Prediction and Early Diagnosis of Bronchial Asthma in Workers Specializing in Argon Welding Based on Fuzzy Mathematical Models

   The purpose of research is to develop fuzzy mathematical models for predicting and early diagnosis of bronchial asthma developing under the influence of toxic substances released during argon welding providing an acceptable quality of decision-making for practical medicine   Methods. The paper uses the method of synthesis of fuzzy mathematical models of forecasting and early diagnosis of occupational diseases provoked by the action of industrial pesticides developed in the South-Eastern Federal University on the basis of the methodology of synthesis of hybrid fuzzy decision rules allowing to take into account the multiplicative nature of the links between the risk of occurrence and development of occupational diseases with the concentration and time of exposure to harmful chemicals.   Results. Using the chosen method, fuzzy hybrid models of forecasting and early diagnosis of industrial bronchial asthma in welding workers were obtained in which, in addition to the effects of ozone, dust and fibrogenic risk factors on the human body, unfavorable environmental and ergonomic factors combined with individual health characteristics are taken into account. In the course of modeling and expert evaluation, it was shown that the obtained mathematical model provides confidence in the correct prognosis of the occurrence of industrial bronchial asthma in welding workers at least 0.85, and early diagnosis at least 0.9.   Conclusion. Hybrid fuzzy models of forecasting and early diagnosis of industrial bronchial asthma in welding workers caused by the presence of ozone and fibrogenic dust in combination with individual, environmental and ergonomic risk factors were obtained, the quality of which was tested on representative control samples. During the control check, it was shown that the prognostic model provides diagnostic specificity and sensitivity at the level of 0.87, and the early diagnosis model - at the level of 0.92, which is a good result for the class of tasks under consideration.