Abstract

The purpose of research is to develop a method for synthesizing models for predicting and diagnosing occupational diseases based on hybrid fuzzy technology, providing an increase in the quality of decision-making in occupational pathology.Methods. It is established that most of the problems related to the topic under study (forecasting, early diagnosis, assessment of the severity and dynamics of the development of professional diseases) belong to the class of poorly formalized problems with fuzzy and incomplete data structure, which are recommended to be solved using the methodology of synthesis of hybrid fuzzy solving rules based on on the interaction of the natural intelligence of doctors and a cognitive engineer with artificial hybrid intelligence. Using the chosen mathematical apparatus, a method for the synthesis of fuzzy models of forecasting and diagnosis of occupational diseases is proposed.Results. As a concrete example, the problem of predicting and diagnosing ischemic heart disease (CHD) in electric train drivers has been solved with the allocation of such classes of conditions as: "healthy and the appearance of CHD is not expected"; "healthy, but the appearance of CHD is expected after the predicted time"; "early stage of CHD"; "CHD disease has been detected". As a result of expert evaluation, it was shown that the confidence in the correct classification is at the level of 0.9. The same result was confirmed by the results of statistical tests on representative control samples in terms of diagnostic sensitivity and specificity.Conclusion. The proposed method of synthesis of hybrid fuzzy models makes it possible to synthesize hybrid decision rules that improve the quality of prediction and early diagnosis of the studied class of diseases both in the presence of training samples and in their absence by compensating for the lack of statistical material by methods of formalization of clinical thinking. As a concrete example, the problem of predicting and diagnosing ischemic heart disease in electric train drivers has been solved. It is shown that the confidence in the correct classification is at the level of 0.9, which allows us to recommend the results obtained for practical use in occupational pathology.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call