Application of the statistical approach in diagnosing in medical and biological researches

Abstract

The task of diagnosis in biomedical research in a number of cases can be solved using a statistical approach. Current research is the possibility of using statistical analysis to diagnose the state of the human respiratory system based on the values of the percentage contributions of particles of different sizes contained in the exhaled air. The aim of the research is to identify certain regularities in the values of the diagnostic signs of the moisture condensation of the exhaled air, which will make it possible to consider the groups under investigation as disjoint classes. Three groups of individuals were examined: healthy people and patients with bronchitis and pneumonia. For each group, the identification of the particles that are the primary diagnostic data using the laser correlation spectroscopy method and the further processing of the data using the discriminant analysis method are performed. Selection of variables discriminating the study groups in the best possible manner is done; the model of variables and classification functions is constructed. There are presented the results of the main steps of the analysis – the set of variables included in the model and the coefficients of the classification functions for the three groups – which formed the basis for the algorithm for the work of the developed software product.