Metadata Open Forum – "15 Years of Industry Discussion"


May 30th, 2012

 

Introduction

ISO/IEC 11179 Metadata Registries (MDR) and ISO/IEC 19763 Metamodel Framework for Interoperability (MFI)

Ms. Denise Warzel, WG2 Convenor, U.S. National Cancer Institute

The ISO JTC1 SC32 WG2 is responsible for Metadata standards within the Data Management and Interchange standards sub-committee.  Two of the primary work products are ISO/IEC 11179 Metadata Registries and ISO/IEC 19763 Metamodel Framework for Interoperability families of standards.  These standards are intended to enable adopters to implement registries that contain metadata descriptions of various kinds of information resources. The syntax and the semantics of these resources are documented using an information resource specific metamodel.  The metamodel is annotated with domain specific language, preferably an ontology, to provide unambiguous meaning. These registries can be used to discover and use information resources consistently and accurately, thus enabling interoperability.  This talk will briefly introduce these standards and provide a update on current WG2 development.

The standards of WG2 include:

ISO/IEC 11179 Metadata Registries (MDR)
11179-1 Framework Ed 3 Committee Draft
11179-2 Classifications Ed 2 International Standard
11179-3 Registry Metamodel and Basic Attributes Ed3 Final Draft International Standard
11179-4 Formulation of Definitions Ed2 International Standard
11179-5 Naming and Identification Principles Ed 3 Committee Draft
11179-6 Registration Ed 3 Working Draft

ISO/IEC 19763 Metamodel Framework for Interoperability (MFI)
19763-1 Framework Ed 2 Working Draft
19763-3 Ontology Registration International Standard
19763-5 Process Registration Committee Draft
19763-6 Registry Summary Committee Draft
19763-7 Service Registration Committee Draft
19763-8 Role/Goal Registration Committee Draft
19763-9 On Demand Model Selection (ODMS) Proposed Draft Technical Report
19763-10 Core Model and Basic Mapping Committee Draft
19763-12 Information Models Committee Draft

This talk will provide a brief overview of the role of these standards in semantics management and discovery.

<Return to Top> 

Talk 1

Healthcare: Mapping EHR Data to a Research Case Report Form: How a Metadata Repository, CDISC’s SHARE, can improve the IHE profile Clinical Research Data (CRD)

Mr. Landen Bain, CDISC Liaison to Healthcare, Ms. Julie Evans, CDISC Director of Technical Services, Mr. Pierre Yves Lastic, Senior Director of Data Privacy and Healthcare Interoperability Standards, Sanofi-aventis

The standards organizations IHE (www.ihe.net) and CDISC (www.cdisc.org) have jointly developed an approach to prefilling a research electronic case report form (eCRF) by leveraging the data and workflow of electronic health records (EHR). This link between healthcare and research systems takes a semantically naive approach which nevertheless has proven valuable as a first step towards full semantic interoperability. This paper explores ways in which a metadata repository, CDISC’s Shared Health and Clinical Research Electronic Library (SHARE), can improve the process of prefilling a case report form using standard documents exported from an EHR.

An IHE content integration profile, Clinical Research Data (CRD), provides the first steps towards mapping an HL7 (www.hl7.org) Continuity of Care Document (CCD) to CDISC’s Clinical Data Acquisition Standards Harmonization (CDASH) dataset as expressed in CDISC’s Operational Data Model (ODM). The CRD profile provides a sample XSLT which succeeds in mapping three categories of well-structured data: demographics, laboratory results, and medications. But the CRD does not tap the more semantically rich aspects of the CCD such as allergies, procedures, and vital signs.

The CDISC SHARE project is a metadata repository of research variables, value lists, and CDISC standards variables that, as its first goal, will structure the metadata in CDISC’s CDASH and Study Data Tabulation Model (SDTM) standards and make it available electronically. An extension of the core SHARE project has begun to map the CDASH and SDTM concepts to the concepts in Clinical Element Model (CEM). The CEM is a set of formal representations of scientific concepts that can be used to share the meaning of data in a computable manner. This metadata linkage of research to healthcare concepts opens new possibilities to improve the CRD’s ability to prefill eCRFs.

<Return to Top> 

Talk 2

Telecommunications: Metadata requirements in telecommunications cloud computing

Ms. Ewelina Szczekocka, Telekomunikacja Polska

Today a huge number of diverse services is available in Internet and via mobile communication, using different devices, like smartphones or tablets. New technologies, like Service-Oriented Architecture, Cloud Computing, rich Internet applications, particularly Telco 2.0, that includes rich communication features, emerge in the last decade, trying to address challenges, concerned with offering relevant and dynamic services to the users.

All these systems and technologies, give Telecoms the new opportunities, to be not only suppliers of a communication medium, but providers of a wide range of new kind of services, combining communication and IT features. Telecoms can collaborate with other business partners, in order to create new values. Telekomunikacja Polska S.A. and other Orange-brand companies, provide strategies focused on creating huge repositories (or clouds of services) offering telecom services, services delivered by business partners and external developers.

We are aware, that effective search, discovery and composition of services by end users or developers, requires additional mechanisms, not only traditional keyword-based search and usual service descriptions. We need to offer services not only in the scope of our repository, but across repositories of our business partners.

Appropriate metadata information, metamodels and semantic mechanisms are required to enable effective and easy navigation through service catalogues of different providers, in order to find services, achieve automatisation of service discovery and composition, as well as interoperability and interchange between providers. Standardised solutions like MFI and MDR, are for us a key technology in order to achieve interchange and interoperability, that will help in boosting the consumption of services.

Our goal is to share experiences, express our needs and have an opportunity for a fruitful discussion. Telekomunikacja Polska S.A. provides several activities, aiming at delivery of automated service search and discovery, like modelling semantic search in service repositories, in scope of internal, partnership and European activities.

<Return to Top> 

Talk 3

Telecommunications: Representing semantics for user oriented service compositions

Mr. Mariano Belaunde, Orange Labs

In order to facilitate the construction of service compositions by non-IT specialists it is important to raise the level of abstraction in service descriptions, which means putting focus not only in structural definitions (like WSDL) but also classifying services and capturing precise semantics of exposed service operations and parameters. In this presentation we describe the approach we had in orange Labs to represent semantics when dealing with service compositions created by users using a mix of natural language and graphical oriented mashup.

<Return to Top> 

Talk 4

Healthcare: ISO/IEC 11179 Metadata Registry

Dr. Steve Harris, Oxford University Computing Laboratory

ISO11179's inclusion in the developing DDI3 standard for the exchange of experimental data in statistical studies has made it an essential part of future experiment registers.  We will look at the Oxford Metadata Group's experience in developing study registers in population studies and clinical trials for hints about how we might take development of the standard forward after the release of edition 3 of the metamodel.

<Return to Top> 

Talk 5

GeoSpatial: The Ingest, Storage, and Retrieval of Geopolitical Code Sets in a 11179 Metadata Registry

Mr. Gramm Richardson, U.S. Department of Defense

The differing conceptualizations of geopolitical code sets used by international organizations present challenges when attempting to align concepts from one code set to another. The various international organizations do not even share a common definition for frequently used
terms such as country, territory, region, government, name, etc.  This makes attempts to perform automated alignment based solely on mapping these terms fraught with inaccuracies and inconsistencies.
The ISO/IEC 11179 metadata standard provides a way to express these concepts in a consistent, standard fashion, enabling different code sets to use the same terms despite differences in meaning. It also provides a framework for depicting the precise relationship between the concepts in different code sets, regardless of the terminology employed by a
particular code set. We have designed processes to successfully and quickly ingest geopolitical code sets provided as RDF or CSV, and transform them into RDF in a format compliant with the ISO/IEC 11179 registry metamodel. The concepts in the code sets are mapped to one another using automated techniques based on name similarity. However,
manual mapping is required in a few notable cases when the name of a particular geopolitical entity does not contain adequate information to uniquely identify this entity.
We use an OWL ontology (ALCHIQ(D) DL expressivity) as the framework for constructing and validating the RDF triples representing the metadata, which are stored in an RDF graph store.  The graph store also contains some non-11179 metadata to take advantage of efficient querying, and to store metadata that has not been converted to the 11179 standard representation.  The latest SPARQL 1.1 Query and Update draft provides useful capabilities to ease the manipulation and querying of the RDF data.

<Return to Top> 

Talk 6

Healthcare: A National Metadata Repository for Empirical Research in Germany

Professor Jürgen Stausberg, IBE Ludwig-Maximilians-Universität München, Germany

Drepper J, TMF - Technology, Methods, and Infrastructure for Networked Medical Research, Berlin, Germany

Löbe M, Institute for Medical Informatics (IMISE), Universität Leipzig, Germany

Ngouongo S, IBE Ludwig-Maximilians-Universität München, Germany

Verplancke P, XClinical, München, Germany

The identification of the appropriate parameters to describe outcome variables, interventions and confounders, the definition of value lists, the preparation of standard operating protocols, and the selection of plausibility checks is a complex and time-consuming task in medical empirical research. Using standards and reusing existing definitions is the right answer to reduce this overload.

A national metadata repository (MDR) is seen as a platform to distribute standards and foster the reuse of empirical research work already existing in Germany. Funded by the German Ministry of Research and Education and supported by the TMF - Technology, Methods, and Infrastructure for Networked Medical Research, an umbrella organization of research networks, a project was set up to implement a metadata repository as a national service in Germany. The project team decided to consider the metamodel of ISO/IEC 11179 “Information Technology - Metadata Registries (MDR)” draft version 3 as the heart of the German MDR.

The metamodel of ISO/IEC 11179 “Information Technology - Metadata Registries” draft version 3 is the most challenging approach for the implementation of an MDR. The evaluation confirmed the capability of ISO’s metamodel to represent the relevant standards. However, the integration of such an MDR into the processes of study centers and research networks is not as common as initially expected. Therefore, the projects’ efforts will further concentrate on the practical integration of the MDR into the daily work processes.

A first prototype of the German metadata repository is planned for mid 2012. At the moment, users are involved via workshops to adapt the metadata repository to the users’ needs as closely as possible. Some efforts are made to encapsulate the complex metamodel of ISO/IEC 11179 in the user interface of data managers. Metadata of clinical trials and registers as well as standards for item definitions and terminological entities are imported or manually entered into the prototype. According to the scrum method used for software development, the metadata repository is incrementally realized.

<Return to Top> 

Talk 7

Healthcare: The TMF and its Metadata and Linked Data Forum

Mr. Matthias Löbe M, Institute for Medical Informatics (IMISE), Universität Leipzig, Germany

Drepper J, TMF - Technology, Methods, and Infrastructure for Networked Medical Research, Berlin, Germany

Ngouongo S, IBE, Ludwig-Maximilians-Universität München, Germany

Stausberg J, IBE, Ludwig-Maximilians-Universität München, Germany

Verplancke P, XClinical, München, Germany

The TMF - Technology, Methods, and Infrastructure for Networked Medical Research - is an organization aimed at supporting clinical, epidemiological and translational research in academic networks. Within the working group for IT infrastructure and quality management, a forum was founded to facilitate the growing number of projects using carefully designed, multiple-used metadata.

Starting with a repository of data elements for clinical trials, it soon became obvious that a singular catalog of isolated items would not satisfy researcher's needs. Using ISO 11179 the solves the problem of designing a sophisticated data model, but a modern clinical data dictionary requires conventions for constructing complex data elements as well as support for a bottom-up community approach and integration of medical terminologies. Furthermore, current trends in clinical research like the secondary use of treatment data, the integration of biobank and genomic data and evolving biomedical ontologies call for providing richer context for data elements: usage in certain vocabularies, revised versions tracking modifications, community ratings, querying of similar/related variants, mapping of conceptual attributes and transformation of representational attributes e.g. for meta-analysis, and regulations for assuring data quality and data privacy.

This contribution will give an overview of current research topics and ongoing implementation in projects in Germany as well as in Europe.

<Return to Top> 

Talk 8

Healthcare: Implementation of an ISO/IEC 11179 based Metadata Registry to foster interoperability of health telematics applications

Mr. Bernhard Rimatzki, FB Informatik,

Prof. Dr. Peter Haas

Introduction/Background

The institution spanning organization and documentation of patient treatments to foster integrated and managed care tends to result in increasing demands for intelligent interoperability of institutional information systems. Thus generic realized interoperability requires overall availability of computer-interpretable structure and semantics of clinical concepts. So a Metadata Registry(MDR)-server with web services as an element of a health telematics platform is essential. This MDR-server can also be used as repository for developers. A MDR is an adequate approach, to administrate object type and attribute definitions or definitions of clinical concepts.

Results

On basis of class model of ISO/IEC11179 V2 [1] we implemented a persistence layer of MDR and additional web services for administration and retrieval of MDR content. Furthermore a versioning of the data definitions is provided. For manual maintenance we implemented a web based user interface, that still use the above mentioned services. Additionally a form generator was realized to create dynamic web forms based on the definitions in the MDR [2]. The input data of these forms are stored in CDA-Level1 documents.

To administrate the semantics of the entities of the MDR, we added associations to the data model of a terminology server based on CTS2 [3].

Conclusion

The specified MDR is suitable to administrate jointly used entity definitions for interoperability of health telematics applications and to provide them computer-interpretable and independent of location and time. The information model is well suitable to administrate complex clinical concepts. For generating dynamic forms additional classes for the form oriented aggregation and representation information must be added. The classes in ISO/IEC 11179 V2 for representing terminologies are insufficient, so therefore we referred to CTS2. For interoperability health telematics applications can retrieve MDR-content and use it for generic interpretation and visualization.

Keywords: Data-Dictionary-Server, Metadata Registry, web service, health telematics platform, dynamic forms

Literature

[1] ISO/IEC11179. 2004. Information technology — Metadata registries (MDR). 2004.

[2] Nadkarni, Prakash M. 2011. Metadata-driven Software Systems in Biomedicin - Designing Systems that can adapt to Changing Knowledge. Heidelberg : Springer, 2011.

[3] HL7-CTS2. CTS 2.0 Specification page. CTS 2.0 Specification page. http://wiki.hl7.org/index.php?title=CTS2.

<Return to Top> 

Talk 9

Sharing of OID Registry Information

Dr. Christof Gessner (TMF e.V., Berlin)
Sylvia Thun (Hochschule Niederrhein, Krefeld)
Stefan Sabutsch (ELGA GmbH, Wien)
Kai Heitmann (Heitmann Consulting and Services, Hürth)

An OID (Object Identifier) is a globally unique identifier for an information object as defined by ISO/IEC 9834  and ITU-T X.660. The OID tree forms a hierarchical structure. Each branch under the root corresponds to an assigning authority. These assigning authorities may in turn designate their own set of assigning authorities that work under its auspices. The authorities assign OIDs within their sub-tree. The leaf nodes on the tree represent the information objects.

In healthcare OIDs are used to reference terminologies, ontologies, identifier namespaces and other information objects in data exchange processes, often across institutional or even national boundaries. OID registries are established in order to maintain and provide information about an OID, its definition and additional information that is needed for any meaningful use of the object identified. The process of registering an OID includes its descriptive data. These metadata also enable implementers to search the registry for a given concept, effectively discouraging the assignment of multiple OIDs to identical or similar information objects and encouraging re-use of concepts already defined and in use.

Today, due to lack of standardization of metadata content and structure, existing OID registries are not compatible. Contents, attributes and rules of the assignment of OIDs of existing registries are incompatible and often dissimilar. An information model was developed providing all classes, attributes and their properties to support a common understanding of the requirements of an OID registry. Mandatory and optional metadata to be recorded in any repository of OIDs are specified based on this information model. Data structures for the exchange of OIDs and their metadata between registries are also specified.

The national affiliates of HL7 in Austria, Switzerland and Germany work together in setting up compatible OID registries and in establishing the infrastructure for the exchange and use of common terminologies and identifiers.

<Return to Top> 













Copyright © 2012 Metadata Open Forum