Search academic texts

Information × Registration Number 0218U000361, 0117U000410 , R & D reports Title Mathematical and software to classify tissue surgical wound during surgery on the organs of the neck popup.stage_title Head Dyvak Mykola Petrovych, Доктор технічних наук Registration Date 26-01-2018 Organization Ternopol National Economic University popup.description2 The object of research is the signals received as a result of stimulation the tissues in a surgical wound on the neck organs by an alternating current with given frequency. The subject of the research is mathematical tools and software for classification the tissues in surgical wound during the operation on the neck organs. The purpose of research is creation the modern mathematical models and software for classification the tissues in surgical wound on the neck, which have guaranteed prognostic features and, on this basis, in combination with electrophysiological methods, make it possible to reduce the risk of damage the nerve tissues to a guaranteed level which is determined solely by the errors of electrophysiological methods. Research methods are the methods of probability theory, the multiple-set and interval approaches, as well as the theory of evolutionary algorithms were used for development the methods and algorithms for structure and parameter identification of the characteristics of signal as reaction to tissue stimulation in surgical wound in the form of interval discrete models; computer simulation, methods of stochastic programming and optimization methods were used to research criteria and optimality indicators of structures of the interval discrete models, convergence and time complexity of methods and algorithms of structure and parameter identification; methods of mathematical programming are applied for solving the interval systems of nonlinear algebraic equations. Research results and their scientific novelty: new conceptual solutions for reducing the risk of damage the recurrent laryngeal nerve (RLN) during surgery on the neck; the method of identification the interval models based on radial basis functions, which combines the procedures for forming centers of the receptor field of radial basis functions and analysis of interval data, which ensures simultaneous adjustment of the model structure and obtaining guaranteed estimates of its parameters; an interval model of characteristics of the environment of RLN monitoring based on radial basis functions, which, unlike the existing ones, provides an estimation of the distance from the stimulation point to the RLN; improved information technology for RLN identification, which, unlike existing ones, is based on mathematical models with radial basis functions and simultaneously reduces the risk of RLN damage and reduces the time of surgery on thyroid gland. Product Description popup.authors Дарморост Ірина Анатоліївна Кедрін Євген Станіславович Ковбасістий Андрій Вікторович Козак Олександра Леонідівна Манжула Володимир Іванович Маслияк Юрій Богданович Мельник Андрій Миколайович Падлецька Наталія Ігорівна Порплиця Наталія Петрівна Пукас Андрій Васильович Сусла Михайло Володимирович Тимець Володимир Іванович popup.nrat_date 2020-04-02 Close