Upload
nooshin-allahyari
View
76
Download
0
Tags:
Embed Size (px)
DESCRIPTION
This project overviews the requirement of prescription medication process, analyzes its processes and dependencies, and designs a intelligent decision support system to overcome some of its issues.
Citation preview
Knowledge Management Needs in Prescription-Medication ProcessAllahyari Nooshin, Das Aby.March 30, 2011 1
Agenda
•The Current Process•Knowledge Identification•Portfolio of Systems•Selected Portfolio•Modeling and Analysis of KM Information Technologies •Role of Ontologies
2
Motivation
3
Motivation (cont.)
4
Problematic Prescription
The Current Process–Flow Diagram
5
The Current Process – Logical DFD at Level Zero
6
The Current Process – Logical DFD at Level One
7
The Current Process – Logical DFD at Level One (cont.)
8
The Current Process – Strategic Dependency Model
9
i* Model10
Possible scenarios to decrease preparation errorsPossible Scenario Advantage Disadvantage
Automated System Decreases human errors.
Cost
E-Bulletin for guidelines Aids pharmacist’s decision
Outdated, Time consuming.
Forums Pharmacist’s can contribute experiences with other pharmacists.
Time consuming, Lack of incentives.
Human-Factor Engineering
Improves efficiency Not periodically updated.
11
But preparation depends on the pharmacist …
12
Knowledge Identification
13
Roles, Responsibilities, and Dependencies
14
Current Knowledge Stores• Pharmacy Local Database• Stores customer customer information, prescription information,
and inventory of drugs. • Physical Prescription • Drug information, dosage.
• Indexing System• Locates drug information, patient history in pharmacy
• Physician’s Database• Physician’s access to patient’s records
15
Knowledge Management Portfolio• Decision Support System (DSS)• Provide expert knowledge to aid pharmacist’s decision.
• Electronic Publishing System • Access to patient information using electronic means.
• Web Portal• Provides access to tools like wikis, form, email, search, and
retrieval tools.• Intranet• Facilitates access to patient past prescription and records across
different pharmacies.
16
Selected Portfolio
• Decision Support System (DSS)
• Electronic-Prescription (EP or E-Prescription)(combines with Electronic Publishing System)
17
Modeling and Analysis of Selected KM Technologies
18
DSS
19
E-Prescription
20
Change in dependencies
21
SD Model before EP
SD Model after EP
22
Interaction of actors and stakeholders with DSS
23
Goal Evaluation for DSS technology
24
Interaction of actors and stakeholders with E-Prescription
25
Goal Evaluation for EP technology
26
Impact of Employed Technology on each other
27
Temporal and Spatial Context
28
Role of ontologies
29
Why Ontologies?
• Unified Healthcare System• Ontologies can explicit conceptualize the semantics of the data • Ontologies can make deductions and reasoning.
30
What do ontologies do?• Ontology application are classified into:• Semantic integration• Search• Decision Support System(DSS).
31
32
Which languages?• RDF• RDF schema• OWL• Rules• KIF• Common Logic• FOL
Powerful logical languages
Conceptual Graph
33
Relationship between diseases, drug, and instructionsFor all (x)
(If disease(x) (exists (y) exists (z) and ( and medicine(y) has_medicine(x,y))
(and instruction(z) has_instruction(x,y,z)))).
34
35
Physician, Pharmacist, and EP RelationshipFor all(x) (iff prescription(x) (exists(y) exists(z) and (and Physician(y) prescribed_by(x,y)) (and pharmacist(z) used-by(x,z)))).
36
Drug prescription error
For all(x) (if Disease(x)
( not exists (y) (and medicine(y) has_medicine(x,y)) (exists(z) exists (w) ( and message(z) physician(w) has_ message(x,w,z)).
37
Benefits of Using our Ontology• It does not have other languages or ontologies constraints • all other semantic web languages are constriction of FOL.
• It is powerful in making deduction and reasoning• It could make inference between different ontologies.
Horrocks et al. (2005), Semantic Web Architecture: Stack or Two Towers?
38
Ontology Structure• Our ontology is using different existing ontologies and using
ontology mapping techniques to connect them to each other. • WSMO (Web Services Modeling Ontology)• DOPE (Drug ontology)• Disease ontology
39
WSMO (Semantic Web)
Romana et al. (2005) , Web Service Modeling Ontology
Questions?
40
Thank You!
41