A Semantic-Web Representation of Clinical Element Models Cui Tao, PhD Mayo Clinic
Acknowledgement Christopher G. Chute, MD, DrPH Robert R. Freimuth, PhD Stanley M. Huff, MD Guoqian Jiang, PhD Thomas A. Oniki, PhD Jyotishman Pathak, PhD Craig Parker, MD Deepak Sharma Feichen Shen Sridhar D. Yadav Qian Zhu, PhD
EHR SHARPn drools Normalized EHR XML Database Schema CEM XML Schema Database OWL Triple Store SWRL drools
Introduction Clinical Element Model (CEM): Logical models ensure semantic interoperability for: Data representation Data interpretation Data exchange within and across heterogeneous sources and applications Represented in CEML/CDL (Constraint Definition Language) Define syntax and grammar Not semantics
Introduction Semantic Web Explicit and formal semantic knowledge representation Web Ontology Language (OWL) /Resource Description Framework (RDF): Define relationships Define classes Define constraints Consistency checking Link to other domain terminologies Harmonize with other clinical data modeling languages Semantic reasoning
Detailed Clinical Element Model Ontology Layer1:meta-ontology Meta-CEM.OWL Layer2: detailed clinical element models Detailed Clinical Element Model Ontology Layer3:patient instances RDF Triple Store Joe Doe’s record Jean Doe’s record Joe Doe’s record
Semantic Reasoning Consistency checking Classification and reasoning Cardinality constraints Data types Property allowed domains and ranges Permissible values in value sets Classification and reasoning
Consistency Checking
Consistency Checking
Consistency Checking
Automatic Classification Semantic Definition of Normal DBP Data Two DBP Measurements
Automatic Classification Semantic Definition of Normal BP Infer that both exam1 and exam2 are within the normal range
Semantic Reasoning Example Has two outpatient (if possible) measurements of SBP >140 or DBP > 90 at least one month after taking anti- hypertensive drugs. Ontology: “Labetalol 100mg” (SCT318445000) is an anti-hypertensive drug (SCT1182007). CNTRO temporal relation reasoner: 5 weeks > one month Duration (May 1 and March 1) > one month
Implementation Status Meta Ontology Basic category classes Properties Cardinality constraints Two OWL experts and two CEM experts have evaluated the meta-ontology to ensure it can faithfully cover the original contents Automatic Convertor: detailed CEM ontologies
CEM-OWL convertor Meta-Ontology Shared CEM Ontology import Shared CEM Ontology import CEM Core Model Ontologies import CEM Secondary Use Ontologies
Download the ontologies Shared Models: http://informatics.mayo.edu/sharp/CEM2OWL/CEM-Shared.owl CORE Models: http://informatics.mayo.edu/sharp/CEM2OWL/COREModels/COREStandardLabObs.owl http://informatics.mayo.edu/sharp/CEM2OWL/COREModels/COREStandardLabObsCoded.owl http://informatics.mayo.edu/sharp/CEM2OWL/COREModels/COREStandardLabObsIntervalQuantitative.owl http://informatics.mayo.edu/sharp/CEM2OWL/COREModels/COREStandardLabObsNarrative.owl http://informatics.mayo.edu/sharp/CEM2OWL/COREModels/COREStandardLabObsOrdinal.owl http://informatics.mayo.edu/sharp/CEM2OWL/COREModels/COREStandardLabObsQuantitative.owl http://informatics.mayo.edu/sharp/CEM2OWL/COREModels/COREPatient.owl http://informatics.mayo.edu/sharp/CEM2OWL/COREModels/CORENotedDrug.owl http://informatics.mayo.edu/sharp/CEM2OWL/COREModels/COREDiseaseDisorder.owl Secondary Use: http://informatics.mayo.edu/sharp/CEM2OWL/SecondaryUse/SecondaryUseDiseaseDisorder.owl http://informatics.mayo.edu/sharp/CEM2OWL/SecondaryUse/SecondaryUseNotedDrug.owl http://informatics.mayo.edu/sharp/CEM2OWL/SecondaryUse/SecondaryUsePatient.owl http://informatics.mayo.edu/sharp/CEM2OWL/SecondaryUse/SecondaryUseStandardLabObsCoded.owl http://informatics.mayo.edu/sharp/CEM2OWL/SecondaryUse/SecondaryUseStandardLabObsIntervalQuantitative.owl http://informatics.mayo.edu/sharp/CEM2OWL/SecondaryUse/SecondaryUseStandardLabObsNarrative.owl http://informatics.mayo.edu/sharp/CEM2OWL/SecondaryUse/SecondaryUseStandardLabObsOrdinal.owl http://informatics.mayo.edu/sharp/CEM2OWL/SecondaryUse/SecondaryUseStandardLabObsQuantitative.owl http://informatics.mayo.edu/sharp/CEM2OWL/SecondaryUse/SecondaryUseStandardLabObsTiter.owl
Conclusion and Future Directions Meta-Ontology: semantically defined the basic classes, properties, their relationships, and constraints Convertor: CDLOWL Represent SHARPn normalized data using RDF Investigate SWRL/Drools combination for phenotyping