The structure of the CDSS information repository based on the ontological approach

The construction of an information repository for a clinical decision support system (CDSS) in the weakly formalized subject area of treating bronchopulmonary diseases is considered. An overview of approaches to creating knowledge bases in this subject area is provided. A method for extracting knowledge is described, which is based on rules from clinical recommendations and the search for dependencies between words in sentences, taking into account the sequence of rule application. The information repository of the CDSS is populated with ontological and production knowledge bases using the proposed knowledge extraction method. An ontology for the selected subject area was developed, and studies of its quality were conducted through an analysis of the graph topology using cognitive ergonomics metrics. The effectiveness of the described knowledge extraction method is demonstrated. An original architecture for a clinical decision support system has been developed.