Application of logical data analysis in a software package for the diagnosis of gastritis

Abstract

A software package for computer diagnostics of gastritis is proposed based on available indicators and the experience of specialists. It is based on a logical approach of data analysis and is designed to determine the most likely types of diseases with a given accuracy. The input data includes patient examination indicators, their diagnoses and the experience of specialists in diagnosis accumulated in medical practice. Systems of multivalued predicate logic are used to represent data. The result of the work of the software package is to select the most appropriate types of diseases with a predetermined accuracy, based on the diagnostic data of patients. If the accuracy of the diagnosis cannot be achieved with the specified accuracy, it is proposed either to change the accuracy of the decision, or to undergo an additional examination.