization, high cost, and
Digital twin
dynamic analysis. immature application.

We first applied this brand-new technology in the musculoskeletal system [8]. During
the construction of the DT system, we used CT images to develop an FE model of the
lumbar spine. The scanning conditions were set as follows: 120 kV, 125 mA, layer thickness
0.625 mm, layer spacing 0 mm, scanning around the spiral from top to bottom. The raw
data were stored in DICOM format and we used Mimics, 3dMax, CATIA V5, and Unity3D
software for hybrid modeling designs. Body movement was calculated by the inverse
kinematics algorithm. The average time delay introduced by the real-time computation
was approximately 24.70 ms, which included 8.04 ms of calculating time and 16.67 ms of
refresh and transmission time of the 60 Hz sensor. It satisfied the real-time requirements of
the biomechanical lumbar spine. As far as we know, this was the first DT example used
in a human musculoskeletal system. Finally, a 3D virtual reality system was developed
with the help of Unity3D software to record the real-time biomechanical performance of
the lumber spine, providing a new and effective method of real-time planning in the field
of spine treatments. Moreover, we used DT construction technology to establish a DT of
a human lumbar spine that dynamically displayed the action status and biomechanical
performance in real-time. It faithfully mapped the health information of the lumbar spine,
demonstrated the digital representation, and could be predicted through simulation. Early
warnings regarding dangerous postures for the human lumbar can help avoid damage to
the spine and provide a more accurate and individualized method for the prevention and
monitoring of spinal diseases.
Other researchers then constructed DTs based on orthopedic surgery models to provide
valuable quantitative information for clinical treatment, optimize treatment methods, and
reduce the occurrence of surgical risks [49,50]. Ahmadian et al. constructed a DT system
that combined computational fluid dynamics and level-set methods to simulate the motion
of the bone cement and bone marrow separation interface [49]. After predicting the cement
morphology, the data co-registration algorithm was used to transform the fluid dynamics
model into a high-fidelity continuous damage mechanics model that enhanced the vertebral
body. It could predict the morphology and fracture response of cement injection within
the bone microstructure. Hernigou et al. developed a more accurate model with DT
technology to improve the method of assessing the downward axis and determining joint
orientation [51]. They used CT data from patients without osteoarthritis to determine the
direction of the subtalar joint axis. Subtalar compensation was assessed based on knee
angle deformity and the direction of the inferior axis, and the inaccuracy of the manual
selection of anatomical markers was greatly reduced. Aubert et al. [52] constructed DT
models using patients’ postoperative 3D X-ray images. They simulated 12 conditions in the
DT system using four stabilizing methods and three bone healing results. Biomechanical
characteristics such as stress distribution in the musculoskeletal system after surgery were
assessed by applying a load during a motion process to assess the risk of recurrent fractures.
Therefore, we are confident that DT will be paid more attention regarding the prevention
and treatment of musculoskeletal diseases in the future.
Bioengineering 2023, 10, 627 6 of 11

6. Disadvantages and Possible Improvements of DT

The use of DT for the biomechanical analysis of the musculoskeletal system is a
further improvement against the limitations of traditional approaches, especially in terms
of technical safety and a lack of real-time dynamic analyses. However, verifying the results
is potentially a significant difficulty for DT. Current comparison methods for modeling and
assessment mainly use cadaveric data or data from previous studies, lacking individualized
assessments. In other words, the results depend on the assumptions that have been made
already. The effective integration of a large number of different component models is also
an urgent problem for the application of DT in the medical field. The accuracy of the
simulation partly determines the effectiveness of the technique. Therefore, data collection
based on multiple dimensions and the use of experimental data for simulation training or
fitting verification will be the main trends of digital simulation in the future. More accurate
validation methods are also needed to allow DT to be more widely used. Currently, there
are difficulties in obtaining real pressure data of the body during movement processes and
fusing different categories of medical information. There is a gap between the results of
model calculation and actual situations, and how to reduce this gap is another difficulty for
DT. For example, there is a possibility that risks not predicted or estimated during modeling
may occur during the process of surgery. Ethical issues are one of the main reasons why DT
may be disadvantageous, and privacy seems to be the most remarkable challenge [53,54].
The standardized definition of DT facilitates a more systematic analysis of ethical risks in
personalized health [55]. Additionally, the cost of DT health systems can lead to injustice in
medical conditions, which may exacerbate existing economic disparities.
In our previous study, DT of the human lumbar spine was designed to dynamically
display action status and biomechanical performance in real-time. It could not be com-
manded to make physical entities work optimally. In further research, technologies such
as exoskeleton devices or other actuators may be available in the human body to carry
out commands through virtual twins so that the human musculoskeletal system functions
more optimally. The development of DT requires the integration of multiple disciplines,
including medicine, engineering, and health law. A standard set of treatment protocols or
tools can facilitate the internationalization of treatment and ethics while enabling patients
to subjectively evaluate approaches. In the future, the accessibility of this approach and the
demand for related products will gradually increase. Safety, efficiency, and ethical practice
will be critical for the development of DT.

7. Discussion
To realize personalized medicine, personalization and patient-centric modeling are
crucial [56]. DT provides a quantitative method concerning diseases, and it is considered
to be a driving force for personalized medicine [57,58]. It offers a potential solution for
personalized diagnosis and the treatment of musculoskeletal system diseases which might
revolutionize the development of medicine. There are many categories of medical infor-
mation, and how to integrate the different categories of data is key to the establishment
of DT in medicine. In general, the technologies behind DT are classified into two broad
categories: mechanical models that integrate multi-scale information and statistical models
that must learn from categories of data [44,59]. Mechanistic models incorporate physiologi-
cal knowledge and the fundamental laws of physics and chemistry, providing a framework
to integrate experimental and clinical data [60]. It can be used in surgical simulation, drug
discovery, and regulatory decision making. Statistical models include all mathematical
methods for inferring relationships from experimental data, using mathematical rules to
extract and optimize combinations of personalized biomarkers [61]. Faster-solving sur-
rogates of models may be developed in the form of machine learning models to solve
problems in real-time, which can help solve the technical hurdles during the creation
of high-fidelity models [62]. This will help accelerate model development and improve
the quality of models through the consideration of feedback from interactions between
various sub-processes.
Bioengineering 2023, 10, 627 7 of 11

Figure 2 shows the connections between the physical and digital twins during the
construction and application of DT. The information in the physical space is acquired by
sensors and CT scans at key locations and then imported into the numerical calculation
system for analysis and calculation. Human reverse dynamics and biomechanics are used
to complete the fusion of various hetero-dimensional data such as human bone movement
information, mechanical information, and human spatial position data. Moreover, the
AI model is constructed by fused data and training data. Finally, through visualization
technology, the calculation results are rendered as high-fidelity digital twins of the human
musculoskeletal system. It can provide a credible digital dynamic model to show the
biomechanical performance of the musculoskeletal system and provide reference data for
subsequent human musculoskeletal medical studies. In the future, we hope to build a DT of
the whole musculoskeletal system and collect various data required by sensors via wearable
VR devices at important parts of the body and key bone positions (Figure 3). The data will
then be processed and imported into the numerical computing system for analysis and
calculation, which is primarily accountable for the security of network system, platform,
medical data, user personal privacy and information [63]. Then, we will use the data
transmission system to encode, read, decode, and write the collected data. Technologies
such as access control, anonymous generalization, storage networking technology, and
dynamic passwords can help the security system effectively prevent malicious attacks from
third parties, data theft, and information tampering [64]. Finally, through the DT display
system, computer graphics technology will be used to render the calculation results into a
high-fidelity DT of the human musculoskeletal system, providing an intuitive and credible
digital dynamic model. The real-time biomechanical performance will be provided on
the digital model, which will be highly compatible with a real human musculoskeletal
system. It will be able to predict the dangerous moments of our bodies in advance through
simulation and prediction so as to avoid damage to the musculoskeletal system
Bioengineering 2023, 10, x FOR PEER REVIEW
and
8 of 12
provide valuable research data and solutions for the prevention, treatment, and monitoring
of musculoskeletal diseases.

Figure 2.
Figure Theconnections
2. The connectionsbetween
betweenphysical
physicaland
anddigital
digitaltwins.
twins.
CTCT = computerized
= computerized tomography;
tomography; AI
AI = artificial
= artificial intelligence.
intelligence.
Bioengineering 2023, 10, 627 8 ofAI
Figure 2. The connections between physical and digital twins. CT = computerized tomography; 11
= artificial intelligence.

Figure 3. Digital twin of the whole musculoskeletal system for personalized diagnosis and treat-
Figure 3. Digital twin of the whole musculoskeletal system for personalized diagnosis and treatment.
ment.
At present, application examples in musculoskeletal systems represent the initial
At present,
exploration application
of DT, and there examples
are stillinmany
musculoskeletal
shortcomings systems represent the
and limitations, initial
such ex-
as the
ploration of DT, and there are still many shortcomings and limitations,
accuracy of simulations and the integration of data from different sources. Additionally, such as the accu-
racy
thereofis simulations
no consensusand on the
the integration
extent to which of data
DTsfrom different sources.
can transform Additionally,
healthcare systems over there
the
is no consensus on the extent to which DTs can transform healthcare systems
next decade due to technical, regulatory, and ethical barriers [65]. New technologies, such over the next
decade duecloud
as big data, to technical,
computing, regulatory,
blockchain,andandethical
VR, barriers
which act[65]. New technologies,
as important interactions such as
of the
big data, cloud computing, blockchain, and VR, which act as important
system, provide more motivation concerning the medical application of DT. The DT system interactions of the
system,
is based provide more motivation
on the acquisition concerning
and simulation the medical data;
of multi-source application
thus, itof
is DT.
veryThe DT sys-
sensitive to
tem
dataisprocessing
based on the acquisition
efficiency and simulation
and required latency.ofUsing
multi-source
laboratory data; thus, itfor
datasets is very sensi-
simulation
tive to data
training processing
or fitting efficiency
verification willand
helprequired
improvelatency. Usingof
the efficiency laboratory datasets forAtsim-
digital simulations. the
ulation training or fitting verification will help improve
same time, modulus-based DT optimization and soft sensor-based efficiency the efficiency of digital simula-
monitoring
tions.help
will At the same time,
to develop DT modulus-based
systems with better DT optimization
performance.and soft sensor-based
However, this requires efficiency
a trade-
monitoring will help to develop DT systems with better performance.
off between the cost of deep learning and the performance of the DT system. Only However, this re-
in
quires
this way a trade-off
can bothbetween
reduce thethe cost
cost ofof deep
time learning and themaking
and materials, performance
DT anof the DT system.
effective tool for
Only in this
analyzing theway can both reduce
musculoskeletal systemthe and
costhuman
of timehealth.
and materials, making DT
In the traditional an effective
framework, the
dependency structure of each module reflects the dependence between biological processes,
which is difficult to modify or extend. The open-source modular software platform for
model simulation and integration proposed by Masison et al. can realize distributed
model construction and integration and support a decentralized, community-based model
construction process [66]. This provides new ideas for the application of DT technology
in the medical field. Accordingly, we hope that this article will inspire researchers to
pursue more scientific applications of DT in the field of medicine, especially regarding the
musculoskeletal system.

8. Conclusions
The application of DT will be increasingly popular in the future of healthcare services
and will become a new platform for the health management of the musculoskeletal system.
Furthermore, DT technology in combination with intelligent medical technologies will have
a significant impact on personalized medicine. Simulations are performed based on the DT
system of the patient to select more accurate, appropriate, and individualized treatments.
In conclusion, it could be the solution for personalized diagnosis and the treatment of
musculoskeletal system diseases.
Bioengineering 2023, 10, 627 9 of 11

Author Contributions: Conceptualization, Z.L.; writing—original draft preparation, T.S. and J.W.;
writing—review and editing, M.S., X.L. and H.H.; literature and data collection, J.Z. and W.Z.; All
authors prepared and contributed to the manuscript. All authors have read and agreed to the
published version of the manuscript.
Funding: This study was supported by the Science and Technology Innovation Foundation of
Dalian (2022JJ12SN045) and the Natural Science Foundation of Liaoning Province (2022-MS-322).
The funders had no role in the study design, data collection and analysis, decision to publish, or
preparation of the manuscript.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Veronese, N.; Maggi, S. Epidemiology and social costs of hip fracture. Injury 2018, 49, 1458–1460. [CrossRef] [PubMed]
2. Safiri, S.; Kolahi, A.A.; Cross, M.; Hill, C.; Smith, E.; Carson-Chahhoud, K.; Mansournia, M.A.; Almasi-Hashiani, A.; Ashrafi-
Asgarabad, A.; Kaufman, J. Prevalence, Deaths, and Disability-Adjusted Life Years Due to Musculoskeletal Disorders for 195
Countries and Territories 1990–2017. Arthritis Rheumatol. 2021, 73, 702–714. [CrossRef] [PubMed]
3. Breen, A.; Breen, A. Uneven intervertebral motion sharing is related to disc degeneration and is greater in patients with
chronic, non-specific low back pain: An in vivo, cross-sectional cohort comparison of intervertebral dynamics using quantitative
fluoroscopy. Eur. Spine J. Off. Publ. Eur. Spine Soc. Eur. Spinal Deform. Soc. Eur. Sect. Cerv. Spine Res. Soc. 2018, 27, 145–153.
[CrossRef] [PubMed]
4. Glaessgen, E.; Stargel, D. The digital twin paradigm for future NASA and US Air Force vehicles. In Proceedings of the 53rd
AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference 20th AIAA/ASME/AHS Adaptive
Structures Conference 14th AIAA, Honolulu, HI, USA, 23 April 2012; p. 1818.
5. Tao, F.; Zhang, H.; Liu, A.; Nee, A.Y. Digital twin in industry: State-of-the-art. IEEE Trans. Ind. Inform. 2018, 15, 2405–2415.
[CrossRef]
6. Tao, F.; Zhang, M.; Cheng, J.; Qi, Q. Digital twin workshop: A new paradigm for future workshop. Comput. Integr. Manuf. Syst.
2017, 23, 1–9.
7. Sun, T.; He, X.; Li, Z. Digital twin in healthcare: Recent updates and challenges. Digit. Health 2023, 9, 20552076221149651.
[CrossRef]
8. He, X.; Qiu, Y.; Lai, X.; Li, Z.; Shu, L.; Sun, W.; Song, X.J.D.T. Towards a shape-performance integrated digital twin for lumbar
spine analysis. Digit. Twin 2021, 1, 8. [CrossRef]
9. Sun, T.; He, X.; Song, X.; Shu, L.; Li, Z. The Digital Twin in Medicine: A Key to the Future of Healthcare? Front. Med. 2022, 9,
907066. [CrossRef]
10. Pathria, M.N.; Chung, C.B.; Resnick, D.L. Acute and Stress-related Injuries of Bone and Cartilage: Pertinent Anatomy, Basic
Biomechanics, and Imaging Perspective. Radiology 2016, 280, 21–38. [CrossRef]
11. Lindsey, D.P.; Parrish, R.; Gundanna, M.; Leasure, J.; Yerby, S.A.; Kondrashov, D. Biomechanics of unilateral and bilateral sacroiliac
joint stabilization: Laboratory investigation. J. Neurosurg. Spine 2018, 28, 326–332. [CrossRef]
12. Nash, K.E.; Ong, K.G.; Guldberg, R.E. Implantable biosensors for musculoskeletal health. Connect. Tissue Res. 2022, 63, 228–242.
[CrossRef] [PubMed]
13. Burger, E.H.; Klein-Nulend, J.; Veldhuijzen, J.P. Mechanical stress and osteogenesis in vitro. J. Bone Miner. Res. 1992, 7 (Suppl. 2),
S397–S401. [CrossRef]
14. Chen, C.N.; Chang, H.I.; Yen, C.K.; Liu, W.L.; Huang, K.Y. Mechanical Stretch Induced Osteogenesis on Human Annulus Fibrosus
Cells through Upregulation of BMP-2/6 Heterodimer and Activation of P38 and SMAD1/5/8 Signaling Pathways. Cells 2022, 11,
2600. [CrossRef] [PubMed]
15. Desmoulin, G.T.; Pradhan, V.; Milner, T.E. Mechanical Aspects of Intervertebral Disc Injury and Implications on Biomechanics.
Spine (Phila Pa 1976) 2020, 45, E457–E464. [CrossRef]
16. Vergroesen, P.P.; Kingma, I.; Emanuel, K.S.; Hoogendoorn, R.J.; Welting, T.J.; van Royen, B.J.; van Dieën, J.H.; Smit, T.H. Mechanics
and biology in intervertebral disc degeneration: A vicious circle. Osteoarthr. Cartil. 2015, 23, 1057–1070. [CrossRef]
17. Gellhorn, A.C.; Katz, J.N.; Suri, P. Osteoarthritis of the spine: The facet joints. Nat. Rev. Rheumatol. 2013, 9, 216–224. [CrossRef]
[PubMed]
18. Dunlop, R.; Adams, M.; Hutton, W. Disc space narrowing and the lumbar facet joints. J. Bone Jt. Surgery. Br. Vol. 1984, 66, 706–710.
[CrossRef]
19. Ke, S.; He, X.; Yang, M.; Wang, S.; Song, X.; Li, Z. The biomechanical influence of facet joint parameters on corresponding segment
in the lumbar spine: A new visualization method. Spine J. Off. J. N. Am. Spine Soc. 2021, 21, 2112–2121. [CrossRef] [PubMed]
Bioengineering 2023, 10, 627 10 of 11

20. Kraeutler, M.J.; Welton, K.L.; Chahla, J.; LaPrade, R.F.; McCarty, E.C. Current Concepts of the Anterolateral Ligament of the Knee:
Anatomy, Biomechanics, and Reconstruction. Am. J. Sports Med. 2018, 46, 1235–1242. [CrossRef]
21. Stordeur, A.; Grange, S.; Servien, E.; Blache, Y.; Klasan, A.; Putnis, S.E.; Boyer, B.; Farizon, F.; Philippot, R.; Neri, T. Optimal
Combination of Femoral Tunnel Orientation in Anterior Cruciate Ligament Reconstruction Using an Inside-out Femoral Technique
Combined with an Anterolateral Extra-articular Reconstruction. Am. J. Sports Med. 2022, 50, 1205–1214. [CrossRef]
22. Hu, B.W.; Lv, X.; Chen, S.F.; Shao, Z.W. Application of Finite Element Analysis for Investigation of Intervertebral Disc Degeneration:
From Laboratory to Clinic. Curr. Med. Sci. 2019, 39, 7–15. [CrossRef] [PubMed]
23. Li, Z.; Chen, D.J.; Liu, Z.; Tang, B.; Zhong, Y.; Li, G.; Wan, Z. Motion characteristics of the lower lumbar spine in individuals with
different pelvic incidence: An in vivo biomechanical study. Clin. Biomech. 2021, 88, 105419. [CrossRef] [PubMed]
24. Wilson, R.L.; Bowen, L.; Kim, W.; Cai, L.; Schneider, S.E.; Nauman, E.A.; Neu, C.P. In vivo intervertebral disc deformation:
Intratissue strain patterns within adjacent discs during flexion-extension. Sci. Rep. 2021, 11, 729. [CrossRef] [PubMed]
25. D'Lima, D.D.; Fregly, B.J.; Colwell, C.W., Jr. Implantable sensor technology: Measuring bone and joint biomechanics of daily life
in vivo. Arthritis Res. Ther. 2013, 15, 203. [CrossRef]
26. Phellan, R.; Hachem, B.; Clin, J.; Mac-Thiong, J.M.; Duong, L. Real-time biomechanics using the finite element method and
machine learning: Review and perspective. Med. Phys. 2021, 48, 7–18. [CrossRef]
27. Li, Y.; Huang, G.; Zhang, X.; Wang, L.; Du, Y.; Lu, T.J.; Xu, F. Engineering cell alignment in vitro. Biotechnol. Adv. 2014, 32, 347–365.
[CrossRef]
28. Zheng, H.D.; Sun, Y.L.; Kong, D.W.; Yin, M.C.; Chen, J.; Lin, Y.P.; Ma, X.F.; Wang, H.S.; Yuan, G.J.; Yao, M.; et al. Deep
learning-based high-accuracy quantitation for lumbar intervertebral disc degeneration from MRI. Nat. Commun. 2022, 13, 841.
[CrossRef]
29. Chen, X.; Wang, X.; Zhang, K.; Fung, K.M.; Thai, T.C.; Moore, K.; Mannel, R.S.; Liu, H.; Zheng, B.; Qiu, Y. Recent advances and
clinical applications of deep learning in medical image analysis. Med. Image Anal. 2022, 79, 102444. [CrossRef]
30. Chen, Y.N.; Chang, C.W.; Li, C.T.; Chang, C.H.; Lin, C.F. Finite element analysis of plantar fascia during walking: A quasi-static
simulation. Foot Ankle Int. 2015, 36, 90–97. [CrossRef]
31. Genant, H.K.; Glüer, C.C.; Lotz, J.C. Gender differences in bone density, skeletal geometry, and fracture biomechanics. Radiology
1994, 190, 636–640. [CrossRef]
32. Warnock, J.M.; Karayiannis, P.N.; Gallagher, N.E.; Hill, J.C.; Beverland, D.E. Are There Gender-Specific Errors in Restoration
of Hip Biomechanics That Affect Outcome following Total Hip Arthroplasty? J. Arthroplast. 2020, 35, 1424–1431. [CrossRef]
[PubMed]
33. Pisu, M.; Kopperdahl, D.L.; Lewis, C.E.; Saag, K.G.; Keaveny, T.M. Cost-Effectiveness of Osteoporosis Screening Using Biome-
chanical Computed Tomography for Patients With a Previous Abdominal CT. J. Bone Miner. Res. 2019, 34, 1229–1239. [CrossRef]
[PubMed]
34. Keaveny, T.M.; Clarke, B.L.; Cosman, F.; Orwoll, E.S.; Siris, E.S.; Khosla, S.; Bouxsein, M.L. Biomechanical Computed Tomography
analysis (BCT) for clinical assessment of osteoporosis. Osteoporos. Int. 2020, 31, 1025–1048. [CrossRef] [PubMed]
35. Grieves, M.; Vickers, J. Digital twin: Mitigating unpredictable, undesirable emergent behavior in complex systems. In Transdisci-
plinary Perspectives on Complex Systems; Springer: Berlin/Heidelberg, Germany, 2017; pp. 85–113.
36. Tao, F.; Zhang, M.; Liu, Y.; Nee, A.Y. Digital twin driven prognostics and health management for complex equipment. Cirp Ann.
2018, 67, 169–172. [CrossRef]
37. Emmert-Streib, F.; Yli-Harja, O. What Is a Digital Twin? Experimental Design for a Data-Centric Machine Learning Perspective in
Health. Int. J. Mol. Sci. 2022, 23, 13149. [CrossRef] [PubMed]
38. Benson, M. Digital Twins for Predictive, Preventive Personalized, and Participatory Treatment of Immune-Mediated Diseases.
Arterioscler. Thromb. Vasc. Biol. 2023, 43, 410–416. [CrossRef]
39. Coorey, G.; Figtree, G.A.; Fletcher, D.F.; Snelson, V.J.; Vernon, S.T.; Winlaw, D.; Grieve, S.M.; McEwan, A.; Yang, J.Y.H.; Qian, P.;
et al. The health digital twin to tackle cardiovascular disease-a review of an emerging interdisciplinary field. NPJ Digit. Med.
2022, 5, 126. [CrossRef]
40. Rahmim, A.; Brosch-Lenz, J.; Fele-Paranj, A.; Yousefirizi, F.; Soltani, M.; Uribe, C.; Saboury, B. Theranostic digital twins for
personalized radiopharmaceutical therapies: Reimagining theranostics via computational nuclear oncology. Front. Oncol. 2022,
12, 1062592. [CrossRef]
41. Pesapane, F.; Rotili, A.; Penco, S.; Nicosia, L.; Cassano, E. Digital Twins in Radiology. J. Clin. Med. 2022, 11, 6553. [CrossRef]
42. Wang, S.; Lai, X.; He, X.; Qiu, Y.; Song, X. Building a Trustworthy Product-Level Shape-Performance Integrated Digital Twin With
Multifidelity Surrogate Model. J. Mech. Des. 2022, 144, 031703. [CrossRef]
43. Rivera, L.F.; Jiménez, M.; Angara, P.; Villegas, N.M.; Tamura, G.; Müller, H.A. Towards continuous monitoring in personalized
healthcare through digital twins. In Proceedings of the 29th Annual International Conference on Computer Science and Software
Engineering, Toronto, ON, Canada, 4–6 November 2019; pp. 329–335.
44. Corral-Acero, J.; Margara, F.; Marciniak, M.; Rodero, C.; Loncaric, F.; Feng, Y.; Gilbert, A.; Fernandes, J.F.; Bukhari, H.A.; Wajdan,
A.; et al. The ‘Digital Twin’ to enable the vision of precision cardiology. Eur. Heart J. 2020, 41, 4556–4564. [CrossRef] [PubMed]
45. Geris, L.; Lambrechts, T.; Carlier, A.; Papantoniou, I. The future is digital: In silico tissue engineering. Curr. Opin. Biomed. Eng.
2018, 6, 92–98. [CrossRef]
Bioengineering 2023, 10, 627 11 of 11

46. Yang, D.; Karimi, H.R.; Kaynak, O.; Yin, S. Developments of digital twin technologies in industrial, smart city and healthcare
sectors: A survey. Complex Eng. Syst. 2021, 1, 3. [CrossRef]
47. Erol, T.; Mendi, A.F.; Doğan, D. The Digital Twin Revolution in Healthcare. In Proceedings of the 2020 4th International
Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT), Istanbul, Turkey, 22–24 October 2020; pp. 1–7.
[CrossRef]
48. Hernigou, P.; Olejnik, R.; Safar, A.; Martinov, S.; Hernigou, J.; Ferre, B. Digital twins, artificial intelligence, and machine learning
technology to identify a real personalized motion axis of the tibiotalar joint for robotics in total ankle arthroplasty. Int. Orthop.
2021, 45, 2209–2217. [CrossRef]
49. Ahmadian, H.; Mageswaran, P.; Walter, B.A.; Blakaj, D.M.; Bourekas, E.C.; Mendel, E.; Marras, W.S.; Soghrati, S. A Digital Twin
for Simulating the Vertebroplasty Procedure and its Impact on Mechanical Stability of Vertebra in Cancer Patients. Int. J. Numer.
Method. Biomed. Eng. 2022, 38, e3600. [CrossRef]
50. Ahmadian, H.; Mageswaran, P.; Walter, B.A.; Blakaj, D.M.; Bourekas, E.C.; Mendel, E.; Marras, W.S.; Soghrati, S. Toward an
artificial intelligence-assisted framework for reconstructing the digital twin of vertebra and predicting its fracture response. Int. J.
Numer. Method. Biomed. Eng. 2022, 38, e3601. [CrossRef]
51. Hernigou, P.; Safar, A.; Hernigou, J.; Ferre, B. Subtalar axis determined by combining digital twins and artificial intelligence:
Influence of the orientation of this axis for hindfoot compensation of varus and valgus knees. Int. Orthop. 2022, 46, 999–1007.
[CrossRef]
52. Aubert, K.; Germaneau, A.; Rochette, M.; Ye, W.; Severyns, M.; Billot, M.; Rigoard, P.; Vendeuvre, T. Development of Digital Twins
to Optimize Trauma Surgery and Postoperative Management. A Case Study Focusing on Tibial Plateau Fracture. Front. Bioeng.
Biotechnol. 2021, 9, 722275. [CrossRef]
53. Bruynseels, K. When nature goes digital: Routes for responsible innovation. J. Responsible Innov. 2020, 7, 342–360. [CrossRef]
54. Popa, E.O.; van Hilten, M.; Oosterkamp, E.; Bogaardt, M.-J. The use of digital twins in healthcare: Socio-ethical benefits and
socio-ethical risks. Life Sci. 2021, 17, 6. [CrossRef]
55. Huang, P.H.; Kim, K.H.; Schermer, M. Ethical Issues of Digital Twins for Personalized Health Care Service: Preliminary Mapping
Study. J. Med. Internet Res. 2022, 24, e33081. [CrossRef]
56. Sisodiya, S.M. Precision medicine and therapies of the future. Epilepsia 2021, 62 (Suppl. 2), S90–S105. [CrossRef]
57. Coorey, G.; Figtree, G.A.; Fletcher, D.F.; Redfern, J. The health digital twin: Advancing precision cardiovascular medicine. Nat.
Rev. Cardiol. 2021, 18, 803–804. [CrossRef] [PubMed]
58. Kamel Boulos, M.N.; Zhang, P. Digital Twins: From Personalised Medicine to Precision Public Health. J. Pers. Med. 2021, 11, 745.
[CrossRef] [PubMed]
59. Peirlinck, M.; Costabal, F.S.; Yao, J.; Guccione, J.M.; Tripathy, S.; Wang, Y.; Ozturk, D.; Segars, P.; Morrison, T.M.; Levine, S.; et al.
Precision medicine in human heart modeling: Perspectives, challenges, and opportunities. Biomech. Model. Mechanobiol. 2021, 20,
803–831. [CrossRef] [PubMed]
60. Lauzeral, N.; Borzacchiello, D.; Kugler, M.; George, D.; Rémond, Y.; Hostettler, A.; Chinesta, F. A model order reduction approach
to create patient-specific mechanical models of human liver in computational medicine applications. Comput. Methods Programs
Biomed. 2019, 170, 95–106. [CrossRef]
61. MacLean, A.L.; Harrington, H.A.; Stumpf, M.P.; Byrne, H.M. Mathematical and Statistical Techniques for Systems Medicine: The
Wnt Signaling Pathway as a Case Study. Methods Mol. Biol. 2016, 1386, 405–439. [CrossRef]
62. Gunasegaram, D.; Murphy, A.; Barnard, A.; DebRoy, T.; Matthews, M.; Ladani, L.; Gu, D. Towards developing multiscale-
multiphysics models and their surrogates for digital twins of metal additive manufacturing. Addit. Manuf. 2021, 46, 102089.
[CrossRef]
63. Lee, C.-C.; Chung, P.-S.; Hwang, M.-S. A Survey on Attribute-based Encryption Schemes of Access Control in Cloud Environments.
Int. J. Netw. Secur. 2013, 15, 231–240.
64. Liu, Y.; Zhang, L.; Yang, Y.; Zhou, L.; Ren, L.; Wang, F.; Liu, R.; Pang, Z.; Deen, M.J. A novel cloud-based framework for the
elderly healthcare services using digital twin. IEEE Access 2019, 7, 49088–49101. [CrossRef]
65. Armeni, P.; Polat, I.; De Rossi, L.M.; Diaferia, L.; Meregalli, S.; Gatti, A. Digital Twins in Healthcare: Is It the Beginning of a New
Era of Evidence-Based Medicine? A Critical Review. J. Pers. Med. 2022, 12, 1255. [CrossRef] [PubMed]
66. Masison, J.; Beezley, J.; Mei, Y.; Ribeiro, H.; Knapp, A.C.; Sordo Vieira, L.; Adhikari, B.; Scindia, Y.; Grauer, M.; Helba, B.; et al.
A modular computational framework for medical digital twins. Proc. Natl. Acad. Sci. USA 2021, 118, e2024287118. [CrossRef]
[PubMed]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.