Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model Ontology-based reinterpretation of the SNOMED CT context model Catalina Martínez -Costa Stefan Schulz Medical University of Graz (Austria) ICBO 2013 Montreal
20
Embed
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model Ontology-based reinterpretation of the SNOMED CT.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
Ontology-based reinterpretation
of the SNOMED CT context model
CatalinaMartínez-Costa
Stefan Schulz
Medical University of Graz (Austria)
ICBO 2013Montreal
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
"… integrating resources that were developed using different vocabularies and different perspectives on the data. To achieve semantic interoperability, systems must be able to exchange data in such a way that the precise meaning of the data is readily accessible and the data itself can be translated by any system into a form that it understands."
Jeff Heflin and James Hendler (2000) Semantic Interoperability on the Webhttp://www.cs.umd.edu/projects/plus/SHOE/pubs/extreme2000.pdf
Semantic Interoperability requires distinction between vocabularies and perspectives
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
"… integrating resources that were developed using different vocabularies and different perspectives on the data. To achieve semantic interoperability, systems must be able to exchange data in such a way that the precise meaning of the data is readily accessible and the data itself can be translated by any system into a form that it understands."
Jeff Heflin and James Hendler (2000) Semantic Interoperability on the Webhttp://www.cs.umd.edu/projects/plus/SHOE/pubs/extreme2000.pdf
Semantic Interoperability requires distinction between vocabularies and perspectives
Terminologies / Ontologies:
Provide codes that denote entity types
Information models
Provide structure and context for data
Ontology Epistemology
SNOMED CT,ICD,…
openEHRHL7 RIMEN 13606,proprietary information models
"Heart failure, suspected"
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
Epistemic Intrusion is common in biomedical vocabularies
Bodenreider O, Smith B, Burgun A. The ontology-epistemology divide: A case study in medical terminology. Proceedings of the Third International Conference on Formal Ontology in Information Systems (FOIS 2004) : IOS Press; 2004. p. 185-195.
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
"… integrating resources that were developed using different vocabularies and different on the data. To achieve semantic interoperability, systems must be able to exchange data in such a way that the precise meaning of the data is readily accessible and the data itself can be translated by any system into a form that it understands."
Jeff Heflin and James Hendler (2000) Semantic Interoperability on the Webhttp://www.cs.umd.edu/projects/plus/SHOE/pubs/extreme2000.pdf
Epistemic Intrusion is common in biomedical vocabularies
Terminologies / Ontologies:
Provide codes that denote entity types
Ontology
"Heart failure, suspected"
Bodenreider O, Smith B, Burgun A. The ontology-epistemology divide: A case study in medical terminology. Proceedings of the Third International Conference on Formal Ontology in Information Systems (FOIS 2004) : IOS Press; 2004. p. 185-195.
SNOMED CT: 394887005:
Most epistemic-laden SNOMED CT concepts are in the hierarchy "Situation with explicit context (CM)"
• Body structure (body structure)• Clinical finding (finding)• Environment or geographical location (environment / location)• Event (event)• Linkage concept (linkage concept)• Observable entity (observable entity)• Organism (organism)• Pharmaceutical / biologic product (product)• Physical force (physical force)• Physical object (physical object)• Procedure (procedure)• Qualifier value (qualifier value)• Record artifact (record artifact)• Situation with explicit context (situation)• Social context (social concept)• Special concept (special concept)• Specimen (specimen)• Staging and scales (staging scale)• Substance (substance)
• ~ 3,000 precoordinated concepts
• patterns for postcoordination
• information model inside SNOMED CT
SNOMED CT:
~ 300,000 concepts
"CM"
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
CM concepts express absence, presence, history, uncertainty, …
alexia and agraphia present
COMPLEX CLINICAL SITUATIONS AND EPISTEMIC STATES
heart failure excluded
suspected neoplasm of brain
history of cat allergy
heart failure excluded
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
Hierarchy: "Heart failure excluded (situation)" isA "No cardiac failure (situation)" isA
"Heart disease excluded (situation)" isA "Disorder excluded (situation)" Full definition:
"Heart failure excluded (situation)" isA "No Cardiac Failure (situation)"
Group1Associated finding: Heart failure (disorder)Finding context: Known absent (qualifier value)Temporal context: Current or specified time (qualifier value)Subject relationship context: Subject of record (person)
OWL EL++: (according to transformation script) 'Heart failure excluded (situation) ' subclassOf RoleGroup some (('associated finding' some 'Heart failure') and ('finding context' some 'Known absent') and ('temporal context' some 'Current of specified') and ('subject relationship context' some 'Subject of record'))
Example: SNOMED CT CM concept "Heart failure excluded"
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
Hierarchy: "Heart failure excluded (situation)" isA "No cardiac failure (situation)" isA
"Heart disease excluded (situation)" isA "Disorder excluded (situation)" Full definition:
"Heart failure excluded (situation)" isA "No Cardiac Failure (situation)"
Group1Associated finding: Heart failure (disorder)Finding context: Known absent (qualifier value)Temporal context: Current or specified time (qualifier value)Subject relationship context: Subject of record (person)
OWL EL++: (according to transformation script) 'Heart failure excluded (situation) ' subclassOf RoleGroup some (('associated finding' some 'Heart failure') and ('finding context' some 'Known absent') and ('temporal context' some 'Current of specified') and ('subject relationship context' some 'Subject of record'))
Example: SNOMED CT CM concept "Heart failure excluded"
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
Ontological homogeneity of CM ? Do they represent clinical situations or information about
situations? What is the difference between X (finding) and X present
(situation)? Current frame-like CM ontologically inappropriate if
transformed to EL++ negations (absence of) value restrictions expression of uncertainty
Computational consequences if using OWL DL?
Problems of current interpretation of SNOMED CT CM concepts
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
Ontological homogeneity of CM ? Do they represent clinical situations or information about
situations? What is the difference between X (finding) and X present
(situation)? Current frame-like CM ontologically inappropriate if
transformed to EL++ negations (absence of) value restrictions expression of uncertainty
Computational consequences if using OWL DL?
Problems of current interpretation of SNOMED CT CM concepts
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
Ontological clarification of relevant toplevel categories Identification of modelling patterns in the CM Re-engineering of selected pattern using OWL DL and
considering basic principles of formal ontologies Peer-review and discussion of patterns, identification of
controversial design decisions Creating SNOMED CT modules using the re-engineered
portions Benchmarking classification time with description logics
reasoners
Methodology
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
Ontological clarification of relevant toplevel categories Identification of modelling patterns in the CM Re-engineering of selected pattern using OWL DL and
considering basic principles of formal ontologies Peer-review and discussion of patterns, identification of
controversial design decisions Creating SNOMED CT modules using the re-engineered
portions Benchmarking classification time with description logics
reasoners
Methodology
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
Reinterpretation of Clinical Situation as part of life when certain condition(s) present / absent
PATIENT
tBrain neoplasm situation
Alexia and agraphia situationCat allergysituation
Clinical finding present:
Suspected:
Past history of:
EHR
DOCTOR
Schulz S, Rector A, Rodrigues JM, Spackman K. Competing interpretations of disorder codes in SNOMED CT and ICD. AMIA Annu Symp Proc. 2012;2012:819-27.
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
PATIENT
tBrain neoplasm situation
Alexia and agraphia situationCat allergysituation
Clinical finding present:
Suspected:
Past history of:
EHR
DOCTOR
Schulz S, Rector A, Rodrigues JM, Spackman K. Competing interpretations of disorder codes in SNOMED CT and ICD. AMIA Annu Symp Proc. 2012;2012:819-27.
Particular
Information object Process Material object
Clinical Situation Biological Life
BioTopLite:
Ontology Patterns for reinterpretation SNOMED CT CM concepts
"P": Clinical Finding present (situation):'Alexia and agraphia present (situation)' equivalentTo 'Alexia (finding)' and 'Agraphia (finding)'
Catalina Martínez-Costa, Stefan Schulz: Ontology-based reinterpretation of the SNOMED CT context model
SNOMED CT context model (CM): important patterns of combing clinical terms with epistemic information
Current OWL-EL representation of the CM flawed Redesign patterns controversial regarding
interpretation of all SNOMED CT findings / disorders as clinical situations the linkage between information entities and clinical situation classes
Bipartition of current CM: parts of CM concepts re-interpreted as clinical situations, parts as information entities
"Friendly" computational behaviour of implemented OWL-DL redesign patterns
Future work: identification of all pattern in the current CM, consensus process regarding redesign within SemanticHealthNet, scripting for producing OWL models for complete CM
Identification of epistemic-laden content in other parts of SNOMED CT