Application of Adaptive Orthogonal Transform for Diagnosing Physiological Signals

摘要

In this paper we propose a method of extracting informative characteristics when analyzing physiological signals and other rhythmic processes with no-stationary parameters. The Principe of the method consists in decomposing the signals in a system of parametric basic functions obtained during the synthesis of an orthogonal operator adaptive to the standard of the class of the signals to be analyzed. This makes it possible to obtain vectors of informative features of minimum size which allows the rate of certainty of diagnosis.