Customized clinical domain ontology extraction for knowledge authoring tool

Clinical Decision Support Systems (CDSS) require a shareable and adaptable knowledge base. However, sharing and reusing the expert's knowledge is a challenging task. The proposed approach designs a web based application that acquires and adapts the clinical expert's knowledge into shareable knowledge base. The system, Intelligent Knowledge Authoring Tool (I-KAT) creates rules in the form of Medical Logic Module (MLM) using HL7 standard Arden Syntax. These rules are easily shareable with HL7 complaint clinical institutions and organizations. To achieve interoperability using MLM, the system uses a mechanism for integration of terminology standard (SNOMED CT) concepts with CDSS standard (Virtual Medical Record (vMR)). The SNOMED CT ontology is comprehensive; containing more than 0.3 million concepts but 10--15% concepts of total ontology is normally used in rule creation for a specific domain. Semantically defining relationships between SNOMED CT concepts and vMR concepts require domain ontology development from the SNOMED ontology. In this paper we focus on automatic extraction of domain ontology from overall SNOMED CT ontology on the basis of vMR schema concepts and their attributes mapping with corresponding SNOMED CT concepts. The extracted domain ontology will increase the efficiency and effectiveness of searching mechanism in contextual selection process.