Modeling and Segmentation of Ecg Signals through Chebyshev Polynomials

ECG (Electrocardiogram) signals originating from heart muscles, generate massive volume of digital data. They need to be compressed or approximated for efficient transmission and storage. ECG signal compression is traditionally performed in three ways: direct, transform and parameter extraction. Polynomial approximation which is a form of parameter extraction method, is employed here. This paper proposes a new technique of ECG modeling using Chebyshev interpolation. The technique is implemented on a standard ECG signal with a Chebyshev polynomial of order 100 with a maximum error less than 0.1. Segmentation of ECG signal further decreases the order of the Chebyshev polynomial and thereby the maximum absolute error.