Ontological representation-oriented term normalization and standardization of the Research Domain Criteria.

Area of Science:

• Mental Health Research
• Computational Psychiatry
• Ontology Engineering

Background:

• The National Institute of Mental Health's Research Domain Criteria (RDoC) offers a novel dimensional framework for studying mental disorders.
• The RDoC matrix is central to this framework, but its elements, particularly Units of Analysis, present data normalization challenges.
• Ontologies provide strengths in semantic inferencing and automated data processing, making an ontological transformation of the RDoC desirable.

Purpose of the Study:

• To address limitations in the data normalization of Research Domain Criteria elements within the Units of Analysis.
• To propose and evaluate methods for improving the ontological representation of the RDoC framework.
• To establish a robust data foundation for future development of an RDoC ontology.

Main Methods:

• Developed solutions for data normalization focusing on the Units of Analysis within the RDoC.
• Leveraged standard terminologies, specifically the Unified Medical Language System Metathesaurus.
• Employed context-combining queries and integrated domain expertise from mental health professionals.

Main Results:

• Proposed methods demonstrated significant improvements in data normalization for RDoC elements.
• Evaluation showed a positive reception rate exceeding 80% among mental health professionals.
• Successfully created a foundational dataset suitable for ontological representation of the RDoC.

Conclusions:

• The proposed data normalization strategies effectively address limitations in the RDoC framework.
• The positive feedback indicates the practical utility and acceptance of the methods by the target audience.
• This work lays essential groundwork for the future development of a comprehensive RDoC ontology.