[Statistical synthesis of an algorithm for the automatic processing of cardiosignals].

Abstract

A mathematical model of cardiosignal (CS) based on the standard cardiosignals (SCS) is discussed. Algorithm for computing probability of arriving cardiosignal identity with one of the SCS classified by their clinical patterns is synthesized. Computations are based on the evaluation of similarity measure between ACS and SCS which is built as a likelihood logarithm. Mathematical-and-heuristical program development is executed and algorithm is realized with minimum computations and "size" of memorized ACS counts. By digital simulation algorithm serviceability is determined. The algorithm performs the following functions: scaling ACS with duration, normalizing, phase referencing and computing similarity measure according to the presented SCS sample.