A method for the classification of QRS-complexes (Q1385901)

Computer-aided processing of ECG data, obtained with the help of the equipment, which registers ECG during a long interval of time (e.g., the Holter monitoring system) leads to the problem of classifying the heart ventricle complexes (QRS-complexes) according to their morphology. The simplest algorithms solve this problem with the help of the analysis of RR-intervals. The algorithms that use this approach are neither sensitive nor specific. They do not allow one to diagnose accurately the multifocal extrasystole and other kinds of complex cardiac arrhythmia. More elaborate algorithms use data about the form of a QRS-complex: its width, amplitude of waves, and other additional information. The serviceability of such algorithms depends on the accuracy of the determination of the parameters of the QRS-complex that are taken into account. This reduces their noise stability. Algorithms are known that are based on comparing every QRS-complex with a fixed set of ``template gages'' (patterns). The fact that the number of template gages is finite does not allow one to take into account the whole variety of possible forms of QRS. An increase of the number of template gages implies a decrease of the efficiency of the algorithm. A distinctive feature of the method described here is that one chooses template gages among the representatives of QRS-complexes that appear in the record under processing, which makes the algorithm adaptive to a concrete record.