xml 2004 medical ontology, owl, e-business registry & repository managing medical ontologies,...

XML 2004 Medical Ontology, OWL, e-Business Registry & Repository

Managing Medical Ontologies, OWL using an e-Business Registry : Repository

• e-Health Service : Interoperate & Mediate• BCM : Business Centric Guidelines• e-Business : Co-operating Registry & Repository• Semantic Web : Smarter Knowledge Tagging• Medical Documents : Generic & Specific• Clinical Metadata : Communities of Interests• Ontology Management : Controlled Vocabularies• Choice Points : Contextual Validation

• Semantic Specifications : Concept URI Collaboration

Deep Dive


Carl Mattocks @ CHECKMi.com

• Co-Chair (ISO/TS 15000) ebXMLRegistry Semantic Content SC • Co-Chair Business Centric Methodology TC • CEO CHECKMi

– Multiple Agent Training Engine (MATE) binding 1000 databases– Federation of Vendor Repositories

• Supporting Operational Metadata Information Needs at Metlife (CA AllFusion, Logidex, Unicorn, Tivoli)

– Knowledge Compendium (.Net open source code)• Crosswalk UDDI / ebXML Registry / IMS Repository• Multiple Taxonomies, Z39.50 SRW Gateway, XACML


21st Century : e-Knowledge Economy

Human CapitalCompetitive Capability

Co-opetition Skills Capacity Productivity & Efficiency

Communities of Interest Community TrustCommunities of Practice

e-Ology e-Business e-Population

Co-operating Registries Peer to Peer Repositories

Service Architecture & Networking Infrastructure

Standards for Business Process Workflow & Quality

of Service Agreements

Best Practice Collaboration

Protocols & Policies

Knowledge & IP

Semantic Information Grids


e-Society : Freely Exchanged Knowledge

e-Health

e-Learning

e-Research

e-Government

e-Community of Interest

e-Community of Practice

e-Business

e-Science

Knowledge Mediation &

Interoperation


e-Health Service Objectives : Inform, Interoperate Involve, Improve

Bringing information tools to the point of care by the EHR (Electronic Health Record) systems used in physician offices and hospitals aka "Inform Clinical Practice"

Building an interoperable health information infrastructure, so that records follow the patient and clinicians have access to critical health care information when treatment decisions are being made aka "Interconnect Clinicians"

Using health information technology to give consumers more access and involvement in health decisions aka "Personalize Care"

Expanding capacity for public health monitoring, quality of care measurement, and bringing research advances more quickly into medical practice aka "Improve Population Health"


e-Health Care Knowledge Requirements : Observe, Experience, Expand

Publish Observations ‘whenever patient and professionals need it ‘

Accurate - prevent errors originating with partial notes

Effective - eliminate duplicative care records, lowering administration costs

Compliant - with regulations, privacy laws, etc e.g. HIPAA

Disseminate Experience regarding Treatment Options

Speed the application of research findings; reducing (estimated 17 years) time it takes to be fully integrated into practice

access to specialty information; especially beneficial for medically under served areas

Expand Sharing of Public Health Alerts

Mercury Global Bio-Threat Monitoring exercise

– http://www.checkmi.com/wpapers/MercBioDCheckmiScout.pdf


e-Health Knowledge Mediation Goals : Re-Use, Re-Usable, Re-Useful

• Support Communities of Practice & Communities of Interest

– Locally, regionally, nationally, internationally

• Use Natural Knowledge models suitable for wide audience not just IT

support specialists e.g. the Patient

– Ability to mitigate business vocabularies and multi-lingual challenges

• Automate sharing Semantically rich MetaData

– provide discovery, understanding, and exchange of Electronic Health Record

(EHR) information

– Enable Providers, Pharmacy, Health Plans / PBM’s (pharmacy benefit

management), and other Agencies to create service processes that support

internationally agreed business process definitions, trust mechanisms and

process control methods


e-Prescribing Knowledge Application : Appropriate, Alternate, Approve

Source: Electronic Prescribing: Toward Maximum Value and Rapid Adoption - eHealth Initiative


Business Centric Methodology : SOA Knowledge

• Enables precise communication between business users and technical experts & enterprise applications and their respective business partner systems

• Involves a layered approach for strategically managing Service Oriented Architecture artifacts and constraints while achieving semantic interoperability


BCM Conceptual Layer : Knowledge Semantics

Role - semantically connect facts and facets

Provide trace-ability from business vision to system implementation

Ensure alignment of business concepts with automated procedures

Facilitate faster information utilization between business parties

Enable accurate information discovery and synchronization

Integrate information by interest, perspective or requirement

Resource - Controlled Vocabularies, Thesauri, Ontologies, RDF, OWL

Leverage COP knowledge organization systems

SKOS Core http://www.w3.org/2004/02/skos/core/

Classification schemes for Federated e-Health artifacts and constraints e.g. archetypes


Federated Knowledge Management Needs

• A Federated Registry not only acts as an interface to a Repository of stored content, it formalises how information is to be registered and shared beyond a single enterprise or agency.

• Specifically, a federated content management capability is required when there is a need for managing and accessing metadata across physical boundaries in a secure manner. Those physical boundaries might be the result of community-of-interest, line of business, system, department, or enterprise separation.

• Irrespective of the boundary type, a Federated Registry must enable information users to seamlessly access, share and perform analysis on information, including :

– map of the critical path of information flowing across a business value chain;

– quality indicators such as statements of information integrity, authentication and certification; and

– policies supporting security and privacy requirements.


Co-operating Knowledge Registry & Repository : Federated Content

Mappings for the critical path of information flowing across a business value chain e.g. ordering & payment of e-Prescription

Quality indicators such as statements of information integrity, authentication and certification e.g. electronic signature used for e-prescribing

• Policies supporting security and privacy requirements e.g. HIPAA

• e-Business XML (e.g. XSDs, Elements, Templates) and Semantics

Registry D

Registry E

Registry B

Registry A

Registry C


Universal XML Schema Management Knowledge

Interacting entities need re-usable data items with known semantics

• These items are required as XML Schema components

• The Schema (document) and the Components must be shared within and often outside the organization

• Components may use external components such as code lists

• Domain Components must be managed cooperatively

• Core components must be common

• Data definition arbitration agreements are essential

• Data types must be standardized


ISO/TS 15000 OASIS ebXML Registry

One interface (HTTP, SwA, ebMS) to classified knowledge content :

– Electronic Forms– Web Services WSDL / WSRP– Collaboration Agreements– Business Process

Requirements, Objects, Data– Domain specific Semantics

and Relationships between Assets & Artifacts

– SQL queries and APIs

Registration

Business Process and Information Models(Compliant to the ebXML Meta Model)

Model to XML Conversion

InternalBusiness

Application

Implementers

Retrieval of Profiles &new/updated ebXML Models

Retrieval of Profiles &new/updated ebXML Models

RegisterCollaboration

Protocol Profile(CPP)

Retrieval of ebXMLModels and Profiles

Build Build

CollaborationProtocol

Agreement (CPA)

Payload

CPA

Gov

erns

CPP

Business ServiceInterface

InternalBusiness

Application

Business ServiceInterface

Registry ServiceInterface

Registries

RegisterCollaboration

Protocol Profile(CPP)


Web Service Knowledge Registration & Classification

• WSDL is an XML format for describing network services as a set of endpoints operating on messages containing either document-oriented or procedure-oriented information. The operations and messages are described abstractly, and then bound to a concrete network protocol and message format to define an endpoint. Related concrete endpoints are combined into abstract endpoints (services).

• Registration & Classification provides for a number of uses: Find a single web service element from among many

Convey semantic content that may be incompletely specified by other attributes - such as names and definitions

Disambiguate between data elements of varying contextual classifications


ebXML Registry Information Model (RIM) - Knowledge Classification

RegistryObject

ClassificationNode Classification

RegistryPackage ExtrinsicObject Service

AssociationRegistryEntry

ClassificationScheme

The structure of the classification scheme may be defined internal or external to the registry

A Classification instance identifies a ClassificationScheme instance and taxonomy value defined within the classification scheme

Classification trees constructed with ClassificationNodes are used to define the structure of Classification schemes or ontologies

Person


e-Business Knowledge : XML Registry Version 3

• Registration and classification of any type of object

• Namespaces defined for certain types of content

• Messages defined as XML Schemas

• Taxonomy hosting, browsing and validation

• Links to external content

• Built-in security (SAML)

• Event notification, Event-archiving (audit trail)

• Life cycle management of objects

• Federation for inter-registry relocation, replication, references – metadata is stored in one registry; a registry may cooperate with multiple

federations for the purpose of federated queries


Semantic Spectrum of e-Business Knowledge

Business Usage Semantics

Data Usage Semantics

System Interoperability

Service Quality

Data Dictionary

Info Engineering Model

Object Oriented Model / Thesaurus

DTD / XML Schema / Dublin Core

RDFS / UML / Topic Map / UNSPSC

OWL-S Ontology + Templates + SKOS + Archetypes


WC3 Semantic Web : Smarter Knowledge Tagging

XML provides syntax for structured documents, but imposes no semantic constraints on the meaning of these documents.

XML Schema is a language for restricting the structure of XML documents and also extends XML with datatypes.

RDF is a datamodel for resources and relations between them.

RDF Schema is a vocabulary for describing properties and classes of RDF resources, with a semantics for generalization-hierarchies of such properties and classes.

• OWL adds more vocabulary for describing properties and classes e.g. relations between classes, cardinality, equality.

• SKOS-Core is an [RDF & OWL] schema for representing thesauri and similar types of Simple Knowledge Organization Systems .


WC3 Semantic Web : RDF (Resource Description Framework) Knowledge Graphs

• Uniform Resource Identifiers e.g. http://www.w3.org/people/em/

• Application Statements

• individual label #me

– genus #person

– property #fullname value ‘Eric Miller’

– property #mailbox value ‘mailto:[email protected]’


RDF / XML : RDF statement using XML-based syntax

• Eric Miller Graph

• <?xml version="1.0"?>

• <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"• xmlns:contact="http://www.w3.org/2000/10/swap/pim/contact#">

• <contact:Person rdf:about="http://www.w3.org/People/EM/contact#me">• <contact:fullName>Eric Miller</contact:fullName>• <contact:mailbox rdf:resource="mailto:[email protected]"/>• <contact:personalTitle>Dr.</contact:personalTitle> • </contact:Person>

• </rdf:RDF>


RDF Schema & OWL (Web Ontology Language)

• RDF Schema is an extension to RDF framework used to describe application-specific classes and properties e.g. Dr. e-Prescribing.

• Classes in RDF Schema allow resources to be defined as instances of classes, and subclasses of classes. – rdfs:subClassOf, rdfs:subPropertyOf

• OWL describes the structure of a domain in terms of classes and properties– owl:sameClassAs, owl:samePropertyAs, owl:disjointWith,

– owl:sameIndividualAs, owl:differentIndividualFrom,

– owl:inverseOf, owl:transitiveProperty,

– owl:functionalProperty, owl:inverseFunctionalProperty


OWL-S (OWL web Service) Service Profile

• When registering and linking descriptors of a Web service, WSDL and OWL-S a key factor is the Service Profile which is used to concisely represent the service in terms of capabilities, provenance, and operational parameters

– contact information that refers to the entity that provides the service

– functional description of the service is expressed in terms of the transformation produced by the service

– inputs required by the service and the outputs generated

– the preconditions required by the service and the expected effects that result from the execution of the service

– category of the service within a classification system– quality rating of the service

– feature specific constraints e.g. geographic availability


e-Medical Service : HL7 Clinical Document Architecture

• HL7 (Health Level) V3 Reference Information Model (RIM)• > domain information models (DIMs) > restricted message information

models (RMIMs) > common message element types (CMETs) > hierarchical message definitions (HMDs) > message schemas in XML

• e.g. e-prescription would reference multiple DMIM's (Domain Message Information Model) and payload “value sets” for Orders, Observations, Pharmacy, Medications, Patient Administration.

• HL7 Clinical Document Architecture (CDA) a generic model for the communication of clinical documents

• Document-level, section-level and entry-level Templates can be used to constrain the generic CDA


CEN ENV 13606-2 : Electronic Health Record

• European Committee for Standardization Electronic Health Record employs four sub-categories:– Folder: High-level subdivisions of the entire EHR for a patient,

usually grouping entries over long time-spans within one organization or department, or for a particular health problem.

– Composition: A set of record entries relating to one time and place of care delivery; grouped contributions to an aspect of health care activity; composed reports and overviews of clinical progress.

– Headed Section: Sub-divisions used to group entries with a common theme or derived through a common healthcare process.

– Cluster: Low-level aggregations of elementary entries (Record Items) to represent a compound clinical concept.


Good European Health Record GEHR

• Extensions on the GEHR approach uses a formal semantic model, known as the GEHR Object Model (GOM) which provides concepts at a number of levels: – EHR and Transaction level

– Navigation level

– Content (e.g. observation, subjective, instruction) level

• Clinical models are expressed in the form of XML-Schema archetype constraint definitions – Transaction types, e.g. contact, summary, etc;

– Navigational headings e.g. EPR workcards

– Clinical content types e.g. lab-results, prescriptions, including their structure (list, table, series etc)


openEHR Container Classes

• openEHR specifications separate generic features of any EHR from the domain-specific information about the patient or data subject– EHR Extract class contains all information that is to be transferred to

another HER e.g. HL7 CDA.

– EHR Folder class allows information within an EHR to be organized. • Composition class that contains information committed to the EHR by a

clinician. • Section - this class allows information within a composition to be

segmented • Entry - this class contains meaningful information that is to be processed

by the machine and read by the clinician

– Templates & Clinical Vocabulary Archetypes together frame clinical domain content e.g. valid values of e-Prescription XForm


Consolidated Health Informatics (CHI) portfolio of Clinical Vocabularies

• National Council on Prescription Drug Programs (NCDCP) standards for ordering drugs from retail pharmacies.

• The Institute of Electrical and Electronics Engineers 1073 (IEEE1073) series of standards that allow for health care providers to plug medical devices into information and computer systems

• Digital Imaging Communications in Medicine® (DICOM®) standards that enable images and associated diagnostic information to be retrieved and transferred from various manufacturers' devices

• Laboratory Logical Observation Identifier name Codes® (LOINC®) to standardize the electronic exchange of clinical laboratory results.


LOINC 10,000 + Archetype Material

• Logical Observation Identifiers Names and Codes developed for use as observation identifiers in standardized messages exchanged between clinical computer systems

– Includes entries for vital signs, hemodynamics, intake/output, EKG, obstetric ultrasound, cardiac echo, urologic imaging, gastroendoscopic procedures, pulmonary ventilator management, selected survey instruments, and other clinical observations


Integrating the Healthcare Enterprise (IHE) XDS

• The IHE Integration XDS Profile for Cross Enterprise Clinical Documents Sharing is document-content neutral.

• XDS documents contain simple text, formatted text, images or structured and vocabulary coded clinical information.

• Based on the ebXML Registry specification (implemented with freebXML Registry) the XDS defines the (document) Registry as an actor that maintains metadata about each registered document in a document entry. – Enforces policies at the time of document registration

– Registry meta data includes a link to the (document) Repository where the actual document is stored which in turn assigns and maintains a unique identifier for each document


Key XDS Concepts

• A XDS Affinity Domain is made of a well-defined (federated) set of Document Repositories and Document Consumers that have agreed to share the clinical documents.

• A shared clinical record is called an EHR-LR (Longitudinal Record)• The care delivery systems used within an enterprise for managing episodes of

patient care are called the HER-CR (Care Delivery Record)• An XDS Document is a composition of clinical information that contains

observations and services for the purpose of exchange with the following characteristics: Persistence, Stewardship, Potential for Authentication, and Wholeness (characteristics that are well defined in the HL7 CDA)

• An XDS Folder provides a collaborative mechanism for several XDS Document Sources to group XDS Documents for a variety of reasons. XDS Documents may be placed into an existing Folder at any time, as long as they relate to the same patient.

• A Document Source may only contribute documents with Document Codes and Health Facility Codes that draw from a Vocabulary Value Set that is approved by

the Affinity Domain.


NIST HL7 Experimental freebXML Registry DMIMs, RMIMs, HMDs, and MessageTypes

• HL7-specific classification schemes, especially the code hierarchies for the structural attributes in the RIM.

• RMIM static models from technical domains, including finance, Patient Administration, Scheduling, Laboratory Orders, Research Trials, Pharmacy, Medical Records, Common Message Types, Message Control, Master File, and Clinical Documents

• RMIM static models from the CMETs domain, each with a an association to the DMIM it is derived from (using external identifiers to the HL7 identification schemes) with external links to diagrams and descriptions, and a number of classifications by specific codes that are fixed by constraints on the RMIM.

• RMIMs leads to one or more derived HMDs and MessageTypes.• Each registered artifact has ExternalLinks to its base UML diagrams, long html

descriptions, and other visual display aids for presentation of base classes, attributes, relationships, and constraints.

• OWL-related template artifacts submitted by HL7 participants active in the Templates technical committee


Ontology Annotated e-Health Services

• e-Health service information can be managed by two different e-Health service entities using different message structures.

• ARTEMIS project providing a standard way of accessing the data by registering & storing – ontologies based on existing healthcare standards,

– the semantic mapping between these ontologies,

– invoke each others web services by semantic mediation.

• Discovery of Services stored in a Registry need semantic service registry query mechanisms that leverage previous research linking OWL to the Registry Information Model


Linking OWL to ebXMLRegistry Objects


Community of Interest : Multiple-levels of Ontologies

• BCM COI & XDS Affinity Domain defined by formal or informal organizational structures

• Upper level ontology is focused on the non-volatile language and principles of a domain

• Lower ontology is focused on the knowledge specific to particular community of practice (as formulated by the recognized experts)


BCM Choice Points : XForm & Templates

• Electronic PRocess (EPR) Web Service Portal Technology e-Health Service employs semantically rich XML templates

• Mayo Clinic Template types– Constraint : operate on code combinations permitted in a code phrase – Style : allow the distribution of semantic details between name and value in

name-value pairs– Pattern : defining specific class of items– Document : construct of particular data elements. – Profile : dataset of selected items that describes an existing entity– Extension : constrains or modifies a profile template – Meta : ordered aggregation of multiple other templates, such as an EHR

template for the Electronic Health Record


BCM Choice Points : Templates & Archetypes

• Template & Archetype Choice Points significantly aids to comprehensibility, alignment, while promoting tracing and accountability when : – archetype is a computable expression of a domain level (clinical)

concept in the form of structured constraint statements, based on some reference model (RMIM)

– archetypes are aligned with Affinity Domain concepts

– archetypes all have the same formalism i.e. may be part of a COP ontology but belong to only one or other ontological level

– template is used to narrow the choices of archetypes for local or specific purposes (DMIM).

– archetype defines constraints on reference model instances which express valid structure (i.e. composition, cardinality).

– archetype defines constraints on instances of a reference model which express valid types and values.


ebXML Registry Semantic Content Management

• Providing support for semantic links based on RDF and OWL the check list of requirements includes :– Re-usable dictionaries of noun definitions for specific industry

domains

– Re-useful dictionaries of Business Process catalogues

– Ontology searching and browsing

– Collaborative (COP) ontology development

– Classify content using OWL Ontology class hierarchies

– Discover content using semantic queries

– Attach semantic Objects to Stored Content

– Semantic reasoning …


With thankful links to :

Members of ebXMLRegistry Semantic Content SChttp://www.oasis-open.org/committees/sc_home.php?wg_abbrev=regrep-semantic

Members of Business Centric Methodology TChttp://www.oasis-open.org/committees/download.php/5931/BCM%20Executive%20Brochure.pdf

Contact: [email protected]

http://www.oasis-open.org/committees/sc_home.php?wg_abbrev=regrep-semantic











xml 2004 medical ontology, owl, e-business registry & repository managing medical ontologies,...

Documents

medical ontology

medical practice

health decisions

population health slide

health information technology

medical documents

medical ontologies

public health monitoring