Resource Descriptive Chart

System Structure

FHIR-Ontop-OMOP offers a unique design that enables users to query an OMOP database using SPARQL, thereby enhancing interoperability between the FHIR RDF and OMOP CDM data models. At its core, OMOP CDM serves as the underlying database structure. A crucial layer of mapping is applied over this database to transform the OMOP CDM data into Turtle format, ensuring it aligns with FHIR RDF standards. This transformation is key to bridging the two standards. Furthermore, FHIR-Ontop-OMOP provides an Ontop endpoint, enabling users to execute SPARQL queries. This feature is vital in making complex data queries more accessible and efficient within the combined framework of these standards.

Diagram depicting the structured framework of the FHIR-Ontop-OMOP system, organized into three main modules. At the bottom, the 'Input Module' contains components such as FHIR Ontology, OMOP-FHIR Mapping, and the OMOP database, each symbolized by distinct icons. Above, the 'CKG Generation Module' shows the process flow beginning with the virtualization of the input data into a FHIR RDF knowledge graph using the 'Ontop' tool, and progressing to materialization and validation through FHIR ShEx, illustrated with appropriate symbols. The topmost 'Semantic Query Module' demonstrates the use of SPARQL queries to interact with the data, with arrows moving from a SPARQL icon to a query results grid, signifying the flow of data querying and result generation.

The resources included in the database are:

OMOP Table FHIR Resource
PERSON Patient
VISIT_OCCURENCE Encounter
CARE_SITE Location
CONDITION_OCCURENCE Condition(Problem)
DRUG_EXPOSURE MedicationStatement
LOCATION Location
MEASUREMENT Observation
PROCEDURE_OCCURENCE Procedure
PROVIDER Practitioner/PractitionerRole
CONCEPT CodeableConcept/Coding
CONCEPT_RELATIONSHIP ConceptMap
CONCEPT_ANCESTER ConceptMap

Resource Descriptive Chart

To better demonstrate the relationships existed among all the resources in our database, we have constructed a resource descriptive chart to provide visual explanation.

(hover your mouse over the image below to zoom in)

OMOP CDM Entity Relationship Diagram
(ERD of all tables designed in the common data model)
Source: https://ohdsi.github.io/CommonDataModel/cdm54erd.html

Entity-Relationship Diagram of the OMOP Common Data Model version 5.4. The diagram visually organizes healthcare data into multiple color-coded sections, representing different data categories such as clinical details, healthcare system interactions, economic factors, and metadata. Each section contains tables like 'Person', 'Visit Occurrence', 'Condition Occurrence', and 'Cost', detailing specific attributes and relationships. Lines connecting the tables show how data are related. A legend in the bottom right corner helps decode the color scheme used for different table categories.

FHIR RDF Resource Descriptive Chart
(resources currently available in our virtual database)

Diagram depicting the FHIR rdf resources available in the virtual database. The chart is organized into interconnected blocks representing various healthcare entities such as Patient, Observation, Procedure, Condition, Practitioner, and Location, among others. Each block details specific resource attributes. Lines between the blocks illustrate the relationships between these resources. For instance, the Observation block is linked to both Patient and Practitioner, indicating their roles in observation resource.

Data stored in RDF (Resource Description Framework) formats uniquely preserves information about itself while linking to other resources. More importantly, it imbues these linkages with semantic meanings, thereby facilitating higher-level information inferences. This capability of RDF is instrumental in building more explainable AI models, as it allows for a deeper understanding of data relationships and contexts, essential in making AI decisions transparent and interpretable.


Resources in the database are:
Resource Description
Patient Demographics and other administrative information about an individual or animal receiving care or other health-related services.
Observation Measurements and simple assertions made about a patient, device or other subject.
Procedure An action that is or was performed on or for a patient, practitioner, device, organization, or location.
Condition(Problem) A clinical condition, problem, diagnosis, or other event, situation, issue, or clinical concept that has risen to a level of concern.
Encounter An interaction between a patient and healthcare provider(s) for the purpose of providing healthcare service(s) or assessing the health status of a patient.
Practitioner A person who is directly or indirectly involved in the provisioning of healthcare or related services.
MedicationStatement A record of a medication that is being consumed by a patient. A MedicationStatement may indicate that the patient may be taking the medication now or has taken the medication in the past or will be taking the medication in the future.
Address An address expressed using postal conventions (as opposed to GPS or other location definition formats).
PractitionerRole A specific set of Roles/Locations/specialties/services that a practitioner may perform at an organization for a period of time.
CodeableConcept A CodeableConcept represents a value that is usually supplied by providing a reference to one or more terminologies or ontologies but may also be defined by the provision of text.
PractitionerRole A specific set of Roles/Locations/specialties/services that a practitioner may perform at an organization for a period of time.
ConceptMap A statement of relationships from one set of concepts to one or more other concepts - either concepts in code systems, or data element/data element concepts, or classes in class models.
Utilizing SPARQL to query resources in RDF graphs opens up a myriad of possibilities for meaningful data analysis in healthcare. With SPARQL, it's feasible to identify patient groups with specific diseases, uncover potential correlations between symptoms, and analyze demographic data to identify populations more vulnerable to certain diseases. These insights can significantly influence clinical diagnoses and treatment plans, demonstrating the power of SPARQL in extracting valuable information from complex data sets.