Удосконалена інформаційна технологія для моніторингу зворотного гортанного нерва

Abstract

The information technology has been proposed and substantiated to detect the location of the laryngeal nerve in the surgical wound during operations on the thyroid gland. The information technology, in contrast to existing, reduces the risk of the laryngeal nerve damage and the time of operation on the thyroid gland due to the developed interval model with radial basis functions. The mathematical models, which basis functions are radial, have a simple generalized structure and their structural identification is reduced to calculating the number of centers of radial basis functions. The main characteristics of the surgical wound tissue during surgery on the thyroid gland are its type and the distance from the point of irritation of the surgical area to the laryngeal nerve. As a result of research and processing of information signals of different patients, indicators of the type of the surgical wound tissue are substantiated. Such characteristics are the spectral components with the maximum amplitude. The patient information signals resulting from the response to muscle and nerve irritation at different distances to the laryngeal nerve have been analyzed. The information signals of patients, which respond to the irritation of the surgical wound with alternating current at equal distances to the laryngeal nerve of different patients, are similar, but the characteristics of the signals are different. The interval data analysis methods to eliminate the uncertainty and heterogeneity of the data sample have been proposed A clustering algorithm of inhomogeneous data samples has been used, which is based on the integration of similar spectral characteristics of information signals into groups described at intervals. The true distance from the irritation point of the surgical area to the back laryngeal nerve is within the constructed distance interval. This approach makes it possible to construct a single mathematical model based on radial basis functions to predict the distance from the irritation point to the laryngeal nerve. The results of the experiments confirmed the efficiency of advanced information technology for the laryngeal nerve monitoring