Titova A. Methods and models of information technology for an automated system for processing diagnostic information on the basis of thermal images

The purporse of the research is to increase the efficiency of analyzing the characteristics of thermograms by developing methods and models of information technology for automated systems for processing diagnostic information. For the first time, an analytical model for forming a thermal images was developed based on the transformation of the temperature indices of a contact digital thermograph into an image of the color model RGB, which allowed the use of image segmentation and filtration methods without pre-processing the generated thermal images. The method of determining the type and parameters of the membership function of the terms of linguistic variables in the formalization of the characteristics of thermal images of breast, which allowed, on the basis of a set of thermographs, to determine the functions of the membership of the terms of the linguistic variables. A fuzzy model was developed to support decision-making in setting a preliminary diagnosis during the analysis of thermograms, which allowed to reduce the time for conducting diagnostics by 8.97%.