Upload
ragia-ibrahim
View
113
Download
0
Tags:
Embed Size (px)
Citation preview
Pdf name UMLS_tutorial.pdf
"this doc is a copy and past from original one
Tutorial T20
September 8, 2004
Olivier Olivier Bodenreider Bodenreider Jan Willis Jan Willis
William Hole William
Outline
1.1 What is UMLS?
How to use to use the the UMLS?
Obtaining Obtaining a license license
Remote Remote access access
Knowledge Knowledge Source Server (UMLSKS) Source Server (UMLSKS)
UMLSKS Application UMLSKS Application programming interface (API) interface (API)
Local installation Local installation and customization (MetamorphoSys)
Questions
What is the UMLS? NOT END- USER Application
1.2 Introduction
Overview through an example
The three UMLS Knowledge Sources
UMLS Metathesaurus
UMLS Semantic
UMLS Lexicon
Univifed medical language system
Started in 1986
National Library of Medicine National Library of Medicine
“Long “Long-term R&D project” term R&D project”
Complementary to IAIMS
1 The Unified Medical Language System What is it and how to use it?
1
1.3 Overview through example
1.3.1 Addison’s disease
Addison's disease is a rare endocrine disorder الصماء الغدد اضطرابات
Addison's disease occurs when the adrenal glands الكظريه الغدد
do not produce enough of the cortisol
For this reason, the disease is sometimes
Called
chronic adrenal insufficiency, or
hypocortisolism
1.3.2 Adrenal insufficiency "Clinical variants"
Primary / Secondary
Primary: lesion االفه of adrenal glands themselves
Secondary: inadequate Secondary: inadequate كافية هرمون of ACTH االفراز secretion غيرby the Acute / Chronic نخامي الغده
Isolated / Polyendocrine
deficiency syndrome نقص متالزمه
1.3.3 Addison’s disease Symptoms
Fatigue
Weakness
Low blood pressure
Pigmentation of the skin (exposed and non Pigmentation of the ………..
2
1.3.4 AD in medical vocabularies Synonyms: different terms different
Contexts: different hierarchies
1.4 Organize terms Synonymous terms clustered into a concept Preferred term Unique identifier (CUI)
Addisonian syndrome eponym اسم على مسمى
Bronzed disease المرض البشره برونزيsymptoms االعراض
Addison melanodermaAsthenia pigmentosa الصباغي الوهنPrimary adrenal deficiency
Clinical variantsPrimary adrenal insufficiency
Primary adrenocortical insufficiencyChronic adrenocortical insufficiency
3
eponym
4
5
6
Organize concepts
One graph instead of instead of
Multiple trees
(Multiple inheritance)
B
C
7
1.5 Relate to other concepts Additional hierarchical relationships
o link to other trees o make relationships explicit
Non-hierarchical relationships Co-occurring concepts Mapping relationships
Relate to other concepts
8
1.6 Categorize concepts High -level categories (semantic types) Assigned by the Metathesaurus editors Independently of the hierarchies in which these concepts are located
1.7 How do they do that? Lexical knowledge
Semantic pre-processing
UMLS editors
1.7.1 Lexical knowledge
1.7.2 Semantic pre-processing Metadata in the source vocabularies
Tentative مؤقت categorization
Positive (or negative) evidence for tentative lexical features synonymy relations based on lexical features
9
1.7.3 UMLS editors: Additional knowledge
1.8 UMLS Summary
Synonymous Terms clustered into concepts Unique identifier Finer granularity Broader scope Additional hierarchical relationships Semantic categorization
10
Metathesaurus
o conceptso Inter-concept relationships
Semantic Network o Semantic types o Semantic network relationships
Lexical resources o SPECIALIST Lexicon SPECIALIST Lexicono Lexical tools Lexical tools
1.9 Biomedical terminologies
1.9.1 General vocabularies Anatomy drugs ( drugs ( RxNorm, First DataBank,Micromedex) , Micromedex) SPN) medical devices (UMD, SPN)
1.9.2 Several perspectives clinical terms (SNOMED CT) Information sciences (MeSH, CRISP) administrative terminologies (ICD-9-CM, CPT CM, CPT-4) data exchange terminologies (HL7, LOINC)
1.9.3 Specialized vocabularies nursing dentistry (CDT) oncology (PDQ) psychiatry (DSM, APA) adverse reactions (COSTART, WHO ART) primary care (ICPC)
1.9.4 Terminology of knowledge bases (AI/Rheum, DXplain, , QMR)
The UMLS serves as a vehicle for the regulatory standards
2 UMLS Knowledge Sources
11
1.10Metathesaurus Basic organization
1.10.1 Concepts Synonymous terms are clustered into a conceptProperties are attached to concepts, e.g., Unique identifier UniqueDefinition
1.10.2 Inter-concept relationshipsConcepts are related to other conceptsProperties are attached to relations, e.g., Properties are attached to relations, e.g.Type of relationship Type of relationship
1.10.3 Addison’s Disease: Addison’s Disease: Concept
12
1.10.4 Metathesaurus Concepts
13
1.10.5 Cluster of synonymous terms of synonymous terms
1.10.6
1.10.7
1.10.8
1.10.9
1.10.10
1.10.11
1.10.12
1.10.13
1.10.14
1.10.15
1.11Metathesaurus RELATIONSHIP Symbolic relations: Symbolic relations: ~9 M pairs of concepts Statistical relations ~7 M pairs of concepts co-occurring concepts Mapping relations: Mapping relations: 100,000 pairs of concepts Categorization: Relationships between concepts: and semantic types from the Semantic
Network
1.11.1 Symbolic relations Relations:- Pair of “atom” identifiers Pair of “atom” identifiers Type Attribute (if any) List of sources (for type and attribute)Semantics of the relationship: Defined by its type [and attribute]Source transparency:
The information is recorded at the “atom” level
14
1.11.1.1 Symbolic relations type
Hierarchal
Parent / Child PAR/CHD
Broader / Narrower than RB/RN
Derived from hierarchies
Siblings (children of parents)
SIB
Associative
Other RO
Various flavors of near Various flavors of near-synonymy
Similar RL
Source asserted synonymy SY
Possible synonymy
RQ
1.11.1.2 Symbolic relations attributes Hierarchical
o isa -a ( is a kind-of) o part of
Associative o location-of o caused byo treats
Cross -references (mapping)
15
16
1.11.1.3
1.12 Semantic Network
1.12.1 Semantic typesTree structure 2 major hierarchies 2 major hierarchies
Entity o Physical Object o Conceptual Entity
Event o Activity o Phenomenon or Process
1.12.2 Semantic network relationships hierarchical (isa = is a kind of )
o among types Animal isa Organism Enzyme isa Biologically Active Substance
o among relations treats isa affects
non-hierarchical o Sign or Symptom Sign diagnoses Pathologic Function o Pharmacologic Substance treats Pathologic Function
Biologic Function” hierarchy (isa)
17
18
Associative (non--isa) relationships
Semantic serve as high level categories Semantic assigned to Metathesaurus concepts, concepts, independently of their position in a hierarchy A relationship between 2 Semantic Types (ST) is a possible link between 2 concepts that have been assigned to STs The relationship may or may not hold at the concept level Other relationships may apply at the concept levelRelationships can inherit semantics Relationships can inherit semantics
19
1.13 Lexical resources
1.13.1 SPECIALIST Lexicon Content
o English lexicon o Many words from the biomedical
200,000+ lexical items Word properties Word properties
o morphology o orthography o syntax
the lexical tools Used by the lexical tools
1.13.1.1 Morphology Inflection
o noun nucleus ,nucleio verb cauterize, cauterizes, cauterized, cauterizingo adjective red, redder, reddest
Derivation o verb noun cauterize --- cauterization o adjective noun red -- redness
1.13.1.2 Orthography Spelling variants Spelling variants
o oe/e oesophagus - esophaguso ae/e anaemia - anemiao ise/ize cauterise - cauterizeo genitive mark Addison's disease
Addison disease Addisons disease
1.13.1.3 Syntax Complementation
o verbs intransitive I'll treat. transitive He treated the patient. ditransitive He treated the patient with a drug.
o nouns prepositional phrase Valve of coronary sinus
Position for adjectives
20
21
1.13.2 Lexical tools Lexical tools Tomanage lexical variation in biomedical terminologies Major tools
o Normalization o Indexes o Lexical Variant Generation program (lvg)
Based on the SPECIALIST Lexicon engines Used by noun phrase extractors, search engines
1.13.2.1 Normalization
e.g. and steps Hodgkin’s diseases, NOS
Remove genitive Hodgkin diseases, NOS Remove stop words Hodgkin diseasesLowercase hodgkin diseases,Strip punctuation hodgkin diseasesUninflect hodgkin diseaseSort words disease hodgkin
Normalization Applications Model for lexical resemblance Help find lexical variants for a term
o Terms that normalize the same usually share the same LUI Help find candidates to synonymy among terms Help to map input terms to UMLS concepts
22
1.13.2.2 Indexes Word index Word
o word to Metathesaurus strings o one word index per language
Normalized word index Normalized word indexo normalized word to Metathesaurus stringso English only
Normalized string index o normalized term to Metathesaurus stringso English only English only
1.13.2.3 Lexical Variant Generation program (lvg) Tool for specialists (linguists) Performs atomic lexical transformations
o generating inflectional variants o lowercase lowercaseo …
Performs sequences of atomic transformations o a specialized sequence of transformations provides the normalized form of a term
(the norm program)
23
Obtaining a license Remote access
o Knowledge Knowledge Source Server (UMLSKS) o UMLSKS Application UMLSKS Application programming interface (API)
Local installation and customization ( MetamorphoSys)
1.14Obtaining a license
3 How to use the UMLS?
24
1.15Remote access UMLS Knowledge Source Server:
http://umlsks.nlm.nih.gov Web search interface Application programming interface (API)
1.15.1 UMLS Knowledge Source Server: Web search interface
25
26
1.16
27
1.17
28
29
30
31
32
33
1.17.1
1.17.2 Knowledge Source Server :Application Programming Interface
34
35
1.17.3 Local installation and ( MetamorphoSys) Tool distributed with the UMLS Multi-platform Java software The UMLS installation and customization wizard
o Installs Knowledge Sources to local o Subsets and customizes a local Metathesaurus
…………………….i didn’t complete implementation issues
36