Semantic Interoperability with Decision Support for ...
Transcript of Semantic Interoperability with Decision Support for ...
RESEARCH POSTER PRESENTATION DESIGN © 2012
www.PosterPresentations.com
Infectious diseases outbreaks demand a timely and proportional
response to mitigate effects on public health. Management of
these outbreaks is becoming a growing concern in public health,
as it requires extreme actions and coordination between
governing authorities at both state and national levels. Dealing
with large numbers of incoming reports and alerts requires an
automated system which performs real time analysis on a
centralized repository of collected clinical information.
Health authorities use these collected information to identify,
manage, and investigate infectious diseases outbreaks. Such
data, when visually and adequately represented, support the
education of healthcare providers within participating facilities
and improves the outcome of disease outbreaks management.
The challenges includes
INTRODUCTION
OBJECTIVES
Infectious Disease – Eco System
Architecture
CONCLUSION
This research work is driven by the immense needs and
posed challenges for integrating different healthcare
systems to monitor and manage public health indicators
such as infectious diseases. As such, this work can be
extended in many aspects such as putting forward a policy-
based framework to allow injecting regulators rules. This
direction should follow the approach specified by El-
Hassan et al. [13] which allows specifying rules for
accessing resources (e.g.. patients data) in both normal
and emergency situations
REFERENCES
1. Pandiyan, M., El Hassan, O.,Maamar Z.,Rajasekaran, P.
"Semantic Interoperability for Infectious Diseases Reporting
System", Sep 2011, Computer Science and Information Systems
(FedCSIS).
2. M. K. Smith, C.Welty, and D. McGuinness. OWLWeb Ontology
Language Guide. W3C Recommendation,
http://www.w3.org/TR/owl-guide/ ,May 14, 2011.
3. W3C. Owl web ontology language-reference. LSDIS Lab,
University of Georgia,2004. http://www.w3.org/TR/owl-ref/.
4. Iqbal, A.M.; Shepherd, M.; Abidi, S.S.R., “An Ontology-Based
Electronic Medical Record for Chronic Disease Management” in
Jan. 2011 44th Hawaii International Conference pp. 4–6.
5. Sampalli, T. Shepherd, M. Duffy, J.., “A Patient Profile
Ontology in the Heterogeneous Domain of Complex and Chronic
Health Conditions” in Jan. 2011 System Sciences (HICSS), 2011
44th Hawaii International Conference.pp 4-6.
6. Arch-int, N.; Arch-int, S “SEMANTIC INFORMATION
INTEGRATION FOR ELECTRONIC PATIENT RECORDS
USING ONTOLOGY AND WEB SERVICES MODEL” in April.
2011 Information Science and Applications (ICISA),
International Conference.pp 3-5
7. International Classification of Diseases.
http://www.cdc.gov/nchs/icd/icd10.htm, May 14, 2011.
8. International Classification of Diseases, Ninth Revision (ICD-9)
http://www.cdc.gov/nchs/icd/icd9.htm, May 24, 2011.
9. Intersystems Cache. http://www.intersystems.com/cache/, May
24, 2011.
10.Business Objects
http://www.sap.com/solutions/sapbusinessobjects/index.epx, May
24, 2011
11.Jena A Semantic Web Framework for Java.
http://jena.sourceforge.net/ , May 14, 2011.
12.Hong Jiang, Hua-qiong Wang, Hong-lei Zhang, Peng-fei Li,
Jing-song Li. “Modeling for the Semantic Integration of Clinical
Pathways with Related Medical Systems”, 2012 International
Symposium on Information Technology in Medicine and
Education.
13.El-Hassan, Osama and Fiadeiro, Jos'e Luiz and Heckel, Reiko
“Managing socio-technical interactions in healthcare systems” in
2008 Proceedings of the 2007 international conference on
Business process management.
ACKNOWLEDGEMENTS
Thanks go to Dubai Health Authority team and particularly the
Director of Health Data and Information Analysis department whose
guidance and support are vital for completing this research
1. To gather information from various sources (i.e., healthcare
facilities) in real-time because of the diverse and
heterogeneous nature of healthcare applications.
2. To come up with a nation-wide policy-based infectious disease
monitoring system which can be implemented across several
healthcare regulators, has the ability to process generated
infectious diseases reports at different diagnosis stages (pre &
post confirmation) and statistically compute accurate infectious
diseases rates.
3. From a regulatory body point of view, since the infectious
disease is a growing concern in public health, it is necessary to
collaborate with other health authorities to exchange and
manage the related information and alerts.
4. For this automatic assessment of degree of accuracy of diagnosis
process and management are prime importance.
Murugavell Pandiyan(Student), Osama ElHassan (Head of eHealth – Dubai Health Authority), Pallikonda Rajasekaran (Professor, KLU, Tamil Nadu, India)
Semantic Interoperability with Decision Support for Infectious Disease
CHALLENGES
1. Each healthcare facility has its own software to manage patients
data e.g., Physician Practice Management System (PPMS) and
Electronic Medical Records (EMR). Moreover, even standardized
EMR systems might be interfaced differently and sub-systems
such as Lab systems are isolated in terms of their used standards
(i.e., proprietary standards) and thus raise extra integration
challenges.
2. Regulatory authority has to make sure the data retrieved from
clinics /laboratories have undergone through proper
policies/procedures.
3. Treating different policies/workflows of reporting a certain
infectious disease.
4. The Challenge of computing the accurate infectious diseases
rates is exacerbated by the inclusion of incomplete
information(e.g. incomplete diagnosis) or misdiagnose. Diagnosis
workflows of certain infectious diseases e.g. tuberculosis
5. It is necessary to collaborate with other health authorities to
exchange and manage the related information and alerts.
Disease centric model
Ontology model on SNOMED
Ontology model on ICD 9/ ICD 10
Relationship / Other codes
Ontology model on Health
Authority rules / regulations and
Diseases.
Ontology model on healthcare provider and
their licensure
Ontology model on Facility
Licensing Rules and the Quality
Certification Results
Ontology model on licensure
and their treatment procedure
relationship
Ontology modelDecision support - flow
neuron
X1
X2
Xn
W1
W2
Wk
bias
X2
Sigmoidal Function
Net Weightage.
Net𝑋𝑘 = 𝑘=1𝑘=𝑛𝑋𝑘𝑊𝑘 + 𝜃𝑘
Sigmoidal FunctionF(Sig) = 1/(1+𝑒−𝑥)
Transforming the net weightage.F(x)= Net𝑋𝑘 * F(Sig)
The current weightage change
∆𝑊𝑖=-ηΔ𝐸
𝑊𝑖
Net Xk
TO DETERMINE DIAGNOSIS USING ARTIFICIAL NEURAL NETWORK
LOINC Code Result Weightage Satisfied
X1 N1 W1 True
X2 N2 W2 True
X3 N3 W3 True
X4 N4 Wn False
From the statistical database calculate the incident rate
(𝜏).
Quantity of vaccination ∝ proportional to incident rate.
Incident rate likelihood over the period can be calculated
from Poisson Distribution.
𝑒−𝜏 ∗ 𝜏𝐾
F(K; 𝜏) = ---------------------
k!
CALCULATE THE VACCINATION QUANTITY REQUIREMENT
<ha:regulation>
<rule:licensurerule licensetype=”X”>
<rule:allowedProcedures>
<rule:treatmentcode codingsystem=”SNOMED-CT”>
xxx-yyyy
</rule:treatmentcode>
<rule:treatmentcode codingsystem=”SNOMED-CT”>
xxx-zzzzz
</rule:treatmentcode>
</rule:licensurerule>
<rule: diagnosispolicies seq=“1”>
<disease:tuberculosis code=”ddd-eeee” codingsystem=”SNOMED-CT”>
<disease:coretests>
<labtest:procedure code=”17296-5” codingsystem=”LOINC”>
<labtest:description>
Mycobacterium tuberculosis complex rRNA [Presence] in Unspecified specimen by
DNA probe
</labtest:description>
<labtest:weightage>W1</labtest:weightage>
</labtest: procedure >
</disease:coretests>
</disease:tuberculosis>
</rule:diagnosispolicies>
<rule:authorizationpolcies>
<rule:signingauthorities>
<regulator:provider code=“physician1”>
<regulator:providername>
</regulator:providername>
</regulator:provider>
</rule:signingauthorities>
</rule:authorizationpolcies>
……………………….
<ha:IsQuaCertificationRule>
<disease:tuberculosis code=”ddd-eeee” codingsystem=”SNOMED-CT”>
<ha:rule code=“x1” refers=“isquacode1” >
<rule:action call=“certificationCheckList()”>
<action:param>isquacode</actionparam>
<action:paramValue>$isquacode</action:paramValue>
</rule:action>
</ha:rule>
</disease:tuberculosis>
</ha:IsQuaCertificationRule>
</ha:regulation>