Chmykhova O. Increasing the reliability of glucose tolerance testing systems under conditions of uncertainty in glycemic dynamics

The work is devoted to the improvement of information and measuring support of glucose tolerance testing of diabetic disorders while reducing the number of invasive glucose tolerance tests. The existing mathematical models of carbohydrate metabolism and dynamics of glycemia are analyzed and the possibility of using regression polynomials of limited order as such models is proved. To assess the informativeness of physiologically adequate parameters of glycemic dynamics, a parametric model of one-way analysis of variance, which allowed to calculate the amount of diagnostic information. The methods of obtaining redundant information on polynomial models on the parameters of localized spectral nonstationarity of changes in the level of glycemia were studied, which allowed to use additional information on polynomial models that reflect the dynamics of glycemia. A block diagram and computerized system for multifaceted identification of glycemic dynamics models based on significantly limited volumes of glucose tolerance tests have been developed, which has increased the reliability of diabetes diagnosis for a statistically reasonable minimum volume of biochemical measurements. An information computerized system has been developed for the recommendatory diagnosis of complications in diabetes mellitus and gastrointestinal disorders with the possibility of quantitative evaluation of diagnostic information for biochemical parameters.