Terminology

1
Terminology

• CG Chute
• HR Solbrig

2
Who we are and what we are about

• Biomedical Informatics research group with
  emphasis
• Classification and Ontologies
• Natural Language Processing
• Information Retrieval
• Active in standards communities
• HL7, ISO, ANSI HISB, PharmGKB,
• Development and maintenance of effective
  indexing and access to comprehensive patient
  data.

3
What does effective indexing and access to
comprehensive patient data entail?

• Knowing what is available and where to find
  it.... (metadata)
• Shared understanding of the meaning and intent of
  databases, forms, tables, columns and field
  content.
• Managing the boundary between information models
  and structured vocabulary

4
Where we all want to be

• Interoperable information
• Comparable and consistent meanings
• Many systems based on a common, distributed
  information base
• Formal, machine readable metadata linked to a
  shared vocabulary
• Standardized, interlocking vocabularies to glue
  it all together

5
Terminological Resources

• Q How do we convince developers and users to
  adopt a common set of terminologies for modeling
  and content?
• A1 Make doing so cost effective
• Make terminologies readily available
• Standardize access and distribution
• Create open-source software packages that to
  lower the cost of entry
• A2 Create a demand
• Standardize, proselytize and demonstrate

6
Preliminary Terminology ServerDesiderata (AMIA
99)

• Word Normalization
• Word Completion
• Target Terminology Specification
• Spelling Correction
• Lexical Matching

• Term Completion
• Word Level Synonymy
• Semantic Locality
• Term Composition
• Term Decomposition

7
Mayo HeritageTerminology Services

• YATN Yet Another Terminology Navigator 96
• Metaphrase (co-develop with LTI/Apelon) 97
• LQS Lexicon Query Services 98
• MTS Mayo Terminology Services 00

8
Vision

• A heterogeneous, distributed collection of
  ontologies / terminologies...
• ...available through a common API
• ...joined through shared indices
• ...able to be extended, enhanced an annotated in
  a loosely-coupled, distributed fashion
• ...available in web-space on web-time

9
A Heterogeneous, Distributed Collection of
Terminologies
HL7
UNPCS
SNOMED-CT
ICD-10
NCI
ICD-9
ISO CC
10
Available through a Common API
O T S
HL7
UNPCS
O T S
SNOMED-CT
ICD-10
O T S
O T S
NCI
O T S
O T S
ICD-9
O T S
ISO CC
11
Joined Through Shared Indices
Index
M T S
12
Extended, enhanced an annotated in a
loosely-coupled, distributed fashion
O T S
Localized Extensions
13
Available in web space...
HTTP FTP SOAP XML RMI CORBA ...
HTML XML RDF OWL CSV ...
14
... on web-time
Review
Localized Extensions
Review
15
The Lexical Grid
LexGrid Model
16
LexGrid Implementation(open source foundations)

• LDAP Core (name directory)
• Lucene Index
• OTS API (Java/SOAP/Web Services)
• Import/Export/Revision Tools
• LDIF
• RDF/OWL
• Editors / viewers / prototypes

• Protégé Clips
• XML Schema,

17
ImplementationOverview
Protégé
LexGrid Editor
1) Implementation
LexGrid
OTS
Apelon Tools
OTS
LDAP Tools
OWL Tools
V-OWL
18
Today Development Future
Protégé
C T S
LexGrid Editor
Lucene Index
LexGrid
OTS
OTS
Apelon Tools
Transformation Toolkit
LDIF
XML
OWL
Protégé
Apelon
...
OTS
LexGrid Model
LDAP Tools
OWL Tools
V-OWL
19
Standards

• Application Programming Interface
• CTS (Common Terminology Services)
• underway in HL7
• OTS (Open Terminology Services)
• CTS v 2.0
• Model
• LexGrid(LDAP) IETF
• V-OWL W3C
• Protégé (de-facto)

20
Content

• HL7 V3
• RxNorm
• NCI Metathesaurus
• ISO 3166 Country Codes
• ISO Language Codes
• Mime Types

• UNSPSC
• UMLS Semantic Net
• HL7 OID Registry
• ...

21
Content

• ISO4217
• CIHI Clinical Service Reason
• CIHI Adjudication Reason
• Canadian Classification of Health Interventions
• CAN/CSA-Z795-96 (R2001)
• North American Industry Classification System
• IETF RFC 1766
• DICOM Class Label
• CPT-5
• CLIP
• CEN ECG diagnostic codes
• UCUM
• CDC Surveillance
• AS4 Neurophysiology Codes
• American Type Culture Collection
• American College of Radiology finding codes
• MDNS
• WHO Adverse Reaction Terms
• WHO ATC

• Medical Dictionary for Drug Regulatory Affairs
  (ME
• NAACCR
• National drug codes
• Medispan GPI
• Medical Economics Drug Codes
• Medical Economics Diagnostic Codes
• Medispan Diagnostic Codes
• First DataBank Drug Codes
• First DataBank Diagnostic Codes
• COSTART
• Chemical abstract codes
• CDC Vaccine Manufacturer Codes
• Read Classification
• CDC Vaccine Codes
• CDC Methods/Instruments Codes
• CDC Analyte Codes
• Unified Medical Language
• Universal Product Code
• UCDS

22
Content

• SNOMED International
• POS Codes
• Systemized Nomenclature of Medicine (SNOMED)
• Omaha System
• Japanese Chemistry
• IUPAC/IFCC Component Codes
• IUPAC/IFCC Property Codes
• International Classification of Sleep Disorders
• ICHPPC-2
• International Classification of Diseases for
  Oncol
• ICD9
• ICCS
• HIBCC
• ICD-10 Procedure Codes
• Health Outcomes
• FDA K10
• Euclides Lab equipment codes
• Euclides Lab method codes
• Euclides quantity codes

• ICD-10
• The Read Codes Version 2
• The Read Codes Four Byte Set
• Multum Lexicon
• MEDCIN
• ABCcodes
• ISO 11073-10101 Health informatics -
  Point-of-care
• Nursing Management Minimum Data Set
• Perioperative Nursing Data Set
• National Uniform Billing Council, UB 92
• NANDA
• ICD-9CM
• CAMNCVS
• ISBT
• Home Health Care
• NOC
• Nursing Intervention Classification
• HCFA Procedure Codes (HCPCS)
• CDT-2 Codes

23
LexGrid and the UMLS

• UMLS
• Centrally maintained
• Integrated and cross-referenced
• Incorporates many standard terminologies
• LexGrid
• Locally maintained
• Local extensions and enhancements
• Incorporates standard, non-standard and
  experimental terminologies

24
LexGrid and the UMLS

• LexGrid augments UMLS by
• Making UMLS ubiquitous
• Supports sub-setting
• Allowing dynamic UMLS extensions
• Allowing localizations and non-clinical
  terminologies

25
LexGrid and caBIGWhat we have to offer

• A comprehensive terminology model in
• LDAP Schema
• XML Schema
• Java
• (V-) OWL
• Protégé
• Supports broad content
• UMLS, GO, OBO, HUGO, caEVS, etc.

26
LexGrid and caBIGWhat we have to offer

• Distribution Mechanisms
• LDAP push/pull distribution
• LexGrid XML
• XML Schema
• OWL
• CSV
• (whatever else you can put out in XSLT)

27
LexGrid and caBIGWhat we have to offer

• Software (open source)
• CTS implementation
• OTS implementation
• Terminology editor
• Protege integration
• Includes a model for terminology/data integration

28
LexGrid and caBIGWhat we have to offer

• Standard Terminology Content
• In LexGrid format
• Available on the web
• Clonable to local servers
• (see earlier slides)