A method for diagnosing borderline mental disorders based on hybrid fuzzy models

The purpose of the research is to synthesize fuzzy diagnostic models for transient neurotic disorders based on hybrid fuzzy models that improve the quality of decisions. Methods. Exploratory analysis has shown that the data structure describing the desired class of mental disorders is fuzzy, which makes it advisable to use fuzzy decision-making logic, and specifically the methodology for synthesizing hybrid fuzzy decision rules. The composition of informative indicators describing transient neurotic disorders in the composition of signs accepted in traditional medical practice, the level of adaptation of the body as a whole, the electrical imbalance of biologically active points associated with neuroses, confidence in the prognosis of the appearance of neuroses, quantitative characteristics of the functional reserve of target organs and systems is determined. For these groups of indicators, private diagnostic models have been obtained, the aggregation of which gives the final diagnostic model. Results. To assess the quality of admissions, three levels of quality control were used: at the expert level; according to model control samples and according to control samples in which the presence of transient neurotic disorders was checked using independent generally accepted research methods. The quality of the classification was checked by such indicators as diagnostic sensitivity, specificity and diagnostic effectiveness, which exceeded the value of 0,97. Conclusion. The paper provides fuzzy models for the diagnosis of transient neurotic disorders. An assessment of the quality of decisions made using expert assessment methods, mathematical modeling and statistical analysis showed that the hybrid fuzzy models obtained provide acceptable diagnostic quality with a confidence of at least 0,97.