Сучасний стан наукових досліджень та технологій в промисловості (Jun 2021)
THE METHOD FOR PREDICTIVE ASSESSMENT OF THE CONDITION OF PATIENTS WITH ATOPIC DERMATITIS AT DIFFERENT STAGES OF THE DISEASE
Abstract
The subject of the research is the development of a method for prognostic assessment of the condition of patients with atopic dermatitis at different stages of the disease. The goal of the work is to study the informativeness of immunological indicators and data from dermatoscopic examinations in order to expand the possibilities of prognostic objectification of methods for assessing the condition of patients with atopic dermatitis with varying degrees of severity of the disease. The task of the work includes objectifying the blurring of assessment standards when analyzing the transition from one stage of the disease to another. Methods. The solution to this problem is possible when assessing the possibility of using models of parametric recognition (discrimination) using indicators of immunoglobulins in blood serum and indicators of immunograms, as well as color characteristics of skin areas based on the analysis of dermatoscopic images at various degrees of severity of the disease. Result. In the course of the study, the analysis of the color characteristics of the skin showed that when immunological blood parameters are added to the discrimination model, the probability of an error in making prognostic decisions significantly decreases. Predictive assessment of the condition of a patient with atopic dermatitis only by the color characteristics of the skin makes it possible to control this pathology with a higher degree of probability, which makes it possible to use the digital dermatoscopy method independently for express objectification of the condition of a patient with atopic dermatitis without waiting for the data of immunological analyzes. When a new patient appears, the above indicators are calculated for him and the normalized Euclidean distances to the center of the clusters corresponding to the studied pathologies are calculated. The calculated distances can be ranked and the probabilities of correspondence of the given case to specific pathologies can be determined. Conclusions. The prospect of further work is to substantiate the metrological characteristics of the method to eliminate possible systematic errors associated with the method of obtaining optical information.
Keywords