The dissertation is dedicated to addressing the current scientific and practical task of improving the methods of analyzing biomedical signals with locally concentrated features in cardiological decision support systems.
The purpose of the dissertation work is to increase the efficiency of electrocardiological patient studies and reduce the risks of making incorrect diagnostic decisions by improving existing methods of morphological analysis of biomedical signals with locally concentrated features and designing cardiological decision support systems based on these methods.
The object of the study is the process of forming diagnostic making under conditions of significant a priori uncertainty based on functional diagnostics of the state of the cardiovascular system.
The subject of the study is the methods and tools of developing biomedical decision support systems for conducting an еlectrocardiographic study of the cardiovascular system.
Based on the research results, the following scientific novelty have been obtained:
– methods of system analysis have been further developed, which consist in building the generalized model of the electrocardiological study process in the form of the set of functional, informational, and structural models, which made it possible to identify the main sources and flows of information, critical stages of processing diagnostic information in order to increase the efficiency of electrocardiological study;
– for the first time, a structural model of an electrocardiological study has been developed in the form of a probabilistic-time graph, with the help of which analytical expressions were obtained that describe the process of developing diagnostic decisions as a result of an electrocardiological study under the given initial conditions, and criteria for the effectiveness of an electrocardiological study were determined, which made it possible to perform analysis and optimization as the whole process of an electrocardiological study, as well as its individual stages;
– for the first time, a method was developed for automatically determining the heart’s electrical axis, which is based on the calculation of the integral signal from six standard leads from the limbs and cluster analysis, which made it possible to increase the probability of correctly calculating the position of the heart’s electrical axis without the need to conduct a morphological analysis of the electrocardiogram.
The practical significance of the obtained results lies in the fact that the models of the electrocardiological study process and the method for automatic determination of the heart’s electrical axis developed in the dissertation work serve as the scientific and methodological basis for the development of corresponding algorithms and software. The practical value of the obtained results includes the following:
– software has been developed for finding analytical expressions of the transfer function, probability, and average time of conducting an electrocardiological study based on the proposed structural model;
– a structural diagram of hardware and software complex for conducting electrocardiological research has been developed;
– software for determining the heart’s electrical axis has been developed.
The results of the dissertation work were implemented in the form of models, algorithms, and software modules for solving decision support problems in biomedical systems, which is confirmed by the act of implementation at the Department of Computer Engineering and Programming of the National Technical University "Kharkiv Polytechnic Institute" in research and educational work the process of studying the disciplines “Signal and Image Processing”, “Design of Computer Diagnostic Systems”, “Fundamentals of Scientific Research”, as well as coursework and theses for bachelors and masters in the specialty “Computer Engineering” (Kharkov).