Analysis of ECG by multipulse excited linear prediction model

Abstract

We propose a new tiae do.ain aethad for efficient representation of the ECG and delineation of its COIaponent waves. The aethod is based on the .ultipulse Linear prediction (LP) coding which is being widely used in speech processing. The excitation to the LP synthesis filter consists of a few pulses defined by their locations and _plitudes. Based on the _plitudes and their distribution. the pulses are suitably co.bined to delineate the COIapo nent waves. Beatto beat correlation in the ECG signal is used in QRS periodicity prediction. The aethod entails a data COIapression of I in 6. The .ethod reconstructs the signal with an NMSE of less than 5%.