On the implementation of pattern recognition methods in the diagnosis of non-communicable blood diseases

Abstract

The paper proposes hybrid approaches to the diagnosis of blood diseases using pattern recognition algorithms and combining expert methods, machine learning algorithms and formalized diagnostic rules. Within the framework of this study, an analysis of diagnostic signs of blood diseases is proposed and the concept of a hybrid intelligent model is developed. Since non-communicable blood diseases, including various forms of anemia, in hematology are characterized by the complexity of differential diagnosis, the variability of clinical manifestations and the need for a comprehensive assessment of the patient's condition, and the traditional analysis of laboratory data requires a significant amount of time and highly qualified specialists, this work is particularly relevant. The proposed approach makes it possible to formalize the process of analyzing medical data and to provide a preliminary classification of pathological conditions in hematology based on a set of diagnostic signs.

Speaker

Ivan Nikitovich Beglakov
Gagarin's State Technical University of Saratov
Russia

Discussion

Ask question