Developer Driven Framework for Security and Privacy in the IoMT
Ceara Treacy
a
, John Loane
b
and Fergal McCaffery
c
Regulated Software Research Centre & Lero, Dundalk Institute of Technology, Dundalk, Ireland
Keywords: Security, Privacy, IoMT, Threat Modeling.
Abstract: The Internet of Medical Things (IoMT), is a fast growing domain as healthcare moves out of structured health
services into care in the community. As a result, the sensitive personal and health data associated with the
IoMT can potentially flow through a diversity of apps, systems, devices and technologies, public and open
networks. This exposes data in the IoMT to additional attack surfaces, which requires the hardening of the
security and privacy of the data. Accordingly, the data is bound by regulatory safety, security and privacy
requirements. Applying the regulatory compliant requirements is a struggle for developers in small to medium
enterprises due to lack of knowledge, experience and understanding. This paper proposes a framework to
assist in meeting regulatory compliance for security and privacy of data in flow in the IoMT, directed at
developers in small to medium enterprises. The framework considers both security and privacy properties for
data in flow protection in the IoMT. This framework expands on the established threat modeling steps to
consider both security and privacy. To mitigate the identified security and privacy threats, the framework
includes a set of categorised technical security and privacy controls developed through medical device
security standards. The originality of this framework is the inclusion of security and privacy requirements in
the extension of the traditional threat modeling process, as well as the security and privacy controls embedded
in the medical security standards.
1 INTRODUCTION
The Internet of Medical Things (IoMT), is a
connected system, consisting of a variety of networks,
medical devices and applications that collect data that
are then provided to medical healthcare IT systems
(Alsubaei et al., 2019). The IoMT is a growing
domain and as it grows, cybersecurity risks have risen
(Brien et al., 2018; Papageorgiou et al., 2018).
Reports (Cisco, 2017; Ponemon Institute, 2018),
determined that in terms of security maturity and
privacy, the medical healthcare domain is behind
other domains and vulnerable to industry-related
cybersecurity. Small to medium enterprises (SMEs)
particulary struggle. Difficulties in budget
constraints, deficiency in knowledge and lack of
trained personnel (Dhillon, 2011; Cisco, 2017;
Ponemon Institute, 2018), complexity and
compatibility issues in terms of the variety of IoMT
technologies in use (Alsubaei et al., 2019) and
a
https://orcid.org/0000-0001-7722-5628
b
https://orcid.org/0000-0002-9285-5019
c
https://orcid.org/0000-0002-0839-8362
understanding regulatory requirements (Parker et al.,
2017), are some of the issues that contribute to
inadequate cybersecurity and privacy strategies
within this domain (Treacy & McCaffery, 2016). In
addition, the fact that data in flow in the IoMT can be
through various apps, systems, devices, technologies,
public and open networks, which are inherently
insecure such as wireless sensor networks and the
cloud, has led to many security issues (Ponemon
Institute, 2018). Moreover, security and privacy
issues have arisen due to the rush into the lucrative
healthcare domain and the speed the healthcare
domain is embracing IoT without a profound
understanding of the security and privacy risks
(Hatzivasilis et al., 2019; Sun et al., 2018).
Recommendations are that security and privacy are
designed at the beginning of a development project,
into the devices, the communication protocols and the
services (McManus, 2018). To address the regulatory
requirements and the above-cited difficulties, this
paper proposes a framework aimed at SMEs and
Treacy, C., Loane, J. and McCaffery, F.
Developer Driven Framework for Security and Privacy in the IoMT.
DOI: 10.5220/0009828304430451
In Proceedings of the 15th International Conference on Software Technologies (ICSOFT 2020), pages 443-451
ISBN: 978-989-758-443-5
Copyright
c
2020 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
443
developers to assist with regulatory compliance
requirements in addressing security and privacy of
data flow in the IoMT.
This paper is structured as follows. Section 2
positions our work with respect to existing
approaches. Section 3 offers our main contributions.
Section 4 details the framework. Finally, Section 5
provides the future work and conclusion.
2 RELATED WORK
There are many TM tools and methods used to
identify security threats (Hussain et al., 2014).
However, in the existing security TM frameworks,
including STRIDE (Deng et al., 2011), privacy
protection and threats are not emphasised (Gholami
et al., 2014). There are tools available for security TM
such as the Microsoft Tool for TM (Microsoft, 2020),
and the OWASP project, Threat Dragon tool
(OWASP, 2020). For privacy TM there is ongoing
research with privacy extensions for data flow
diagrams (DFDs) that extend on LINDDUN
(Antignac, Scandariato, & Schneider, 2016).
SPARTA is a tool in development for security and
privacy threat analysis (Sion et al., 2018). There is
also research to include in a structured way into the
TM process, the already determined security
decisions and constraints that are known before the
project begins (Sion et al., 2018). Hatzivasilis et al
(2019), presented a study of the main defence
mechanisms in core security and privacy controls, for
providing end-to-end security and privacy in the
IoMT. They state the study can act as a best-practices
guide for general IoT or specialised IoMT
applications.
3 CONTRIBUTIONS
It is now required that privacy and security are built
into the core of technical products, which is applied
by the EU General Data Protection Regulation
(GDPR) requirement of ‘data protection by design
and by default (GDPR, 2016). A requirement of the
GDPR is a Data Protection Impact Assessment
(DPIA). A DPIA is an effective way to assess and
demonstrate the project’s compliance with the data
protection principles and obligations (ICO 2020).
Documenting the framework process provides
evidence that a development project has implemented
the appropriate technical and organizational measures
to ensure and demonstrate compliance with the
regulations (ICO, 2020). For an inclusive DPIA the
current framework would require extension to include
data storage as this research looks explicitly at the
security and privacy of data in flow in the IoMT. A
contribution of the framework is in bringing together
the dispersed standards, best practice and guidelines.
In the existing security TM frameworks,
including STRIDE (Deng et al., 2011), privacy
protection and threats are not emphasised (Gholami
et al., 2014). This framework considers the
application of both security and privacy properties in
a TM process for data in flow in the IoMT. The
framework adopts and expands the Developer Driven
Threat Model Process (DDTM) developed by Danny
Dhillon (2011), which is based on the established TM
steps as referenced in (Myagmar, Lee, & Yurick,
2005; Swiderski & Synder, 2004). The DDTM uses
STRIDE for security threat identification. The
framework adds a TM process that addresses privacy
threat identification called LINDDUN. The
LINDDUN framework is a systematic approach to
assist with the elicitation and mitigation of privacy
threats in software systems (Sion et al., 2018). The
framework also applies security and privacy-aware
data flow diagram extensions to the TM.
Additionally, the framework maps the STRIDE and
LINDDUN threat categories to established security
and privacy properties and provides a categorized set
of data flow security and privacy controls (DFSPCs)
to mitigate the threats, which are also mapped to the
security and privacy properties. This simplifies the
threat mitigation process. The framework supports
collection of the DFDs, vulnerabilities, threats,
annotations and mitigation controls for the
development of a knowledge base library for future
projects.
4 FRAMEWORK
The framework currently focuses on the legal
requirements for PII protection in the EU, which is
regulated by the GDPR (General Data Protection
Regulation (GDPR), 2016). The proposed framework
has six steps: Step 1: Contextual knowledge. Step 2:
System decomposition. Step 3: Threat identification.
Step 4: Threat analysis. Step 5: Identify security and
privacy properties against threats. Step 6: Selection
of controls to mitigate threats.
4.1 Step 1 Contextual Knowledge
Step 1 provides the contextual knowledge to assist
SMEs and new or inexperienced developers to
ICSOFT 2020 - 15th International Conference on Software Technologies
444
understand the security and privacy context in order
to be able to use the framework. There are three parts
to step 1 summarized in the following sections.
4.1.1 Part 1 Security and Privacy Objectives
The framework is based on PII security and privacy
objectives that should reflect those of the organisation
and balance the regulatory data privacy and security
obligations (ISO/IEC, 2017). The framework security
objectives are based in the Information Security
Management System standard ISO/IEC 27001 (2017)
and the privacy objectives in the standard for Privacy
Information Management Systems, ISO/IEC 27701
(ISO/IEC, 2019). ISO/IEC 27701 is an extension to
ISO/IEC 27001 and ISO/IEC 27002 and the privacy
objectives align to the ISO/IEC 27001 security
objectives. As specified in ISO/IEC 27701, the names
and definitions for personal and health data
categorisation and definitions can vary between
different regulatory regimes. The framework requires
the development of a PII classification scheme for an
organisation, as detailed in ISO/IEC 27701, if not
already in place. The classification incorporates
definitions and explanation of: the nature of the PII
e.g. personal health information; PII principals
concerned e.g. PII relating to children. The regulatory
requirements differ in relation to age of consent and
how the information is processed and used; changing
or extending the purposes for the processing of PII,
which will require updating and/or revision of the
legal basis and additional consent for use from the PII
principal user.
4.1.2 Part 2 Security and Privacy Properties
Security and privacy properties are the common goals
that the framework protects for data in flow in the
IoMT. The traditional data security properties are
confidentiality, integrity and availability (CIA).
However, Whitman and Mattord (2011) assert that
the CIA model no longer adequately addresses the
constantly changing environment. The security
properties for the framework considered the
inadequacy of the CIA model and the requirements
for the IoMT. The framework examined network
security and adopted the eight security properties
founded in Part 3 of the ISO/IEC 27033 standard for
network security (ISO/IEC, 2010). Further research
added the security property Authorization added from
the STRIDE model. The security properties of the
framework are: Authentication, Integrity, Non-
repudiation, Confidentiality, Availability,
Authorization, Access Control, Communication or
Transport Security, and Privacy/Opacity. The
decision to incorporate all of these security principles
was taken as data in flow in the IoMT requires
comparable security principles required for network
security. The framework incorporates the privacy
properties Deng et al. (2011) used to develop
LINDDUN (Wuyts, Scandariato, & Joosen, 2014).
The privacy objectives are based on the LINDDUN
model, which are founded in the privacy objectives
defined by Pfitzmann and Hansen (2010). The
privacy objectives used are: Unlinkability,
Anonymity & Pseudonymity, Plausible deniability,
Undetectability & Unobservability, Confidentiality,
Content Awareness and Policy and Consent
Compliance.
4.1.3 Part 3 Administrate Two LINDDUN
Privacy Threat Categories
This part of the framework is the implementation of
two soft privacy properties, Content Awareness and
Policy and Consent Compliance, and management of
their LINDDUN threat categories Unawareness and
Non-compliance respectively. Both of these soft
privacy properties are significant for GDPR
regulatory requirements and the requirements of the
privacy standard ISO/IEC 27701. The non-
compliance threat category means not following the
data protection legislation, the advertised policies or
the existing user consents of the regulatory
jurisdiction (Wuyts & Joosen, 2015).The framework
includes details on the requirements of privacy policy
and consent to address policy and consent compliance
and content awareness. It outlines how development
can adhere to the specifics of the privacy policy. In
consent for collecting data, development obtains the
consent and then applies the correct collection in line
with policy reason and use of the collected data within
the specifics of demographics. Developers need to
understand that if data is collected or used outside the
privacy policy, there is a potential breach of privacy
regulatory requirements.
4.2 Step 2 System Decomposition
Step 2 of the framework is decomposition of the
system. The decomposition for the framework is
founded in DDTM
introduced by Dhillon (2011) as it
was developed to provide a process that incorporated
guidelines on creating DFDs for developers with or
without security expertise. DFDs were also used
because they are an established tool used by TM
(Osterman, 2007; Shostack, 2014) and both STRIDE
and LINDDUN use DFDs for decomposition and use
(Sion, Wuyts, et al., 2018). In addition, their use was
Developer Driven Framework for Security and Privacy in the IoMT
445
aided by the fact that DFDs support following the data
flow through a system and problems tend to follow
the data flow (Shostack, 2014). Likewise, ISO/IEC
27701, advises the use of DFDs as helpful tools to
inform a protection impact assessment and risk
assessment transfer, which assists with regulatory
requirements. Additionally, STRIDE and LINDDUN
provide a set of threat types in relation to the elements
of DFDs. Three key features for creating DFDs to
assist inexperienced developers and SMEs are
outlined in the framework, which are: DFDs
Elements and Symbols, Decomposition Levels and
Annotations.
4.2.1 DFDs Elements and Symbols
It is important that DFDs have a defined common set
of elements, element names and symbols and these
are used consistently throughout by the development
team and within projects (Ibrahim & Yen, 2011).
DFDs use standard symbols called elements, to
graphically represent the interaction between data
stores, processes, data flows, and external entities
(Shostack, 2014).
The framework uses these standard
set of symbols with the adapted change in the process
symbols presented by Ibrahim and Yen (2010, 2011).
The framework also applies the diversity of
boundaries, Machine Boundary and Process
Boundary, presented by Osterman (2007) and
Shostack (2008), and used by Dhillon (2011).
While the use of DFDs for TM security in systems
is well established, their use for privacy is an
emerging and developing area. The framework
incorporates the principles of privacy DFD models
from Antignac et al. (2016) to provide for privacy in
the DFD feature of the framework. Antignac et al.
(2016), applied the designation of the GDPR that
personal data processing involves: collection,
disclosure, usage, record, retrieval, and erasure. They
used these concepts to establish a link with the
privacy requirements of personal data as specified in
standards and regulatory texts. The research provided
a usable extension to the LINDDUN TM process
called Privacy-Aware Data Flow Diagrams (PA-
DFDs). The PA-DFDs privacy extensions reflect the
personal data processing concepts from the GDPR
and references the privacy principles of the ISO/IEC
29100:2011+A1:2018 (ISO/IEC, 2018). The
framework applies the PA-DFDs three different types
of external entities related to privacy, which are data
subjects, data controllers, and data processors and the
new element called Erasure, which pertains to the
purpose of ensuring that erased data comply with
the
GDPR ‘right to be forgottenprinciple. The
framework also uses the two new process elements
introduced, the usage and complex usage process
elements.
4.2.2 Decomposition Levels
The more you know about the system, the easier it is
to uncover threats (Meier et al., 2003). The
framework applies the DFDs decomposition levels
presented in research by (Ibrahim & Yen, 2010) and
Dhillon (2011). The DFDs hierarchical levels are:
Level 0 context diagram, system as a single entity;
Level 1 more detailed DFD through refinement of the
system; Level 2 more complex systems requiring
decomposition of some components with
annotations; Level 3 further decomposition of
complex systems Dhillon (2011). Level 1 and 2 of
DFD abstraction would include the annotations
outlined in the next section. The framework fosters
documentation for both security and privacy
beginning at decomposition Level 0 that include the
security and privacy properties and PII classification
of the project. The security and privacy
documentation will change over the level of
decomposition. The framework requires uniformity
in information and PII use through the levels of
decomposition. Given privacy in TM and DFD
decomposition is an emerging field, the framework
uses the eleven security privacy principles detailed in
the privacy framework standard ISO/IEC 29100
(2018) to provide a knowledge foundation for
developers and SMEs not experienced with privacy.
4.2.3 Annotations
Dhillon (2011) found in his experience in Microsoft
that DFD elements alone do not capture all necessary
details to perform a security threat model effectively.
Annotations to the DFDs consider the interactions of
the system that could provide additional information
to assist in the TM process (Dhillon, 2011;
Scandariato, Wuyts, & Joosen, 2015). Addition of
annotations were revealed to quicken the TM process,
by bringing focus on the typical areas attackers are
interested in and which are common sources of
vulnerabilities (Dhillon, 2011; Scandariato et al.,
2015). The vulnerabilities can be used to identify
interactions that could introduce weaknesses making
the identification process faster (Scandariato et al.,
2015). The framework added annotations for both
security and privacy. Security annotations include
critical security functions such as; authentication,
password management, and cryptographic
operations, network and local dataflow; HTTP, API
call and file I/O. Privacy annotations are linked to the
ICSOFT 2020 - 15th International Conference on Software Technologies
446
elements and interactions between the DFD elements,
as seen in Antignac et al. (2016). The framework adds
the annotations to the usage and complex usage
process elements, the annotation is purp. The
annotation of purpose is to track the purposes to
which the data subject has consented. The data flow
element carrying personal data is labelled pdata with
a link provided through a dotted line to the
corresponding data subject, which the personal data
refers to provide records of where PII is in the
system.(Antignac et al., 2016). It will also support
developers identifying where in development to
provide the added controls to keep the data secure and
private. Understanding where PII is in the system and
assessing the risks to the security and privacy at that
point, could influence how, in what form, the
necessity and why the data is transmitted. The
framework uses annotations from Level 0 of
abstraction, however, annotations can be added at any
level of abstraction. The annotations should evolve in
line with the previous abstractions of the system.
4.3 Step 3 Threat Identification
Step 3 of the framework is the identification of
threats. Categorising threats makes it easier to
understand what the threat allows an attacker to do
and supports in assigning priority and mitigation
(Hussain et al., 2014). Threat identification is central
to the TM process but, is also one of the most difficult
aspects of the process to complete, for developers
with little or no experience (Dhillon 2011). The
framework uses the threat categories from STRIDE
and LINDDUN to address security and privacy threat
identification. Threat identification is completed
through both element-based and per interaction-based
methods used by both LINDDUN and STRIDE and
is centred upon the building blocks of DFDs (Sion et
al., 2018).
Threat identification per-element is where each
element in the DFD is analysed through the threat
categories. Both STRIDE and LINDDUN have
mapped their threat categories to DFD elements. Not
all threat categories from both STRIDE and
LINDDUN apply to all DFD element types. The
STRIDE and LINDDUN threat categories to DFD
elements mapping are presented in Table 1, where the
LINDDUN threat categories are in red. Two of the
categories overlapped. These are non-repudiation and
disclosure of information. The LINDDUN threat
category non-compliance only refers to the regulatory
compliance and can only be addressed by the
organisation and is out of scope for developers.
However, it is vital that developers understand the
regulatory compliance requirements for PII and
comply with the PII classification, consent and the
privacy policy outlined in step 1.
Threat identification per-interaction considers all
interactions taking place in the system and details the
origin, destination and interaction in the system and
identifies threats against them (Shostack, 2014).
Particular focus is with entry and exit points, where
data enters or exits the system (Burns, 2005). The
framework requires identifying the entry and exit
points throughout the system through establishing the
interactions. This is particularly relevant to the
framework as the key categories for entry points to a
system offer a way-in for attackers. The IoMT
provides varied and numerous entry and exit points.
Examples include: communication (especially
wireless), software and physical (also known as
hardware). These categories generally overlap with
trust boundaries. When identifying the per-interaction
threats inside the IoMT, it is important to note that
trust boundaries are not fixed, they are subject to
change (Seeam et al., 2019), as data moves through
the system.
Table 1: STRIDE and LINDDUN threats categories
mapped per DFD element.
Threat
Category
STRIDE &
LINDDUN
Entity
(External)
Data
Flow
Data
Storage
Process
Spoofing X X
Tampering X X X
Repudiation
Non-
repudiation
X X X X
Information
Disclosure
Disclosure of
Information
X X X
Denial of
Service
X X X
Elevation of
Privilege
X
Linkability X X X X
Identifiability X X X X
Detectability X X X
Unawareness X
Non-
compliance
X X X
4.4 Step 4 Threat Analysis
Threat analysis is one of the most difficult aspects of
TM (Dhillon, 2011). Both the Stride and LINDDUN
categories are abstract enough, which means that
Developer Driven Framework for Security and Privacy in the IoMT
447
attacks could apply to one or more of the threat
categories. There is a degree of required
understanding and knowledge to map to tangible
attacks and without security knowledge, STRIDE
can’t be used effectively (Dhillon, 2011). To support
SMEs and developers lack of knowledge and
understanding required to complete threat analysis,
the framework maps the STRIDE and LINDDUN
threat categories to the OWASP Top 10 (OWASP,
2019). The framework recommends using the
mapping as an initial direction for threat elicitation
and analysis. Both STRIDE and LINDDUN have
extensive threat trees, employed in the framework as
guidance and additional assistance. The framework
also provides guidance for SMEs and inexperienced
developers to address threat explosion. Threat
explosion means that the threats can grow rapidly and
the effort required to take into account all threats
exceeds the TM process (Shevchenko et al., 2018;
Wuyts et al., 2018).
The framework threat prioritisation is guided by
the NIST SP 800-30 guide for conducting risk
assessments (NIST, 2012). The framework uses the
formula of Risk = Likelihood x Impact, including the
definitions of the standard. Appendices G and H of
the standard provide sets of exemplary tables for use
in determining adverse likelihood and impacts
quantitatively. Threat prioritization is guided by the
quantitative outcomes of the formula and managed by
the sensitivity of the PII and safety of the patient
associated with the threat and vulnerabilities. NIST
SP 800-30, provides standardised guidance for SMEs
and developers with little or no knowledge and
experience.
4.5 Step 5 Identify Security and
Privacy Properties against Threats
This step of the framework provides a connection
between the prioritised elicited threats and the
security and privacy properties the framework
protects. The threat categories of STRIDE and
LINDDUN map to the security and privacy property
it violates. The security and privacy properties are
affiliated to the standards and regulatory
requirements and step 5 maps the threats identified in
step 4 to the frameworks security and privacy
properties. The purpose of this mapping is to simplify
identification of appropriate security and privacy
controls to mitigate the identified threats. This is a
straightforward step in the framework necessary to
complete step 6. Similar to the threat elicitation and
analysis stages, there will be commonalities and
overlapping of property categories.
4.6 Step 6 Data Flow Security and
Privacy Controls
Step 6 of the framework is the identification of the
countermeasures needed to defend the security and
privacy properties breached by the threats identified
and prioritised. The security and privacy controls have
been classified with respect to the security and privacy
properties. A key objective of this research was the
establishment of a set of technical security controls to
maintain the security and privacy of data in flow. These
are called the Data Flow Security and Privacy Controls
(DFSPCs). The aim of the DFSPCs is to provide a set
of technical controls to assist developers comply with
security and privacy requirements of regulation and
close the gap in knowledge in this area. The DFSPCs
fill the vacuum of specific technical controls for the
security and privacy of data in flow to assist developers
to comply with the regulatory requirements.
The DFSPCs development originate in the
international standard IEC/TR 80001-2-8 (IEC/TR,
2016). IEC/TR 80001-2-8 identified over 300
security controls in a set of tables evaluated for their
relevance in establishing the 19 security capabilities
of IEC/TR 80001-2-2 (Jump & Finnegan, 2017).
Jump and Finnegan (2017), note that this should be
considered an approach for a basic foundation in
security. The controls are to manage risks to CIA and
accountability of data and systems and do not
consider privacy of data. IEC/TR 80001-2-8 does not
encompass the security and privacy properties
identified for the framework. The standards used to
develop the controls for IEC/TR 80001-2-8 were
examined for both the security and privacy controls
of the framework. IEC/TR 80001-2-8 mapped the
controls from six security standards. Two standards,
ISO/IEC 27002 (ISO/IEC, 2013) and IS0 27799 (ISO,
2016), are for operational and administrative security
controls and were not considered suitable for the
DFSPCs as they did not employ technical controls.
There were four standards used for technical controls:
ISO/IEC 15408-2 (ISO, 2016), ISO/IEC 15408-3
(ISO/IEC, 2008), NIST 800-53 Rev. 4 (NIST, 2014)
and IEC 62443-3-3 (IEC, 2013). ISO/IEC 15408-3
defines the assurance requirements of the evaluation
criteria, which was deemed out of scope for the
DFSPCs and was excluded. The remaining three
standards were examined to establish the DFSPCs.
5 CONCLUSIONS
In this paper, we presented a framework to assist
SMEs and inexperienced developers through a TM
ICSOFT 2020 - 15th International Conference on Software Technologies
448
process to protect the security and privacy of data in
flow in the IoMT. The framework considers both
security and privacy objectives and properties based
in standards and addresses regulatory requirements.
The framework is built on established TM processes
and threat and privacy threat identification categories.
In addition the framework provides a set of technical
security and privacy controls, the DFSPCs, from the
standards to mitigate the elicited threats.
Future work is in a comprehensive validation of
the framework by experts and implementation into
the development environment of a SME. Three
experts in the TM domain have been identified to
review the framework. The security controls have
been validated by two industry experts and one expert
from the medical device standards domain. Experts
from the privacy standard domain and industry have
been identified to validate the privacy controls. On
completion of the expert validation of the framework,
it will be implemented in the development teams of
two companies from the medical domain.
ACKNOWLEDGEMENTS
This work was supported with the financial support
of the Science Foundation Ireland grant 13/RC/2094.
REFERENCES
Alsubaei, F., Abuhussein, A., Shandilya, V., & Shiva, S.
(2019). IoMT-SAF: Internet of Medical Things
Security Assessment Framework. Internet of Things, 8,
100123. https://doi.org/10.1016/j.iot.2019.100123
Antignac, T., Scandariato, R., & Schneider, G. (2016). A
privacy-aware conceptual model for handling personal
data. In 7th International Symposium on Leveraging
Applications of Formal Methods, Verification and
Validation - ISoLA’16 (1) (pp. 942–957). Springer,
Cham. https://doi.org/10.1007/978-3-319-47166-2_65
Brien, G. O., Edwards, S., Littlefield, K., McNab, N.,
Wang, S., & Zheng, K. (2018). NIST SP 1800-8
Securing Wireless Infusion Pumps In Healthcare
Delivery Organizations. Retrieved from https://nccoe.
nist.gov/projects/use-cases/medical-devices
Burns, S. (2005). Threat Modeling: A Process To Ensure
Application Security. SANS Institute InfoSec Reading
Room. SANS.
Cisco. (2017). 2017 Annual Cybersecurity Report (Vol. I).
San Jose. https://doi.org/10.1002/ejoc.201200111
Deng, M., Wuyts, K., Scandariato, R., Preneel, B., &
Joosen, W. (2011). A privacy threat analysis
framework: supporting the elicitation and fulfillment of
privacy requirements. Requirements Engineering,
1(16), 3–32. Retrieved from http://web.b.ebscohost.
com.library.capella.edu/ehost/pdfviewer/pdfviewer?si
d=e7ebe3bc-59f7-43a0-ace9-60485dc3acd3%40sessi
on mgr111&vid=1&hid=118
Dhillon, D. (2011). Developer-Driven Threat Modeling:
Lessons Learned in the Trenches. IEEE Security and
Privacy, (9), 41–47. https://doi.org/10.1109/MSP.
2011.47.
General Data Protection Regulation (GDPR). (2016).
Regulation (EU) 2016/679 of the European Parliament
and of the council on the protection of natural persons
with regard to the processing of personal data and on
the free movement of such data, and repealing Directive
95/46/EC (General Data Protection Dir. EU. Retrieved
from https://eur-lex.europa.eu/legal-content/EN/TXT/
PDF/?uri=CELEX:02016R0679-20160504&from=EN
Gholami, A., Lind, A. S., Reichel, J., Litton, J. E., Edlund,
A., & Laure, E. (2014). Privacy threat modeling for
emerging biobankclouds. Procedia Computer Science,
37, 489–496. https://doi.org/10.1016/j.procs.2014.
08.073
Hatzivasilis, G., Soultatos, O., Ioannidis, S., Verikoukis, C.,
Demetriou, G., & Tsatsoulis, C. (2019). Review of
security and privacy for the internet of medical things
(IoMT): Resolving the protection concerns for the
novel circular economy bioinformatics. In Proceedings
- 15th Annual International Conference on Distributed
Computing in Sensor Systems, DCOSS 2019 (pp. 457–
464). IEEE. https://doi.org/10.1109/DCOSS.2019.
00091
Hussain, S., Kamal, A., Ahmad, S., Rasool, G., & Iqbal, S.
(2014). Threat Modelling Methodologies: a Survey.
Sci.Int.(Lahore), 26(4), 1607–1609.
Ibrahim, R., & Yen, S. Y. (2010). Formalization of the Data
Flow Diagram Rules for Consistency Check.
International Journal of Software Engineering &
Applications, 1(4), 95–111. https://doi.org/10.5121/
ijsea.2010.1406
Ibrahim, R., & Yen, S. Y. (2011). A Formal Model for Data
Flow Diagram Rules. ARPN Journal of Systems and
Software, 1(2), 60–69.
ICO. (2020). What is a DPIA? Retrieved January 13, 2020,
from https://ico.org.uk/for-organisations/guide-to-data-
protection/guide-to-the-general-data-protection-
regulation-gdpr/data-protection-impact-assessments-
dpias/what-is-a-dpia/
IEC/TR. 80001-2-8:2016 Application of risk management
for IT-networks incorporating medical devices
Application guidance Part 2-8 : Guidance on standards
for establishing the security capabilities identified in
IEC TR 80001-2-2 (2016).
IEC. (2013). 64223-3-3 Industrial Communication
Networks - Network and System Security Part 3-3:
System security requirements and security levels.
European Commission. Retrieved from http://www.
iec.ch/dyn/www/f?p=103:22:0::::FSP_ORG_ID:1250
ISO/IEC. (2008). 15408-3:2008 Information technology -
Security techniques - Evaluation Criteria for IT
Security - Part 3: Security assurance components.
ISO/IEC. (2010). 27033-3 -Information technology
Security techniques Network security Part 3:
Developer Driven Framework for Security and Privacy in the IoMT
449
Reference networking scenarios Threats, design
techniques and control issues. BSI Standards
Publication.
ISO/IEC. (2013). 27002:2013 - Information technology
Security techniques Code of practice for informati
on security controls. Switzerland.
ISO/IEC. (2017). 27001:2017 Information technology -
Security techniques - Information security management
systems - Requirements (ISO/IEC 27001:2013).
ISO/IEC. (2018). 29100:2011+A1:2018 Information
technology Security techniques Privacy
framework. British Standards Publication.
ISO/IEC. (2019). 27701:2019 Security techniques
Extension to ISO/IEC 27001 and ISO/IEC 27002 for
privacy information management — Requirements and
guidelines.
ISO. (2016). 27799-27002:2016 Health informatics
Information security management in health using,
27002.
Jump, M., & Finnegan, A. (2017). Using Standards to
Establish Foundational Security Requirements for
Medical Devices. Biomedical Instrumentation and
Technology, 51(s6), 33–38. https://doi.org/10.2345/
0899-8205-51.s6.33
Marr, B. (2018). Why The Internet Of Medical Things
(IoMT) Will Start To Transform Healthcare In
2018. Retrieved September 9, 2019, from
https://www.forbes.com/sites/bernardmarr/2018/01/25
/why-the-internet-of-medical-things-iomt-will-start-to-
transform-healthcare-in-2018/#7cbe068b4a3c
McManus, J. (2018). Security by Design: Teaching Secure
Software Design and Development Techniques. J.
Comput. Sci. Coll., 33(3), 75–82. Retrieved from
http://dl.acm.org/citation.cfm?id=3144687.3144710
Meier, J. D., Mackman, A., Vasireddy, S., Dunner, M.,
Escamilla, R., & Murukan, A. (2003). Improving Web
Application Security: Threats and Countermeasures.
Retrieved from https://www.microsoft.com/en-
us/download/confirmation.aspx?id=1330
Microsoft. (2020). Microsoft Threat Modeling Tool - Azure
| Microsoft Docs. Retrieved February 6, 2020, from
https://docs.microsoft.com/en-
us/azure/security/develop/threat-modeling-tool
Myagmar, S., Lee, A. J., & Yurick, W. (2005). Threat
Modeling as a Basis for Security Requirements Suvda.
In In Symposium on requirements engineering for
information security (SREIS) (pp. 1–8).
NIST. (2012). Special Publication 800-30 Guide for
Conducting Risk Assessments. Retrieved from
http://scholar.google.com/scholar?hl=en&btnG=Searc
h&q=intitle:NIST+Special+Publication+800-30#0
NIST. (2014). 800-53 Revision 4 Security and Privacy
Controls for Federal Information Systems and
Organizations Security and Privacy Controls for
Federal Information Systems and Organizations. Joint
Task Force Transformation Initiative. U.S. Department
of Commerce. https://doi.org/10.6028/NIST.SP.800-
53Ar4
Osterman, L. (2007). Threat Modeling, once again – Larry
Osterman’s WebLog. Retrieved November 18, 2019,
from https://blogs.msdn.microsoft.com/larryosterman/
2007/08/30/threat-modeling-once-again/
OWASP. (2019). OWASP Top Ten. https://doi.org/
10.1007/s11623-006-0164-8
OWASP. (2020). OWASP Threat Dragon. Retrieved
March 3, 2020, from https://owasp.org/www-project-
threat-dragon/#
Papageorgiou, A., Strigkos, M., Politou, E., Alepis, E.,
Solanas, A., & Patsakis, C. (2018). Security and
Privacy Analysis of Mobile Health Applications: The
Alarming State of Practice. IEEE Access, 3536(c), 1–
13. https://doi.org/10.1109/ACCESS.2018.2799522
Parker, L., Karliychuk, T., Gillies, D., Mintzes, B., Raven,
M., & Grundy, Q. (2017). A health app developer’s
guide to law and policy: A multi-sector policy analysis.
BMC Medical Informatics and Decision Making, 17(1),
1–13. https://doi.org/10.1186/s12911-017-0535-0
Pfitzmann, A., & Hansen, M. (2010). A terminology for
talking about privacy by data minimization:
Anonymity, Unlinkability, Undetectability,
Unobservability, Pseudonymity, and Identity
Management (Version 0.34 Aug. 10, 2010). Technical
University Dresden. tech. rep. TU Dresden and ULD
Kie. https://doi.org/10.1.1.154.635
Ponemon Institute. (2018). The State of Cybersecurity in
Healthcare Organizations in 2018.
Scandariato, R., Wuyts, K., & Joosen, W. (2015). A
descriptive study of Microsoft’s threat modeling
technique. Requirements Engineering, 20(2), 163–180.
https://doi.org/10.1007/s00766-013-0195-2
Seeam, A., Ogbeh, O. S., Guness, S., & Bellekens, X.
(2019). Threat Modeling and Security Issues for the
Internet of Things. In 2nd International Conference on
Next Generation Computing Applications 2019,
NextComp 2019 - Proceedings (pp. 1–8).
https://doi.org/10.1109/NEXTCOMP.2019.8883642
Shevchenko, N., Frye, B. R., & Woody, C. (2018). Threat
Modeling for Cyber-Physical System-of-Systems:
Methods Evaluation. Retrieved from https://resources.
sei.cmu.edu/library/asset-view.cfm?assetid=526365
Shostack, A. (2008). Experiences threat modeling at
Microsoft. CEUR Workshop Proceedings, 413, 1–11.
Shostack, A. (2014). Threat Modeling: Designing for
Security. John Wiley & Sons.
Sion, L., Van Landuyt, D., Yskout, K., & Joosen, W.
(2018). SPARTA: Security & Privacy Architecture
Through Risk-Driven Threat Assessment. Proceedings
- 2018 IEEE 15th International Conference on Software
Architecture Companion, ICSA-C 2018, 89–92.
https://doi.org/10.1109/ICSA-C.2018.00032
Sion, L., Wuyts, K., Yskout, K., Van Landuyt, D., &
Joosen, W. (2018). Interaction-Based Privacy Threat
Elicitation. Proceedings - 3rd IEEE European
Symposium on Security and Privacy Workshops, EURO
S and PW 2018, 79–86. https://doi.org/10.1109/
EuroSPW.2018.00017
Sion, L., Yskout, K., Van Landuyt, D., & Joosen, W.
(2018). Solution-aware data flow diagrams for security
threat modeling. In Proceedings of the ACM
Symposium on Applied Computing (pp. 1425–1432).
ICSOFT 2020 - 15th International Conference on Software Technologies
450
https://doi.org/10.1145/3167132.3167285
Sun, W., Cai, Z., Li, Y., Liu, F., Fang, S., & Wang, G.
(2018). Security and Privacy in the Medical Internet of
Things: A Review. Security and Communication
Networks, 2018. https://doi.org/10.1155/2018/5978636
Swiderski, F., & Synder, W. (2004). Threat Modeling.
Retrieved March 11, 2016, from https://msdn.
microsoft.com/en-us/library/ff648644.aspx
Treacy, C., & McCaffery, F. (2016). Data Security
Overview for Medical Mobile Apps Assuring the
Confidentiality , Integrity and Availability of Data in
Transmission. International Journal on Advances in
Security, 9(3 & 4), 146–157.
Whitman, M. E., & Mattord, H. J. (2011). Principles of
Information Security Fourth Edition. Boston: Cengage
Learning.
Wuyts, K., & Joosen, W. (2015). Tutorial privacy threat
modeling: a tutorial, Technical Report (CW Reports),
LINDDUN tutorial. Retrieved from https://linddun.org/
downloads/LINDDUN_tutorial.pdf%0Ahttps://distrine
t.cs.kuleuven.be/software/linddun/downloads/LINDD
UN_tutorial.pdf
Wuyts, K., Scandariato, R., & Joosen, W. (2014). Empirical
evaluation of a privacy-focused threat modeling
methodology. Journal of Systems and Software, 96,
122–138. https://doi.org/10.1016/j.jss.2014.05.075
Wuyts, K., Van Landuyt, D., Hovsepyan, A., & Joosen, W.
(2018). Effective and efficient privacy threat modeling
through domain refinements. Proceedings of the ACM
Symposium on Applied Computing, 1175–1178.
https://doi.org/10.1145/3167132.3167414.
Developer Driven Framework for Security and Privacy in the IoMT
451