An approach towards automatic generation of evidence-based decision support systems for clinical diagnosis based on an extensive clinical guideline schema

Machine-executable clinical guidelines are of great importance to the implementation of clinical decision support system (CDSS). Any change to them in traditional CDSSs would lead to redesign and redevelopment of systems, which makes the systems hard to maintain and not adaptive to different environments. We propose, in this paper, an approach towards automatic generation of evidence-based decision support systems for clinical diagnosis based on an extensive clinical guideline schema. Flowchart is used for the representation of best clinical guideline in the graphical knowledge definition tool. Through mapping rules, flowchart can be changed into structural knowledge that stores in knowledge repository. Knowledge interpretation engine is designed to parse the knowledge and provide necessary information for web-based assisted diagnosis. The mutual transformation of flowcharts and structural knowledge based on mapping rules makes the knowledge repository adaptive to different situations and easy to be maintained. The cooperation of mapping rules and the engine makes the proposed system as automatic as possible when system captures new clinical knowledge.