IJCSN - International Journal of Computer Science and Network, Volume 6, Issue 5, October 2017

ISSN (Online) : 2277-5420

www.IJCSN.org
Impact Factor: 1.5

Evidential Modeling for Telemedicine Continual

Security
1
Sofienne Mansouri, 2 Bel G Raggad

1
U of Tunis El-Manar, ISTMT, Lab of Biophysics and Medical Technologies, Tunisia
2
Seidenberg School of CS & IS, Pace U, New York, USA

Abstract - Telemedicine has not advanced at the same pace as IT and its own medical technologies. The long-awaited progress has
been hindered aggressively by security risks that came with innovative information and communication technologies. One major
technological factor to blame for this tardiness in telemedicine is its information security that lead to patients and doctors attrition and
hence system infeasibility. Given the great deal of uncertainties and ambiguities in the telemedicine environment, Bayesian reasoning
does not offer a sound approach to tackle all the security problems menacing telemedicine. For this article proposes an evidential
reasoning model to manage risks due to security uncertainties and ambiguities characterizing most telemedicine environments.
Dempster and Shafer Theory is used to process security management evidence for the purpose to forecast the overall security risks
associated with the continual feasibility of a telemedicine system. This article also provides a numerical example to demonstrate the
working of the proposed evidential reasoning model.

Keywords - Telemedicine, Dempster and Shafer theory, evidential reasoning, Belief functions, security risk.

1. Introduction telemedicine environment including emails and

videoconferencing. A more advanced example would

T elemedicine has been around now for be the use of satellite technology to broadcast a
several decades, but it is not advancing at consultation between providers at facilities in two
the same pace as other technologies. In fact, countries, using videoconferencing equipment and
telemedicine has shown greater deficiencies despite the robotic technology.
advances in most information and telecommunication
technologies. Most of the risks that continued to rise Telecommunication has been enhanced to allow
with the rise of technology has hit telemedicine very physicians and specialists in remote health to deliver
hard and slowed its adoption and progress. These risks health care, diagnose patients, and provide therapy in a
are mainly associated with lacks in information privacy real-time manner. Telemedicine software systems have
and concerns associated with computer and network been also enhanced to allow for medical diagnostic
security. activities but these have been limited to augmenting the
performance of the communication effort and not
Even though you can see many major telemedicine necessarily the decision support part of the diagnosing
networks, like American Well, MDLIVE, and Teladoc, effort [12].
they are all hit with all types of government regulations
across state borders, in addition to major health Major applications involved in the telemedicine effort
standards like HIPAA, HL7, and regional data include radiology, pathology, cardiology and e-health
protection acts ([3], [4], [6]). education. Telemedicine adopted a variety of
approaches including Store-and-Forward, Home-
The significant progress made in the IT community Health, Real-Time, and other asynchronous methods
should certainly advance the telemedicine field. ([10], [11], [8]).
Advancement in secure data sharing among patients,
health providers, and insurance agencies with assured In this paper we refer to telemedicine simply as the use
mutual trust and information privacy established a of technology to provide health care at a distance, and
robust platform for real-time healthcare delivery. The we are mainly concerned with the security of its
entire nation started to accept the deployment of computing environment. . We see that this recent field
electronic health information systems and the federal is certainly playing a great role in delivering healthcare
government has sponsored it ([1], [10]). nationwide and the entire nation is committed to it.
There are then risks associated with the telemedicine’s
The major part of telemedicine that has seen great own information security that considerably depends on
development is the use of telecommunications to the security of all its subsystems on its network of
provide medical information and services. This includes health service deliveries: Remote patients, remote
simple data transfer among members of the doctors, remote hospitals, remote data resources, mobile

559

Copyright (c) 2017 International Journal of Computer Science and Network. All Rights Reserved.
IJCSN - International Journal of Computer Science and Network, Volume 6, Issue 5, October 2017
ISSN (Online) : 2277-5420
www.IJCSN.org
Impact Factor: 1.5

units, insurance agencies, and its distributed computing sM defining the security postures of all subsystems H1
environment. through HM constituting the telemedicine environment;
and the additional security parameter s0 summarizing
This article proposes an evidential reasoning model the continual security efforts applied by telemedicine
aiming at assuring a continual feasibility of a owners to keep its security risks lower than the
telemedicine system through effective information acceptable level defined in the telemedicine system
security. The security posture of the telemedicine security policy.
system is written in terms of the security postures of all
its subsystems and the effectiveness of owners’ efforts
to assure continual security. While the strength of a
security posture is expressed as the belief function of
adequate security, in contrast, the system security risk is
expressed as the plausibility of ineffective security.

2. Evidential reasoning model for the

telemedicine system
Let us consider the environment of a generic
telemedicine system as reported in most of the literature
[] as having multiple subsystems H1, …, HM that are all
connected together for the purpose of a feasible delivery
of telehealth services as required by all state and
national laws and regulations. Given the structure of a
telemedicine system, all performance deficiencies and
security threats on any of the telemedicine subsystems
will affect the overall security of the telemedicine Figure 1: Generic telemedicine security structure
environment. This is to say that any factors affecting the
security of one of the subsystems can propagate to the
main telemedicine system and compromise it. One That is, in order to maintain an acceptable continual
compromised, the telemedicine will lose its feasibility security level for the telemedicine security level, we
as an e-healthcare provider and my shut down. There is need to continuously manage the security of all its
certainly no doubt that the information security of the subsystems {Hi}i=1,M. and also exercise an adequate
telemedicine environment is a consequential variable to information security management level by system
continuously manage below any security risk levels owners in a continuous manner. Let us assume that we
accepted and defined in the system security policy. can capture all information security management
Figure 1 depicts the generic constitution of a information using a belief scheme consisting of M+1
telemedicine environment and how information security assertions, a0, a1, …, aM. The assertions a1, …, aM are
propagates throughout its computing environment. belief structures translating the adequate security
postures of respectively the subsystems H1, …, HM. The
No matter how you define uncertainty, ambiguity, or assertion a0 expresses the adequacy of the security
ignorance [Han et al), Telemed is full it. The great deal posture for the overall telemedicine system.
of uncertainty in all its forms and lack of structure will
make impossible for Bayesian theory to accommodate Of course, there won’t be any way to study those
any decision modeling for most information security assertions unless we can define the basic belief
decision. An evidential reasoning model, using assignments constituting their belief structures. And
Dempster and Shafer Theory, becomes hence very these belief structures can only be shaped up if we can
essential. elicit sufficient evidence to do so. Let us assume that
for every assertion ai, i=1,M, we collect a subset of
Such a security evidential reasoning framework for a evidence Ei constituting of |Ei| belief structures ei1, …,
telemedicine environment has to define all information ei|Ei|. The evidence of telemedicine owners’ efforts to
security parameters that affect the overall security of the maintain a continuously adequate security posture is
telemedicine system. Let us say, after consulting with capture by a single belief structure e0 that will be fused
telemedicine owners and their information security later with the rest of the evidence to produce
management team, that it is sufficient to examine M+1 information of the overall security posture of the
decision parameters: M security parameters s1 through

560

Copyright (c) 2017 International Journal of Computer Science and Network. All Rights Reserved.
IJCSN - International Journal of Computer Science and Network, Volume 6, Issue 5, October 2017
ISSN (Online) : 2277-5420
www.IJCSN.org
Impact Factor: 1.5

system. This evidential reasoning structure is depicted

in Figure 2. Table 1: Evidential structure for security management
Layer 0: Layer 2:
Begin Layer 0: Begin Layer 2:
Layer 0 = Sublayer 01, Sublayer Layer 2= {Sublayer21, …,
02} Sublayer2M}
Sublayer 01: Sublayer21:
{ {
Φ
a0: m[Telemed]: 2 →[ ] E1 = {e11, …, e1|E1|}
Φ 0 = {a01, a02} {
Φ
a01 = Adequate overall e11: m11: 2 →[ ]
Telemed s security posture Φ11 = {e111, e112}
a02 = Inadequate overall e111 = Adequate Security
Telemed s security posture Controls Set 1
} e112 = Inadequate Security
Controls Set 1
Sublayer 02: }
{ ---
E0 = {e0} {
Ω Φ E
e0: m0: 2 →[ ]} e1E1: m1E1: 2 →[ ]
Ω0 = {e01, e02} Φ1E1 = {e1E11, e1E12}
e01 = Adequate Telemed s e1E11 = Adequate Security
Figure 2: Evidential structure defining the security of the owners effort for continual Controls Set |E1|
overall security of the system e1E12 = Inadequate Security
telemedicine system
e02 = Inadequate Telemed s Controls Set |E1|
owners effort for continual }
overall security of the system }
} ---
} Sublayer2M:
End of Layer 0; {
EM = {eM1, …, eM|EM|}
Layer 1: {
ΦM
Begin: Layer 1 eM1: mM1: 2 →[ ]
Layer 1= {Sublayer11, …, ΦM1 = {eM11, eM12}
Sublayer1M} eM11 = Adequate Security
Sublayer11: Controls Set M
{ eM12 = Inadequate Security
Φ
a1: m1: 2 →[ ] Controls Set M
Φ1 = {a11, a12} }
a11 = Adequate H1 s security ---
posture {
ΦMEM
a12 = Inadequate H1 s security eMEM: mMEM: 2 →[ ]
posture ΦMEM = {eMEM1, eMEM2}
} eMEM1 = Adequate Security
--- Controls Set |EM|
Sublayer1M: eMEM2 = Inadequate Security
{ Controls Set |EM|
ΦM
aM: mM2 →[ ] }
ΦM = {aM1, aM2} }
aM1 = Adequate HM s security End of Layer 2.
posture
aM2 = Inadequate HM s security
posture
}
End of Layer 1

561

Copyright (c) 2017 International Journal of Computer Science and Network. All Rights Reserved.
IJCSN - International Journal of Computer Science and Network, Volume 6, Issue 5, October 2017
ISSN (Online) : 2277-5420
www.IJCSN.org
Impact Factor: 1.5

3. Evidential reasoning process

Let us recall what we have achieved so far in terms of
defining all the variable retained to manage the security
of our telemedicine system. Only one assertion is
studied and that is a0 expressing the adequacy of the
overall security of the telemedicine system. We then
selected M assertions a1, …, aM to capture the adequacy
of the security posture of the M telemedicine
subsystems H1, …, HM. Additionally, we defined M
evidence subsets E1, …, EM that will feed the needed
information to construct respectively the assertions a1,
…, aM. The following vectorial scheme shown in Table
1 gives a rather cumbersome representation of the
evidential structure that we thought may help those who
are interested to follow the mathematical computations
of the root assertion associated with the overall security
of the telemedicine environment:

3.1 Demonstration of the working of the ER model

Let us consider a simple telemedicine architecture, as
shown in Figure 3, made of 7 main subsystems:
Computing Environment (CE) subsystem, Patients
subsystem, Doctors subsystem, Hospitals subsystem,
Medical Data Resources subsystem, Mobile Units
subsystem, and Insurance Agencies subsystem. We
assume that the overall security of the telemedicine Figure 4: Evidential architecture for a Telemedicine system
environment depends on 7 security postures associated
with the 7 telemedicine subsystems identified in this 4. Conclusion
example, in addition to the adequacy of the continual
The article looked into the telemedicine tardiness in
security efforts exercised by system owners. The
following current advances in information and
computations are provided in Table 2.
telecommunication technologies and proposed an
evidential reasoning model to tackle its security
problems believed to be one of the major hinders of
telemedicine progress. Given the great deal of
uncertainties and ambiguities in the telemedicine
environment, Bayesian reasoning does not offer a sound
approach to tackle all those security problems menacing
telemedicine. We proposed an evidential reasoning
model to manage risks due to security uncertainties and
ambiguities characterizing most telemedicine
environments.

We employed Dempster and Shafer Theory to process

security management evidence for the purpose to
forecast the overall security risks associated with the
continual feasibility of a telemedicine system.
Figure 3: An example of a generic telemedicine architecture
This article also provided a numerical example to
. demonstrate the working of the proposed evidential
reasoning model.

562

Copyright (c) 2017 International Journal of Computer Science and Network. All Rights Reserved.
IJCSN - International Journal of Computer Science and Network, Volume 6, Issue 5, October 2017
ISSN (Online) : 2277-5420
www.IJCSN.org
Impact Factor: 1.5

References Interoperability, AMIA 2007 Symposium

Proceedings, pp 216-220, (2007).
[1] Blavin, F., et al., Final Report: Lessons from the [7] John Craig, J., and V. Patterson, Introduction to
Literature on Electronic Health Record the practice of telemedicine, Journal of
Implementation, U.S. Department of Health and Telemedicine and Telecare, 11(1), pp 1-9,
Human Services, http://www.healthit.gov/, (2005).
(2013). [8] Kiernan, TE. J. and Demaerschalk, B.M.,
[2] Demiris G, and D. Tao, An analysis of the Nursing Roles within a stroke Telemedicine
network, Journal of Central Nervous System
specialized literature in the field of
Disease, 2, 1–7, (2010).
telemedicine. Journal of telemedicine and
[9] Paul K.J. Han, KJP, Klein, WMP, and Arora,
telecare. 11(6): 316-327, (2010). NK, Varieties of uncertainty in health care: a
[3] De Moor, G. et al., Using electronic health conceptual taxonomy, Medical Decision
records for clinical research: The case of the Making, 31(6), pp 828–838. (2011).
EHR4CR project, Journal of Biomedical [10] Stoten, S., 2009. Health Policy Issue with the
Informatics, Vol 53, pp 162-173, (2015). Electronic Health Record, Online Journal of
[4] Doherty, N., & H. Fulford, Aligning the Nursing Informatics (OJNI), 13(2), 1-14,
information security policy with the strategic (2009).
information a systems plan. Computers & [11] Talmon J. AE., et al., STARE-HI: statement on
Security, 25, pp 55-63, (2006). reporting of evaluation studies in health
[5] Duan, L., Street W.N., and Xu, E., Healthcare informatics. International Journal of Medical
information systems: data mining methods in Informatics, 78(1), pp 1-9, (2009).
the creation of a clinical recommender system, [12] Xu, E., Wermus, M., and D.B. Bauman,
Enterprise Information Systems, 5(2), 169-181, Development of an integrated medical supply
(2011). information system, Enterprise Information
[6] El Fadlya, A.et al., Electronic Healthcare Systems, 5(3), pp 385-399, (2011).
Record and Clinical Research in Cardiovascular
Radiology: HL7 CDA and CDISC ODM

563

Copyright (c) 2017 International Journal of Computer Science and Network. All Rights Reserved.
IJCSN - International Journal of Computer Science and Network, Volume 6, Issue 5, October 2017
ISSN (Online) : 2277-5420
www.IJCSN.org
Impact Factor: 1.5

Table 2: Computations to fuse available evidence to produce the assertions

Budget evidence Adequate security Iterative Fusion

Either Either Either

Adeq Inadeq one Adeq Inadeq one Adeq Inadeq one

Computing Env sec posture 0.3 0.4 0.3 0.5 0.3 0.2 0.5 0.4 0.06

Patients sec posture 0.4 0.6 0 0.5 0.4 0.1 0.44 0.55 0

Doctors sec posture 0.6 0.3 0.1 0.6 0.3 0.1 0.75 0.23 0.01

Hospitals sec posture 0.3 0.7 0 0.7 0.1 0.2 0.56 0.43 0

Med Data sec posture 0.5 0.4 0.1 0.6 0.3 0.1 0.67 0.31 0.01

Mobile U sec posture 0.6 0.3 0.1 0.6 0.2 0.2 0.77 0.2 0.02

Ins agencies sec posture 0.7 0.3 0 0.3 0.5 0.2 0.62 0.37 0

Fuse a1-a2 0.5 0.4 0.06 0.44 0.55 0 0.46 0.48 0.04

Fuse a1-a3 0.46 0.48 0.04 0.75 0.23 0.01 0.74 0.24 0.01

Fuse a1-a4 0.74 0.24 0.01 0.56 0.43 0 0.79 0.2 0

Fuse a1-a5 0.79 0.209 0 0.67 0.31 0.01 0.87 0.1 0.01

Fuse a1-a6 0.87 0.1 0.01 0.77 0.2 0.02 0.95 0.03 0.01

Fuse a1-a7 0.95 0.03 0.01 0.62 0.375 0.01 0.98 0.02 0.01

a0 = Fusion of e0 and a1-a7 0.98 0.02 0 0.43 0.53 0.04 0.975 0.02 0

PS: Due to the approximations used in all computations, the m-values did not add to 1.

564

Copyright (c) 2017 International Journal of Computer Science and Network. All Rights Reserved.