Content uploaded by Shengli Wei
Author content
All content in this area was uploaded by Shengli Wei on Dec 03, 2021
Content may be subject to copyright.
Computer Methods and Programs in Biomedicine Update 1 (2021) 100014
Contents lists available at ScienceDirect
Computer Methods and Programs in Biomedicine Update
journal homepage: www.elsevier.com/locate/cmpbup
Is Human Digital Twin possible?
✩
Wei Shengli
∗
Anyang Institute of Technology, Anyang, Henan Province 455000, China
Keywords:
Human digital twin
Augmented digital twin
Smart healthcare
While Digital Twin nds its applications in many elds, mainly in advanced manufacturing, PLM (Product Life-
cycle Management), and smart healthcare, attempt of using Digital Twin in full lifecycle management of human
being is discussed. The concept of Augmented Digital Twin is put forward as the basis of the concept of Human
Digital Twin (HDT), which is the core idea of the paper. With the experiences of Digital Twin application in
smart manufacturing, PLM and smart healthcare, and the development of other related technologies such as Data
Mining, Data Fusion Analysis, Articial Intelligence, especially Deep Learning and Human Computer Science, a
conclusion can be drawn naturally, that HDT is an enabling way of full lifecycle health management and it is pos-
sible to construct Human Digital Twin, particularly from the technology view. Comparison between Digital Twin
and Human Digital Twin demonstrates the possibility. The concept, conceptual model and characteristics of HDT
are presented for preparation of its construction. How to construct Human Digital Twin is discussed by proposing
Human Digital Twin System Architecture and Implementation Approach. Also, it is pointed out that there will be
a long way to go because of not only its extreme complexity, but also so many aspects being involved, especially
the security and social ethics problem.
1. Introduction
In modern society, people pay more attention to their healthcare.
Especially in China, with its economy development and living level im-
provement in recent decades, Chinese people have taken healthcare into
consideration gradually. Therefore, high level healthcare and medical
resources are required. The demand for medical treatment has changed
from extensive and batch diagnosis to ne and customized services. In
the past, when a patient went to a hospital, there was only a short time
for a doctor to make diagnoses and treatment, at some times, only a
few minutes, because each doctor had to deal with dozens of patients
or even more than 100 in a work day. That can’t satisfy people’s needs.
Using modern data management, Articial Intelligence, Remote Diagno-
sis and other technologies can alleviate the contradiction between more
patient healthcare demands and less doctor services. Human being is
a complex system. Health management is required to help us manage
our health situation which can reduce the number of times to hospitals
and improve the quality and eciency of diagnosis and treatment. So,
nowadays, more and more attention is paid to smart medicine and health
management. To achieve these, the use of Digital Twin (DT) is an eec-
tive way to solve complex problems. Digital Twin has been mainly used
in manufacturing, product life cycle management, industry 4.0 and so
on since it was put forward. But can it be used for human [ 1 , 2 ]? This pa-
✩ This document is supported by key scientic research project plan of higher school in Henan Province (20B520001).
∗ Corresponding author at: Anyang Institute of Technology, Huanghe Road, Anyang City, Henan Province 455000, China.
E-mail address: weishengli401@126.com
per is mainly about the possibility to construct Human Digital Twin, its
architecture, security and other main issues. The remains of this paper
are organized as follows. Section 2 reviews the origin of Digital Twin,
and presents the concept of Augmented Digital Twin (ADT), an extended
Digital Twin which is the basis of Human Digital Twin that will be dis-
cussed in Section 4 . Section 3 describes Digital Twin applications and
recent researches mainly in smart manufacturing, PLM and healthcare.
Section 4 emphasizes the concept, conceptual model and characteristics
of Human Digital Twin and the similarities and dierences between it
and Digital Twin are executed in Section 5 . Section 6 mainly refers to
how to construct Human Digital Twin while at last conclusion is drawn
in Section 7 .
2. Augmented digital twin
2.1. Origin of digital twin
The germination of the idea of Digital Twin can be traced back to
Apollo Program launched by NASA (National Aeronautics and Space
Administration) in 1970s. In that program, at least two identical space
vehicles were built to allow mirroring the conditions of one of them
during the mission [3] . But the concept of Digital Twin would be con-
tributed to Professor Grievers who used the concept in Product Lifecycle
Management. In 2003 in the course of Product Lifecycle Management at
https://doi.org/10.1016/j.cmpbup.2021.100014
Received 23 February 2021; Received in revised form 6 May 2021; Accepted 18 May 2021
2666-9900/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
( http://creativecommons.org/licenses/by-nc-nd/4.0/ )
W. Shengli Computer Methods and Programs in Biomedicine Update 1 (2021) 100014
Fig. 1. Conceptual model of digital twin.
the University of Michigan in the United States, he put forward the idea
of “Virtual Digital Expression Equivalent to Physical Product ”and gave
the denition “one or a group of digital copies of a specic device can
abstract the real device, and can be used as the basis for testing under
real or simulated conditions ”. The concept comes from the expectation
of more clearly expressing the information and data of the device, hop-
ing to put all the information together for higher-level analyses. At that
time this concept was not called Digital Twin and it was called “Mir-
ror Space Model ”in 2003–2005 [4] and “Information Mirror Model ”in
2006–2010 [5] . However, its conceptual model had all the elements of
Digital Twin, namely physical space, cyber space and the connection or
interface between them, so it can be considered as the rudiment of Dig-
ital Twin [6] . In 2011, Professor Grievers normally quoted the Digital
Twin term created by his co-operator John Vickers in his book “Virtually
Perfect: Driving Innovative and Lean Products through Product Lifecy-
cle Management ”[7] . And from that time, Digital Twin has been used
up to now. According to Professor Grievers, The conceptual model of
Digital Twin basically has three main parts:
1) Physical product in physical space;
2) Virtual counterpart of the physical product in cyber space;
3) Data and information interaction interface between physical space
and cyber space.
Fig. 1 illustrates its conceptual model. In the early stage, it was
mainly used in the eld of military industry and aerospace. For exam-
ple, the US Air Force Research Laboratory and NASA have carried out
the application of aircraft health management and control based on Dig-
ital Twin [ 8 , 9 ]. Lockheed Martin has introduced Digital Twin into the
production process of F-35 ghters to improve the process ow, pro-
duction eciency and quality [10] . Since the concept of Digital Twin
originated from PLM by Grievers, it is always the main research area to
which Digital Twin has been applied. Because of the characteristics of
virtual real integration, real-time interaction, iterative operation, opti-
mization, and full factor/full process/full service data-driven [11] , Dig-
ital Twin has been applied to all stages of product life cycle, including
product design, manufacturing, service and operation[ 11 , 12 ].
After it was put forward and formally named [13] by Grievers, more
and more researchers has noticed the importance of Digital Twin. Liu
Yawei believes that Digital Twin is a digital image of a physical entity
or process essentially that can be updated in real time and give a strong
true presence scene to support the decision-making of various activities
in the life cycle of the physical product. Articial Intelligence, Machine
Learning and sensor technology are involved when that “real ” digital
image model is built. The combination of real data and virtual model
analyses can prevent the occurrence of real problems before they oc-
cur, reduce production interruption and cost, and even make plans for
future activities through simulation. With Digital Twin, engineers can
access real-time data, simulation results and solutions, and eciently
perform hundreds of operational tasks from a long distance [10] . Tao
et al. think that Digital Twin establishes a multi-dimensional, multi-
scale, multi-disciplinary and multi-physical dynamic virtual models of
Fig. 2. Augmented digital twin conceptual model.
physical entities in a digital way to simulate and characterize the at-
tributes, behaviors, rules and so on of physical entities in the real en-
vironment. Conclusion about Digital Twin can be drawn as such from
above description.
1) A digital counterpart in the cyber world of the physical entity in real
world.
2) Two-way communication between the physical entity and the digital
counterpart.
3) The digital counterpart is the representation or model of the physical
entity, and the digital model can simulate the physical entity in real
time in a vivid way.
4) There must be full of smart sensors or other instruments to detect
the attributes, parameters, changes etc. of the physical entity and
the surroundings around it, and the information must be transport
to the digital model in time and in a secret way.
5) The model behaves according to the information just as the real en-
tity since it is modeled in terms of its real part working principles or
mechanisms.
6) Besides behaves as its real part (simulation), the model can validate,
optimize, evaluate, diagnose the real part, and give suggestions, pre-
diction by data analyses, Articial Intelligence (AI), etc. for people
to make decisions and exert controls on the real part if necessary.
The whole system constitutes a feedback loop or a circle [14] .
2.2. Augmented digital twin
Digital Twin can be summarized as a digital copy in the virtual space
of the physical entity in the real physical space, and two parts commu-
nicate with each other. The Digital Twin can carry on simulation, val-
idation, optimization, evaluation, and give suggestions, prediction and
controls to the real entity for people to make decisions, to improve the
performance, to prolong the lifecycle of the physical entity.
But here we will expand the concept of Digital Twin, what is called
Augmented Digital Twin whose conceptual model is as Fig. 2 illustrated.
Unlike other researchers only regarding Digital Twin as a digital coun-
terpart of its real entity, we dene Augmented Digital Twin as a complex
system which not only interacts with its real entity but its surroundings
and other Digital Twins. The Augmented Digital Twin system includes
the digital counterpart and its surroundings, relationship with other Dig-
ital Twins, the physical entity and its surroundings, relationship with
other physical entities. They communicate with each other, change si-
multaneously, interact and aect mutually. The system is complicated
and can update and evolve with time. The reason of expanding the con-
cept of Digital Twin is that we would like to use the Augmented Digital
Twin to human beings, which are living beings and dierent from the
inanimate physical entities. In Section 4 , Human Digital Twin is pro-
posed whose conceptual model is coming from that of Augmented Dig-
ital Twin.
2
W. Shengli Computer Methods and Programs in Biomedicine Update 1 (2021) 100014
3. Application of digital twin
3.1. Why is digital twin so important?
The intelligent or smart manufacturing is one of advanced manufac-
turing and it is the development direction in the future and the goal that
human beings have pursued forever. Essentially intelligent manufactur-
ing is an open information system. In the whole manufacturing process,
a variety of informations are transported, exchanged, treated, and so it
is actually a system for treating informations. The most important way
to achieve smart manufacturing is cyber physical integration which be-
comes possible with the development of Internet of Things (IOT), Cloud
Computing (CC), Big Data Analytics (BDA), and AI [16] . But cyber phys-
ical integration is not an easy thing to achieve. It needs a right way to
realize cyber–physical integration. Digital Twin is consider as the en-
abling technique for solving the fusion of CPS (Cyber Physical System)
[15] . And Qinglin Qi etc. believe that it is the key technology of in-
formation integration for smart manufacturing and it can pave a way
for the cyber physical integration, which has been regarded as an im-
portant bottleneck to achieve smart manufacturing before [17] . Digital
Twin can break the barriers between the physical world and the cyber
world of manufacturing [18] .
Now the development of current technology, such as Mathemati-
cal Modeling, Articial Intelligence, Cloud Computing, Block Chain, Big
Data, Data Analysis, IOT, high-speed network, 5G and so on, provide the
possibility for the development of Digital Twin. From this view, Digital
Twin will nd wide use in the future.
3.2. Main applications and researches in manufacturing
Digital Twin is mainly used in smart manufacturing, PLM, and aero
craft etc. Violeta Damjanovic-Behrendt et al. discuss the design of a
Digital Twin demonstrator for smart manufacturing, following an open
source approach for implementation [19] . Claudio Mandolla et al. ana-
lyze the way to build a Digital Twin through the exploitation of Block
Chain used in aircraft industry theoretically and practically for additive
manufacturing [20] . Guivarch et al. present a new approach to develop
Digital Twin of helicopter dynamic systems [21] . Qinglin Qi, Fei Tao
review and compare Digital Twin and Big Data, and discuss how to in-
tegrate them to promote smart manufacturing [17] . In order to improve
the accuracy and eciency for green material optimal-selection in prod-
uct, a new method driven by Digital Twin is proposed in literature [22] .
Literature [23] discusses an object-oriented, event-driven simulation us-
ing Digital Twin of a exible assembly cell coordinated with a robot
to perform assembly tasks alongside human. Literature [24] presents
a data-driven Digital Twin model based on Deep Learning to predict
machining tool condition. In paper [25] , a new method for product de-
sign, manufacturing, and service driven by Digital Twin is proposed.
The paper [26] presents a concept for the composition of a database
and proposes guidelines for the implementation of the Digital Twin in
production systems in small and medium-sized enterprises. In literature
[27] , an AR application that uses Microsoft HoloLens to visualize the
Digital Twin data of a CNC milling machine in a real manufacturing
environment is presented.
3.3. Applications and researches in healthcare
Digital Twin is not only used in areas above mentioned, it also can
nd its application in healthcare area. With technology development
and the application of Big Data, smart healthcare is put forward and get
more and more focus. As the enabling method of smart healthcare, Digi-
tal Twin has been paid much attention, and there are a lot of researches
about it. Neeraj Kavan Chakshu et al. develop a semi-active human Dig-
ital Twin model for detecting severity of carotid stenosis from head vi-
bration [28] . Vera Yakovchenko et al. discuss automated texting system
on mobile phone text messaging to promote patient self-management
[29] . Camille Nebeker et al. investigate what should be done to advance
digital health research by using health data and Articial Intelligence
[30] . Literature [31] discusses Cardiovascular Disease Prevention and
Management by using wireless technology, sensors, apps, and data. A
decision-making checklist tool is developed to support technology se-
lection in digital health research in [32] . Literature [33] reviews the
latest application of digital intelligent diagnosis and treatment technol-
ogy related to liver surgery in the hope that it may help to achieve
accurate treatment of liver surgery diseases. Adam Palanica et al. de-
scribe a VR(Virtual Reality) technology Health Voyager which can pro-
vide patients with engaging visuals and a VR simulation of their indi-
vidual clinical ndings and help them know more eective transfer of
medical knowledge and comprehension of treatment information [34] .
Costas Boletsis et al. test the assumption that the use of smartwatches
for monitoring physical health aspects of dementia patients can benet
formal home-based care [35] . Wen-Tsung Ku’s research is about the at-
titude of Taiwanese Middle-Aged and Elderly Patients’ Acceptance and
Resistance toward the Health Cloud [36] . A Cardio Twin architecture
for Ischemic Heart Disease detection designed to run on the Edge is pre-
sented by Roberto Martinez-Velazquez et al. [37] . While maintaining
a food record is an essential means of health management, Kiyoharu
Aizawa has developed the smartphone application “FoodLog, ”as a new
framework for food recording [38] . The advent of Articial Intelligence
methodologies paves the way towards smarter healthcare by exploit-
ing new concepts such as Deep Learning. Literature [39] presents an
overview of Deep Learning techniques that are applied to smart health-
care. Since emotion situation is important to people’s healthcare, lit-
erature [40] discusses emotional states’ detection approaches by using
Machine and/or Deep Learning techniques and existing physiological
databases for emotion recognition, and also highlights the challenges
and future research directions in this eld. Hadi Habibzadeh et al. study
the emerging trends in smart healthcare applications as well as the key
technological developments that give rise to these transitions, partic-
ularly emphasize threats, vulnerabilities, and consequences of cyber-
attacks in modern smart healthcare systems and last investigates their
corresponding proposed countermeasures [41] . Amrata Bagaria, et al.
give an overview of the existing literature on Digital Twin for personal
health and well-being [42] . Literature [43] provides an overview of
Medical Cyber Physical Systems and the data ow through them which
are platforms to gather, process data from patients with IOT, ML, Cloud
Computing etc., to monitor patients’ healthcare remotely, and to im-
prove quality of life of patients. Document [44] encompasses a brief
literature review on TEC (Technology-Enabled Care) challenges, with
a focus on the key technologies enabling the development of wearable
solutions for remote human motion tracking. Document [45] explores
the most relevant aspects in relation to the outcomes and performances
of the dierent components of a healthcare system with a particular fo-
cus on mobile healthcare applications. Electroencephalography (EEG)
motor imagery signals have recently gained signicant attention due to
its ability to encode a person’s intent to perform an action. Researchers
have used motor imagery signals to help disabled persons control de-
vices, such as wheelchairs and even autonomous vehicles. Hence, the
accurate decoding of these signals is important to Brain–Computer In-
terface (BCI) systems. Deep Learning methods, such as the Convolution
Neural Network (CNN), have achieved remarkable success in computer
vision tasks. Literature [46] proposes a multiple-CNN feature fusion
architecture to extract and fuse features by using subject-specic fre-
quency bands. Abdulmotaleb El Saddik presents a Digital Twin called
Ecosystem-Dtwin to promote human health and well-being [47] . Health-
care applications depend on voluminous data containing rich and impor-
tant insights that can support predictive analysis by discovering these
patterns using Machine Learning algorithms. Unfortunately this data is
not always available for multiple reasons including lack of unied and
standardized architecture for distribution/dissipation of such data. Doc-
ument [48] proposes a framework for Cloud based architecture of a dig-
3
W. Shengli Computer Methods and Programs in Biomedicine Update 1 (2021) 100014
Fig. 3. HDT conceptual model.
ital ecosystem for storing, sharing and predictive analysis of healthcare
data.
4. Human digital twin
4.1. Concept of human digital twin
Documents in Section 3.3 develop related researches of smart health-
care based on Digital Twin. Few of them take full lifecycle of human be-
ing into consideration, and they seldom present the concept of Human
Digital Twin. They just apply Digital Twin to healthcare or a certain dis-
ease, such as heart disease. Here, Human Digital Twin is presented. It
is a copy or counterpart in cyber space of a real person in our physical
world. You can think there is another of you in the cyber space, just as
your soul, but it isn’t a soul actually and it doesn’t know what you think
about. It is just like your incarnation in the cyber space and essentially a
model or database which records your information such as age, weight,
sex, your relatives etc. It is the digital description of you in the digital
manner in a computer or a server in the Cloud. When the information
about you changed, the records change accordingly. Information syn-
chronization depends on the communication between the real person
and the cyber space through Internet, 4G, 5G, WIFI, etc. technologies.
Smart wearing sensors, mobile phones, hospitals or other institutes ob-
tain your information continuously, which will be transferred to the cy-
ber space consecutively, and the information recorded in the database
update accordingly. It doesn’t only record your data, it do a complicated
work. The model analyzes the timely data, historic data, the data from
your relatives and dig out insight from these data by Cloud Comput-
ing, Deep Learning etc. technologies. And then it will provide feedback
information such as diagnoses, prediction or other suggestions to you.
4.2. HDT conceptual model
HDT is based on the Augmented Digital Twin model, and Fig. 3 illus-
trates its model. Besides the two parts, i.e. the physical entity, the virtual
counterpart and the two-way communication between them, surround-
ings and other entities (here are other real persons) are added to the
physical space, and virtual surroundings and other Digital Twins are
added to the cyber space, respectively. The reason for adding those fac-
tors is as follows.
a) Environmental factors have greater inuence on people than on ma-
chines, so environment, circumstance or surroundings is important
to human beings.
b) People are social creatures and need to communicate with others. We
all have relatives, such as parents, brothers, sisters, etc. They have
more similar genes which can be inherit generation to generation.
Adding these factors will help to describe the Human Digital Twin
more comprehensively. In fact, smart manufacturing can also use the
Augmented Digital Twin model, because the same batch of products
produced in the same factory also have some certain common features
which products not produced in the factory or not belong to the same
batch don’t have. For example, products from the same batch may have
similar failures. The Augmented Digital Twin model expands the Digital
Twin concept and can be used to describe the relationship between the
physical space and cyber space comprehensively.
4.3. Characteristics of HDT
1) Each of us in the real world has a corresponding HDT stored in the
cyber space. Each of it has a unique index, which can be used as its
ID and can also be used as an account to log in the HDT.
2) As soon as a person is born, his (let’s assuming the person is a male)
HDT will be created by a qualied person who may be a doctor,
his parents or other experts in a hospital or other institutes. It will
accompany the whole life of the person. Next generations of the per-
son will inherit some of his biological genetic characteristics. And
of course, he inherits biological genetic characteristics from his an-
cestors too. His HDT can inherit these feature from his ancestors’ di-
rectly. People with consanguinity may have similar diseases or char-
acteristics which can be used as one of the basis of diagnoses.
3) Each real person and the corresponding Human Digital Twin would
change synchronously. All the change of the person himself (called
internal factors, e.g. height growth), the inuence on him (called ex-
ternal input, such as treatment, vaccine, injury, etc.) and his response
(called output, such as stimulation, anger, movement, laughter, etc.)
will be transmitted to the cyber space, and the Human Digital Twin
will change accordingly.
4) The data of each examination, treatment (medication, injection,
etc.), immunization (vaccination), physical examination, etc. in
medical institutions which are in the form of text (such as diagnosis
record), image (such as electrocardiogram), number (such as blood
pressure, height and weight of the person) will be transported to the
cyber space by the sta of the medical institutions with the consent
of the real person and the HDT update the data accordingly.
5) Some attributes of the real person, such as age, height will increase
naturally and gradually, and at the same time, according to the law
of human growth, the Human Digital Twin will “grow ” too, which
can be corrected by the information coming from its real counter-
part. For example, height, weight can change according to some kind
of human growth law, and can be collected whenever these data are
measured and sent to the Human Digital Twin by smart equipment.
6) All kinds of wearable sensor data, such as weight, blood pressure,
pulse, heart rate, respiration, blood glucose, exercise volume and
emotional changes, are sent to the Human Digital Twin through com-
munication technology. Family eating habits or recipes and family
buying habits are also transmitted to it for data analyses and inte-
gration. All kinds of education are also imported into the Human
Digital Twin. Also, surroundings data are sent to cyber space, and
accordingly, surroundings in cyber space change synchronously. If
necessary, surroundings Digital Twin can be constructed in the cy-
ber space. This surroundings Digital Twin is just the surroundings
as shown in Fig. 3 . Some information must keep consistency in the
related HDTs and surroundings. If some information changing can
break the consistency, Block Chain technology must be introduced
to trace the information changing and eliminate the inconsistency,
so that related data will return consistency again.
7) The Human Digital Twin can use a variety of data and information
to evaluate the human situation, and give evaluation opinions to
the real person and hospitals related to him. With the help of Ex-
pert Knowledge Systems, health assessment and disease diagnoses
are carried out, and information is fed back to users. The concerned
technologies will include Big Data Analysis, Edge Computing, Inter-
net of Things, Data Fusion, Intelligent Diagnosis, Clustering Analysis,
Articial Neural Network and Decision Tree and so on. The results of
4
W. Shengli Computer Methods and Programs in Biomedicine Update 1 (2021) 100014
health assessment will be fed back to the real person and his doctors
in a hospital as prediction, suggestion, guidance, alarm, or treatment
schedule. These information allows the real person to make certain
improvements, such as strengthening exercise, improving food struc-
ture, further treatment, etc. In case of emergency, his hospital will
dispatch an ambulance immediately and rst aid measure will be
taken.
8) One or his authorized people can login his Human Digital Twin sys-
tem to look into the content. Through VR/AR technology, it can pro-
vide a vivid 3D image of the real person, and can display the health
situation dynamically, visually. One even can observe his internal
organs which are generated by 3D Reconstruction Technology.
9) To maintain privacy, password, ngerprint, iris recognition, autho-
rized login, block chain encryption, authentication, identication,
and other technology can be used for encryption to prevent tamper-
ing. Each HDT needs an account and password to log in and to view
across it. For adults, the account and password can be controlled
by themselves. For children, the account and password can be con-
trolled by their parents or guardians. When necessary, doctors or
professionals can be authorized to inquire about or modify it.
10) Other security mechanism: for trusted access, medical institutions
and individual need to carry out identity authentication and vulner-
ability detection to login the HDT system. For secure trusted com-
munication connection, real address authentication, and xed path
transmission and authentication are needed, that is, medical institu-
tions and home IP addresses are usually xed and need to be regis-
tered; communication path and routing path are usually xed. For
secure and reliable data access, using Edge Computing and cross line
access control to provide data access security. Now medical institu-
tions, families and personal data nodes have the computing power
and storage capacity and can run security algorithms to ensure se-
curity and verication. In order to ensure the reliability, medical
institutions, families, node devices, routing devices and personal ter-
minals need to be registered, and each address identity is embedded
into the address for traceability. After listening to the address assign-
ment in the access network, the device IP and communication port
are dynamically bound to realize the source address trust worthiness.
5. Comparison between DT and HDT
5.1. Commons between DT and HDT
There are some similarities between Human Digital Twin and Digital
Twin.
1) They are all digital replica in virtual cyber space of its real entity in
the physical space.
2) Both need two-way information transmission. It needs to detect the
entity parameters and send them to the cyber space, and the cyber
space needs to send feedback information to the entity.
3) Models need to be built. Digital Twin is the representation or model
of the real entity, which is the most important work when construct-
ing Digital Twin. The model simulate the working mechanism of the
entity. It analyzes, summarizes and judges the input data, and gives
feedback to the entity, and the whole system constitute a closed loop.
4) Security technology is needed to ensure privacy, reliability and in-
tegrity.
5) They share some common technologies, such as Cloud Server, Net-
work, Data Collection, Data Clustering Analysis, Big Data, Edge Com-
puting, Deep Learning, etc.
5.2. Differences between DT and HDT
1) Human beings have mental activities which can be reected by the
fusion of blood pressure, respiration, heart rate, hormone levels and
other data. Detection of mental activities usually relies on biosensors
and brain computer science.
2) People’s reactions are both objective and subjective which can be
supported by research on psychology and human behavior.
3) Human beings are more complex entities and more dicult to rep-
resent in cyber space. Modeling, Data Analysis and Data Fusion re-
quire much more knowledge than what needed when constructing
Digital Twin. For general Digital Twins, modeling depends on math-
ematics, physics, chemistry, mechanics, materials science, etc. and
building HDT models need to focus on physiology, psychology, bi-
ology, chemistry, mathematics and other disciplines or rules. People
know more about laws of physics, chemistry, mechanics, material
science, etc. but know little about laws of human beings. Many laws
of human beings are statistical, not rigid expression.
4) Human beings have the characteristics of heredity and variation and
so a special Data Structure or Class is needed to incorporate these
attributes. It is more dicult to construct HDT models. From the
view of computers, the HDT models in cyber space is a Data Struc-
ture, commonly a Class with inheritance attributes just as referred
in C ++ or Java. Each Class comes from a base Class and can inherit
methods and variables from its parent Class. Each Class needs an ID
to distinguish from others. Not only the Class variables and meth-
ods can be inherited, but also the values of some variables, even the
values of Cross Variables (for example, some eigenvalues come from
both parents, even grandparents) can be inherited. How to design
this Class to better describe human genetic relationship should to be
take into consideration. For example, how recessive gene/dominant
gene can be inherited automatically in a Class is a subject worthy of
study.
5) Human beings are social creatures and more social ethics and inter-
personal relationships need to be considered.
6) Human beings are easily aected by the environment and easily in-
uenced by the region diet. Compared with the machines which can
be aected by temperature, humidity, noise, human being can be af-
fected by the atmosphere, mood, virus, bacteria and air quality and
so on.
6. Constructing a HDT
In this section, how to construct HDT is discussed. In order to in-
terpret the way to construct HDT system, the information owchart of
HDT system is produced rstly, and then HDT Application Layer Ar-
chitecture and its detailed Architecture are presented. At last, System
Implementation Approach is elaborated.
6.1. Information flow chart of HDT system
Fig. 4 illustrates the information owchart of HDT system. A HDT in
the cyber space is responding to the real person in the physical space. All
kinds of data e.g. age, weight, height, ECG, blood pressure, diet, mood
etc. are collected by numerous IOT devices which are common nowa-
days, and may become more widespread, more intelligent, smarter and
transmitted to a network Sink Center(commonly a smart phone to un-
dertake the job), which transfer all the data to the remote Cloud Server.
The Sink Center can also execute Edge Computing sometimes when it is
necessary [37] . In literature [37] , a Deep Learning model is constructed
based on Tensor Flow to deal with the ECG coming from the person,
then judge if the person has a coronary heart disease. If it believes the
person being at the risk of coronary disease, it send alarm to the person,
his family and the nearest emergency hospital to implement a treatment.
This is very important, because early detection and early treatment will
achieve better therapeutic eect.
6.2. HDT system architecture
HDT is usually deployed in the Cloud (SaaS, PaaS or IaaS Platform
provide by infrastructure suppliers, such as Azure, Baidu, etc. as illus-
trated in Figs. 5 and 6 ) which is supported by remote distributed com-
5
W. Shengli Computer Methods and Programs in Biomedicine Update 1 (2021) 100014
Fig. 4. Information ow chart of HDT system.
Fig. 5. HDT application layer architecture.
Fig. 6. Detailed HDT application layer architecture.
puter servers. Besides obtaining the data coming from the person di-
rectly, HDT also obtains a verity of data in diverse form from other
resources, mainly from hospitals. These data may be in many kinds of
forms such as digital, analog, text, gure, image, video, audio, which
should be cleaned, pre-processed and normalized if necessary. And they
are analyzed or input to some mathematic models, or sometimes CNN
models to calculate or simulate. The procedure of analysis or simulation
would be very complicated, and many methods, approaches or technol-
ogy e.g. Data Fusion Analysis, Data Mining, Regression Analysis, Opti-
mization, Experts Systems, AI/DL may be used. After that, results which
Fig. 7. Three layer architecture.
may be prediction, suggestions or diagnoses will be output to the person
or hospitals to take measures to solve the problems.
The HDT Application Layer Architecture is illustrated by Fig. 5 . Each
of the seven layers undertakes certain functionality which are detailed in
Fig. 6 , the Detailed HDT Application Layer Architecture. The Data Col-
lection Layer collects data such as Blood Pressure, Temperature, ECG,
Mood, Movement, and etc. from wearable smart devices, Food Log, Hob-
bies, other social activities from smart phone, and Daily Records, In-
spection, Treatment Records etc. from hospitals or other healthcare in-
stitutes. These data need to be pre-processed, normalized in the smart
devices for subsequent processing since many of which have the capabil-
ity of nishing these kind of job. The pre-processed or normalized data
will be gather to a Sink Center (usually a smart phone or a gateway)
through LANs (Local Area Networks), e.g. BAN (Body Area Network),
Blue Tooth, WIFI etc. The Sink Center transfers them to HDT deployed in
remote Cloud through Internet. Sometimes, Edge Computing is needed
when it is necessary just as document [37] presents. All data are stored
in the Cloud Database, where many mathematical digital models, repre-
sentations and computing platforms which construct the core of the HDT
are deployed to provide function interface to the Application Layer. The
Application Layer provides healthcare management, disease diagnosis,
exercise suggestion, diet recommendation etc. to users.
Usually, the system may adopt B/S (Browser/Server) mode and
Three Layer Architecture as shown by Fig. 7 , i.e. Client, Web Server,
Data Server. The Client may be Browser, Apps, or Applets, etc. and the
Web Server may be IIS, Tomcat, etc. The Data Server may be SQL Server,
Oracle, Distributed DBMS or other Cloud Servers. Other IT technology
concerned will include: HTML, XML, Java, Java Script, Hadoop, python,
Jason, Edge Computing, and Cloud Computing etc. One can login his
HDT system by a browser, or an APP to view his health situation. He
can look up all data, or even investigate his internal organs in a vivid,
visually way with the technology such as hologram photography and 3D
reconstruction.
6.3. System implementation approach
Fig. 8 shows the System Implementation Approach. All kinds of wear-
ables, smart devices nd their use in human beings. In the future, more
and more these kinds of equipment will be used. More and more in-
formation in verity of diverse forms are collect by them. These infor-
mation is send to the Sink Center usually through wireless communi-
cation technology, e.g. BAN, Blue Tooth, Lora, ZigBee, WIFI etc. The
Sink Center maybe a PDA, iPad, and more commonly a smart phone.
It works as a gateway, and transfer data to the cyber space through In-
ternet. People can’t stop going to hospitals from birth, where are places
getting amount of data from people. The inspection data, diagnoses,
treatments, ECG, blood pressure, EMG(Electromyogram), NMR(Nuclear
6
W. Shengli Computer Methods and Programs in Biomedicine Update 1 (2021) 100014
Fig. 8. System implementation approach.
Magnetic Resonance), blood sugar, EEG, CT(Computed Tomography),
PET (Positron Emission Tomography), urinalysis, blood biochemistry,
are also used as input to the HDT through Internet, VPN (Virtual Pri-
vate Network), 5G, etc. In the process, encryption technology should be
adopted to prevent information leakage.
7. Conclusion
Human beings are more complex systems than any machines or man-
ufacturing systems. For human full lifecycle management, it is far more
dicult. Fortunately, Digital Twin is ideal way to solve this kind of
problems [ 14 , 49 , 50 ]. That is why we discuss Human Digital Twin here.
Though the construction of HDT is a hard task, but with rapid promo-
tion of computing capacity, high performance and cheap smart devices,
big data storage, convenient data acquisition, development of AI, Digital
Twin enabling technology and the leap of understanding ourselves, HDT
must have a bright future. We have adequate reason to believe that it is
possible to construct Human Digital Twin, at least from the technology
view. Initially, it may be far away from the goal, but when time goes
on, and with technology development, it will be closer and closer to the
goal. Still, it has a long way to go because of some challenges which
include the complexity of human being, the diculty of modeling, mas-
sive data fusion analysis, the diversity of data sources, data variability
and heterogeneity. But the most challenge may be the social ethic issues
and people’s worrying about the safety.
Declaration of Competing Interest
We declare that we have no nancial and personal relationships with
other people or organizations that can inappropriately inuence our
work, there is no professional or other personal interest of any nature or
kind in any product, service and/or company that could be construed as
inuencing the position presented in, or the review of, the manuscript
entitled “Is Human Digital Twin Possible? ”.
Reference
[1] T. Fei , L. Weiran , Z. Meng , et al. , Five-dimension digital twin model and its ten
applications, Comput. Integr. Manuf. Syst. 25 (1) (2019) 1–18 .
[2] N. Bagaria, F. Laamarti, H.F. Badawi, et al. Health 4.0: Digital Twins for Health and
Well-Being. doi: 10.1007/978-3-030-27844-1 . (2020):143-152.
[3] R. Rosen , G. Wichert , G. Lo , R. Rosen , et al. , About the importance of autonomy
and digital twins for the future of manufacturing, IFAC-Papers OnLine 48-3 (2015)
567–572 .
[4] M. Grieves , Product lifecycle management: the new paradigm for enterprises, Int. J.
Prod. Dev. 2 (1/2) (2005) 71–84 .
[5] M. Grieves , Product Lifecycle management: Driving the Next Generation of Lean
Thinking, McGraw Hill, New York, N. Y., USA, 2006 .
[6] Z. Cunbo , L. Jianhua , X. Hui , et al. , Connotation, architecture and trends of product
digital twin, Comput. Integr. Manuf. Syst. 23 (4) (2017) 753–768 .
[7] M. Grieves , Virtually Perfect: Driving Innovative and Lean Products Through Product
Lifecycle Management, Space Coast Press, Cocoa Beach, Fla., USA, 2011 .
[8] E. Glaessgen , D. Stargel , The digital twin paradigm for future NASA and U.S. air force
vehicles, in: Proceedings of the 53rd AIAA/ASME/ASCE/AHS/ASC Structures, Struc-
tural Dynamics and Materials Conference, AIAA, Reston, Va., USA, 2012 7274-7260 .
[9] E.J. Tuegel , A.R. Ingraea , T.G.
Eason , et al. , Reengineering aircraft structural life
prediction using a digital twin, Int. J. Aerosp. Eng. 2011 (2011) 1–14 .
[10] L. Yawei. Lockheed martin used digital twin to speed up production of F-35[EB/
OL]. (2017-12-27). http://www.sohu.com/a/212980157_613206 . (2017-12-27, in
Chinese).
[11] F. Tao , J. Cheng , Q. Qi , et al. , Digital twin-driven product design, manufactur-
ing and service with big data, Int. J. Adv. Manuf. Technol. 94 (2018) 3563–3576
(9/10/11/12) .
[12] T. Fei , Z. Meng , C. Jiangfeng , et al. , Digital twin workshop: a new paradigm for
future workshop, Comput. Integr.
Manuf. Syst. 23 (1) (2017) 1–9 (in Chinese) .
[13] M. Grieves, Digital twin: manufacturing excellence through virtual factory replica-
tion, 2015 https://www.researchgate.net/publication/275211047 .
[14] K.M. Alam, AEl Saddik, C2PS: a digital twin architecture reference model for the
cloud-based cyber-physical systems, 2017, doi: 10.1109/ACCESS.2017.2657006 .
[15] F. Tao , H. Zhang , A. Liu , et al. , Digital twin in industry: state-of-the-art, IEEE Trans.
Ind. Inform. 24 (1) (2018) 1–18 .
[16] F. Tao, Q. Qi, L. Wang, et al., Digital twins and cyber–physical systems to-
ward smart manufacturing and industry 4.0: correlation and comparison, 2019,
doi: 10.1016/j.eng.2019.01.014 .
[17] Q. Qi, F. Tao. Digital twin and big data towards smart manufacturing and
industry 4.0: 360 degree comparison. vol. 6, 2018. Digital Object Identier
10.1109/ACCESS.2018.2793265:3585-3593.
[18] J. Liu, X. Du, H. Zhou, et al., A digital twin-based approach for dynamic clamping
and positioning of the exible tooling system, in: 26th CIRP Life Cycle Engineering
(LCE) Conference, 2019, doi: 10.1016/j.procir.2019.01.063 .
[19] V. Damjanovic-Behrendt, W. Behrendt, An open source approach to the design and
implementation of digital twins for smart manufacturing, Int. J.Comput. Integr.
Manuf. (2019), doi: 10.1080/0951192X.2019.1599436 .
[20] C. Mandolla, A. Messeni, Petruzzelli, et al., Building a digital twin for
additive man-
ufacturing through the exploitation of blockchain: a case analysis of the aircraft
industry, Comput. Ind. 109 (2019) 134–152, doi: 10.1016/j.compind.2019.04.011 .
[21] D. Guivarch, E. Mermoz, Y. Marino, et al., Creation of helicopter dynamic systems
digital twin using multibody simulations, 2019, doi: 10.1016/j.cirp.2019.04.041 .
[22] F. Xiang, Z. Zhang, Y. Zuo, et al., Digital twin driven green material optimal-selection
towards sustainable manufacturing, 52nd CIRP Conference on Manufacturing Sys-
tems, 2019, doi: 10.1016/j.procir.2019.04.015 .
[23] A. Bilberg, A.A. Malik, Digital twin driven human–robot collaborative assembly.
https://doi.org/10.1016/j.cirp.2019.04.011.
[24] Q. Qiao, J. Wang, L. Ye, Digital twin for machining tool condition
prediction, 52nd CIRP Conference on Manufacturing Systems, 2019,
doi: 10.1016/j.procir.2019.04.049 .
[25] F. Tao, J. Cheng, Q. Qi, et al., Digital twin-driven product design, manufacturing
and service with big data, 2017, doi: 10.1007/s00170-017-0233-1 .
[26] T.H.-J. Uhlemann, C. Lehmann, R. Steinhilper, The digital twin: realizing the cyber-
physical production system for industry 4.0, The 24th CIRP Conference on Life Cycle
Engineering, 2017, doi: 10.1016/j.procir.2016.11.152 .
[27] Z. Zhu, C. Liu, X. Xu, Visualisation of the digital twin data in manufacturing by
using augmented reality, 52nd CIRP Conference on Manufacturing Systems, 2019
http://creativecommons.org/licenses/by-nc-nd/3.0/ .
[28] N.K. Chakshu, J. Carson, I. Sazonov, et al.,
A semi-active human digital twin model
for detecting severity of carotid stenoses from head vibration - acoupled computa-
tional mechanics and computer vision method, 2019, doi: 10.1002/cnm.3180 .
[29] V. Yakovchenko, T.P. Hogan, T.K. Houston, Automated text messaging with patients
in department of veterans aairs specialty clinics: cluster randomized trial, J. Med.
Internet Res. (2020) https://www.jmir.org/2019/8/e14750 .
[30] C. Nebeker, J. Torous, J. Rebecca, et al., Building the case for actionable
ethics in digital health research supported by articial intelligence, 2019,
doi: 10.1186/s12916-019-1377-7 .
[31] S. Wongvibulsin, S.S. Martin, S.R. Steinhubl, et al., Connected health
technology for cardiovascular disease prevention and management, 2019
https://doi.rog/10.1007/s11936-019-0729-0 .
[32] C. Nebeker, R.J.B Ellis, J. Torous, Development of a decision-making check-
list tool to support technology selection in digital health research, 2019,
doi: 10.1093/tbm/ibz074 .
[33] C. Fang, P. Zhang, X. Qi, Digital and intelligent liver surgery in the new era: prospects
and dilemmas, 2019, doi: 10.1016/j.ebiom.2019.02.017 .
[34] A. Palanica, M.J. Docktor, A. Lee, et al. Using mobile virtual reality to en-
hance medical comprehension and satisfaction in patients and their families.
doi: 10.1007/s40037-019-0504-7 .
[35] C. Boletsis, S. McCallum, Brynjar Fowels landmark. The use of smart
watches for health monitoring in home-based dementia care, 2015,
doi: 10.1007/978-3-319-20913-5_2 .
[36] W.-T. Ku, P.-J. Hsieh, Taiwanese middle-aged and elderly patients’ acceptance and
resistance toward the health cloud, 2015, doi: 10.1007/978-3-319-20913-5_9 .
[37] R. Martinez-Velazquez , R. Gamez , A.El. Saddik , Cardio twin: a digital twin of the
human heart running on the edge, 2019 IEEE International Symposium on Medical
Measurements and Applications (MeMeA), 2019 .
[38] K. Aizawa , Image recognition-based tool for food recording and analysis: foodlog,
2019 https://doi.org/10.1007/978-3-030-27844-1_1 .
7
W. Shengli Computer Methods and Programs in Biomedicine Update 1 (2021) 100014
[39] M. Simsek, A.A. Obinikpo, B. Kantarci, Deep learning in smart health: methodolo-
gies, applications, challenges, 2019, doi: 10.1007/978-3-030-27844-1_3 .
[40] D.P.T. Vallejo, A.E. Saddik, Emotional states detection approaches based
on physiological signals for healthcare applications: a review, 2019,
doi: 10.1007/978-3-030-27844-1_4 .
[41] H. Habibzadeh, T. Soyata, Toward uniform smart healthcare ecosys-
tems: a survey on prospects, security, and privacy considerations, 2019,
doi: 10.1007/978-3-030-27844-1_5 .
[42] A. Bagaria, F. Laamarti, H.F. Badawi, et al., Health 4.0: digital twins for health and
well-being, 2019, doi: 10.1007/978-3-030-27844-1_7 .
[43] O.R. Shishvan, D.-S. Zois, T. Soyata, Incorporating articial intelligence into medical
cyber physical systems:
a survey, 2019, doi: 10.1007/978-3-030-27844-1_8 .
[44] P. Daponte, L.D. Vito, G. Mazzilli, et al., Technologies for motion measurements in
connected health scenario, 2019, doi: 10.1007/978-3-030-27844-1_10 .
[45] G. Sannino, G. De Pietro, L. Verde, Healthcare systems: an overview of the
most important aspects of current and future m-health applications, 2019,
doi: 10.1007/978-3-030-27844-1_11 .
[46] S.U. Amin, M. Alsulaiman, G. Muhammad, Deep learning for EEG motor imagery-
based cognitive healthcare, 2019, doi: 10.1007/978-3-030-27844-1_12 .
[47] A.E. Saddik , H. Badawi , R.A.M. Velazquez , et al. , Dtwins: a digital twins ecosystem
for health and well-being, IEEE COMSOC MMTC Commun.–Front. 14 (2) (2019)
March .
[48] B. Qureshi, Towards a digital ecosystem for predictive healthcare analytics, 2014,
doi: 10.1145/2668260.2668286 .
[49] R. He, G. Chen, C. Dong, et al., Data-driven digital twin technology for optimized
control in process systems, 2019, doi: 10.1016/j.isatra.2019.05.011 .
[50] R. Stark, C. Fresemann, K. Lindowa, Development and operation of digital
twins for technical systems and services, CIRP Ann.–Manuf. Technol. (2019),
doi: 10.1016/j.cirp.2019.04.024 .
8
