Radiation Therapy Ontology: Report of AAPM Working Group
M Phillips1*, J Bona2, A Dekker3, P Gabriel4, C Mayo5, (1) University of Washington, Seattle, WA, (2) University Of Arkansas, ,,(3) Maastro Clinic, Maastricht, ,NL, (4) University of Pennsylvania, Philadelphia, PA, (5) University of Michigan, Ann Arbor, MI
Presentations
TU-IePD-TRACK 4-7 (Tuesday, 7/27/2021) 12:30 PM - 1:00 PM [Eastern Time (GMT-4)]
Purpose: To build an ontology that represents the domain of radiation oncology in a manner that makes it useful for medical and scientific progress.
Methods: A working group was established by the AAPM Big Data Subcommittee to organize and lead this effort. The working group has met on a biweekly basis since November 2018. The ontology’s design follows the guidelines of the OBO Foundry whose mission is “to develop a family of interoperable ontologies that are both logically well-formed and scientifically accurate”. To this end, a foundational ontology (Basic Formal Ontology) was chosen to provide the guiding principles and the basic classes and relations. Additional principles include: (a) to reuse whenever possible entities already described in other OBO ontologies to avoid duplication and confusion; (b) to embrace “realism” in the definition of entities and relations to encourage its use in clinical practice and research, and (c) to coordinate our efforts with other organizations working on standards-related projects. On a practical level, the github platform provides the infrastructure, development is accomplished using the Web Ontology Language (OWL), and the ontology software Protégé is used to develop and view the ontology.
Results: A domain ontology has been constructed. In line with the goal of coordinating with related efforts, mapping to the ASTRO minimum data elements has been accomplished. The work has also spawned the development of an applied ontology with more specific goals.
Conclusion: A robust fundamental ontology can be a key element in the development of data and concept-driven research in biomedicine. It provides a link to a number of practical taxonomies. Ontologies have proven to be useful in machine learning efforts in medicine. In the future, the ontology will as a basis for building applied ontologies and coordination with standards-forming organizations.
ePosters
Keywords
Not Applicable / None Entered.
Taxonomy
IM/TH- Informatics: Informatics in Therapy (general)
Contact Email