On optimization of the number of diagnostic signs for breast diseases through thermometric data

The work is devoted, on the one hand, to the study of the correlation interconnection of signs, derived from medical thermometer data intended for the diagnosis of breast diseases. The authors also highlighted the most important signs of a disease for the express-diagnostics of cancer breast tumors.
One of the directions of development of artificial intelligence systems is the development of expert systems for medical diagnostics. Their use helps the doctor to improve the quality of their work. The objective of such systems is not only the definition but also consulting assistance in identifying diseases (one or more) of patients, based on his observations. When creating intelligent advisory systems, developers rely on a number of high-quality diagnostic features. For their mathematical interpretation, a variety of functional relationships between the initial thermometric data can be used. In this regard, the number of possible diagnostic quantitative traits that may be employed in diagnostic systems express increases to several hundred or even thousands. Each of the resulting symptoms was assessed in terms of its information content, which reflects the degree of diagnostic ability of this feature. However, many of the quantified traits are interdependent and interchangeable. Reducing the number of diagnostic features with regard to their independence and maximum information content is one of the ways of increasing the quality of advisory intelligent systems.