Intelligent decision making system of department of medical institutions based on neural network, production and statistical models

The article describes the development of information-analytical system of decision-making in the treatment of abdominal organs of patients. The structure of the system provides for the formation of a preliminary diagnosis of the patient’s condition based on a neural network and statistical analysis of the electronic medical record. For the operating control of the patient’s condition during the operation or getting a quick consultation in the case of a critical situation, the system provides an expert assessment of the circumstances with the possibility of a surgeon’s speech dialogue with the intellectual system. Purpose. Increase the intelligence of decision-making in the department of a medical in-stitution based on neural network, production and statistical models.Materials and methods. Neural networks and the statistical approach for analyzing and processing a large amount of medical data, as well as computer modeling of the practical problem, using the Java programming language, were used to obtain scientific results.Results. The developed prognosis program is a hybrid dynamic expert system, the use of which will improve the efficiency of processes for assessing the severity of the underlying dis-ease, taking into account pathology; predicting the risk of intraoperative complications in the planning mode and in real time; recommendations of surgical tactics with combined surgery; predicting the risk of postoperative complications; determine the volume of intensive care in the postoperative period.Conclusion. The structure of creating a fuzzy model of predicting operational risk for performing simultaneous interventions depending on the patient’s condition based on production rules is considered, the base of which can be corrected in the training regime of the expert system