Reason Functional Profile
This category supports reasoning and using an OWL DL representation of models, and research and clinical data. This addresses a requirement from the BRIDG stakeholder communities. (BRIDG) explicitly called for this capability. This also addresses the CDISC requirement for ontologies
Capability Elaborations
This Functional Profile includes, but is not limited to, the following capability elaborations:
Derived From Requirements
- Gap Analysis::Reason::117.1 - Represent caBIG Information models as ontologies Represent caBIG information models as ontologies in order to be able to use Semantic Web technology to reason over information models and the data that conforms those models
- Gap Analysis::Reason::125 - Reasoning and querying on Ontology models Support reasoning and querying on Ontology models (T-Box) and instance data (A-Box)
- Gap Analysis::BRIDG::BRIDG-1 - Store data elements in a publicly available ontology repository This is representational model of the BRIDG Level 2 and Level 3 content that can be used to reason over BRIDG-derived artifacts. The new model will become a “canonical quality control” by which new content – when presented for inclusion in the evolving 3.x model – will be able to be “analyzed” for semantic consistency “on its way in” so that semantic inconsistencies can be addressed. The BRIDG model is a complex ontology that incorporates all of the HL 7 artifacts into the version 3 side of the model and links these to the UML side of the model via mappings at the class and attribute level.
- Gap Analysis::caEHR::caEHR 3 - Support reasoning on clinical outcome measurements The caEHR provides a means for performing outcome measurements, and aggregating those measurements for analysis. These outcome measurements are defined as being time interval values based on clinical observations. Representation of clinical outcome measurement details is required. This should include the set of observations that a clinical outcome requires, and any constraints that might be necessary. These constraints may need to be represented as rules or value set bindings. This would more likely be a specific profile for an observation class or set of classes with criterion classes associated. These would also be associated with the detailed clinical models that represent the observation.code attribute in added detail. In addition, aggregations occur across measurements. Aggregations may have measurements for single patients or informal collections of patients. In addition it may operate upon formal collections of patients in the form of study cohorts. These measurements may be defined in the form of a treatment plan. The purpose of the KR data for outcomes would allow reasoning to determine related measurements, related cohorts, possible cohort definition criteria, and treatment plan details.
- Gap Analysis::CDISC::CDISC-10 - Use Ontologies to Help Create High Quality Data Element Definitions Use ontologies to help create high-quality data element definitions and avoid duplicate definitions. In particular, Layer 2 of BRIDG 3.0 has an OWL-based ontology of the BRIDG DAM semantics that can be used to help support this normalization and harmonization process. *Source * * CDISC SHARE: Pathway into the Future for Standards Development and Delivery, April 10, 2010, Brow W. Kisler, CDISC Senior Director and CDISC SHARE Pilot Report.
ontologyModelReasoning capability elaboration
Support reasoning and querying on Ontology models (T-Box) and instance data (A-Box)
Represent caBIG information models as ontologies in order to be able to use Semantic Web technology to reason over information models and the data that conforms those models
Store data elements in a publicly available ontology repository
Use Ontologies to Help Create High Quality Data Element Definitions.
reasonOverClinicalOutcomeMeasurements capability elaboration
The caEHR provides a means for performing outcome measurements, and aggregating those measurements for analysis. These outcome measurements are defined as being time interval values based on clinical observations.
Representation of clinical outcome measurement details is required. This should include the set of observations that a clinical outcome requires, and any constraints that might be necessary. These constraints may need to be represented as rules or value set bindings.
This would more likely be a specific profile for an observation class or set of classes with criterion classes associated. These would also be associated with the detailed clinical models that represent the observation.code attribute in added detail.
In addition, aggregations occur across measurements. Aggregations may have measurements for single patients or informal collections of patients. In addition it may operate upon formal collections of patients in the form of study cohorts. These measurements may be defined in the form of a treatment plan. The purpose of the KR data for outcomes would allow reasoning to determine related measurements, related cohorts, possible cohort definition criteria, and treatment plan details.
