USING SEMANTIC WEB IN ICD-11 :THREE YEARS DOWN THE ROAD

1. USING SEMANTIC WEB IN ICD-11 :THREE YEARS DOWN THE ROAD -KHUSHBOO BAGHADIYA

2. OVERVIEW Introduction System Description iCAT in use Evolution of the system Evolution of modeling Evolution of features Evolution of collaboration process Benefits of using Semantic Web Technologies Conclusion

3. INTRODUCTION The World Health Organization is using Semantic Web technologies in the development of the 11th revision of the International Classification of Diseases (ICD-11) Health officials use ICD in all United Nations member countries to compile basic health statistics, to monitor health-related spending, and to inform policy makers

4. INTRODUCTION United Nations member countries are using ICD to analyze the general health of population groups, to monitor the prevalence of diseases and other health problemsandto compile mortality and morbidity statistics They have developed iCAT, a Web-based platform built as an extension of theprotégé system that WHO uses as the technological infrastructure for the development of ICD-11

5. SYSTEM DESCRIPTION WHO is building ICD-11 as an ontology represented in OWL and iCAT is a Web-based platform for authoring it Dr.Musen and group in collaboration with WHO and the Health Informatics Modeling Topic Advisory Group, created the core part of the ICD ontology

6. SYSTEM DESCRIPTION Two main goals with implementing iCAT: Enable domain experts who are distributed around the world to edit the ICD ontology collaboratively Provide an interface that is easy enough to use and that will allow non-ontology experts to edit the ICD ontology

7. SYSTEM DESCRIPTION The iCAT platform is a customization of WebProtégéa Web client for the Protégé ontology editor Similar to WebProtégé, iCAT is implemented in Java, it uses the Google Web Toolkit (GWT) and is open source iCAT takes advantage of many WebProtégé features, such as the concurrent browsing and editing of an ontology by distributed users, the read-write access policy mechanism, form-based editing and full-fletched collaboration

8. SYSTEM DESCRIPTION Gives us a way to hide the underlying OWL representation from the users, and to present them with a simple and familiar editing interface that they can understand Ability to create links between properties of diseases and external terminologies stored in the BioPortal ontology repository

9. SYSTEM DESCRIPTION

10. WORKING OF iCAT Distributed users edit the ontology simultaneously-propagates the changes to all client Web browsers iCAT tracks all the edits a user makes and stores these changes as instances in a Change and Annotation Ontology (ChAO), creating a structured log that we can easily access programmatically Users can Carry out threaded discussions Attach a note to a disease Watch entities or branches in the ontology and receive email notifications about the changes and notes in their watched areas

11. iCAT IN USE Running since November 2009 Different types of users are using the platform in their daily work: Classification experts Domain experts Project managers

12. iCAT IN USE BEFORE AFTER Users : 48 270 Changes : 15,000 260,000 Notes : 5,000 61,000

13. iCAT IN USE

14. EVOLUTION OF THE SYSTEM We focus on three aspects of iCAT evolution: Evolution in terms of modeling Evolution of iCAT features Evolution of the collaboration processes

15. EVOLUTION OF THE SYSTEM

16. EVOLUTION OF MODELLING The initial ICD-11 ontology simply modeled the ClaML schema (Classification Markup Language) Content model(Disease and injury) Meta classes(definition of disease, description of disease) Views(extract different views for a specific use case) Backwards compatibility (ICD-10  ICD11) Post-coordination(Severe, Moderate, Acute)

17. EVOLUTION OF FEATURES Avoiding duplicates External access Export to spreadsheet

18. EVOLUTION OF COLLABORATION PROCESS iCAT started as an open platform Less than a year into the project, the groups participating in the process requested that they should close the access to the platform One way to motivate users to contribute more was to displays a Top 10 of the “most active” ontology branches in ICD

19. BENEFITS OF USING SEMANTIC WEB TECHNOLOGIES Traditional benefits - External benefits Internal benefits

20. EXTERNAL BENEFITS Can rely on well-defined semantics, which ensures that all tools and users will interpret it in a uniform way The previous approach - ClaML -clearly defined Data structures - Different interpretations Using OWL opens the door of reusing ICD in other biomedical ontologies, or for other purposes in applications that need machine-processable information (ICPS, ICTM) ICD in OWL naturally has unique identifiers for all its entities, a critical feature in allowing its reuse in other applications

21. INTERNAL BENEFITS Building and maintaining the tool itself as internal benefit Generality of WebProtege toolcombined with the semantic modeling - deal with the frequently changing requirements(10-56,3-12) The iCAT interface is a domain-expert-friendly rendering of the underlying OWL ontology The core OWL ontology kept changing frequently even after iCAT has gone into production

22. CONCLUSION Significant benefits in using semantic technologies - adjust to changing requirements and to support a changing processes in an agile way The result is one of the most high-profile uses of Semantic Web in the real world The only real-world project supporting the collaborative authoring of ontologies at this scale, and which, at the same time, has a high visibility and impact for the health care around the world

23. THANKYOU