Substantiation of the Result of the Work of the Cyber-Physical System for Breast Cancer Diagnostics

Abstract

The chapter is devoted to the method of data analysis and processing in cyber-physical systems for the diagnosis of patients in medicine. In particular, the approach to processing the results of the thermometric data analysis system is discussed. In this work, the influence of cancer signs on the diagnosis made after the work of the classifier is studied. The influence of each sign on both the correctness and the error of the diagnosis is determined. The initial set of highly informative diagnostic thermometric signs used in classification algorithms was previously obtained by A. G. Losev and V. V. Levshinsky. In this chapter, the next stage of building a modern medical cyber-physical system is considered. This stage is the development of a mechanism for justifying the obtained diagnostic solution in a language that is understandable to the doctor. A factor model is constructed for compiling a hierarchy of signs used to justify the classification result. Further directions of research based on the obtained knowledge are proposed, in particular, the use of the fuzzy inference mechanism to decide the strengthening or weakening of signs in the model.