Burtsev M. Decision Support System in the diagnosis of therapy disease based on the combined decision rule

This work is dedicated to solving important scientific and technical tasks - synthesis of diagnostic decision which purpose is to improve the accuracy and reliability of diagnosing therapeutic diseases, and development on its basis biomedical decision support system. The developed decision rule is intended to combine objective (statistical) and subjective (symptom complex structure description) approaches for the diagnostic conclusion formulation. This paper formalizes the problem and describes the methods of decision rule synthesis, the construction of the diagnoses hierarchical structure, the construct of diagnostic meaningful intervals of numerical signs based on the theoretical distribution approximations with polynomials of the 2nd order, choice of membership functions for symptom complexes description, an approach to building, maintaining, and modification of decision support system knowledge. Developed structure, information and algorithmic software of decision support system was tested on real medical data.