Electrocardiogram Signal Compression and Decompression

Abstract

Electrocardiogram is a method of recording the heartbeat of a patient electronically. Quicker transmission of electrocardiogram signals through channels like across wireless networks, telephone networks, hospital networks is not possible without compressing them. Here I implement an algorithm for ECG (electrocardiogram) signal compression based on Delta Coding and LZW (Lempel–Ziv–Welch) encoding. To carry out compression, real time ECG data has been recorded through EKG sensors and Labview. Since the recorded signal was analogue it was first converted to digital by applying Savitzky Golay Filter, uniform sampler and quantizer sequentially. Then the signal was compressed using delta coding and LZW coding. Then the compressed signal was decompressed and then further converted to analogue. Finally the reconstructed signal was compared with the original signal. This was done for six different analogue samples. The compression ratios and the reconstruction errors for all were calculated. The calculated compression ratios range from 1:10.51 to 1:12.56 and PRDs range from 3.33% to 4.79%. Mean compression ratio came out to be 1:11.36 and mean PRD came out to be 4.18%. The number of quantization levels was kept sufficiently large in order to reduce the quantization error zero levels. In the pre-processing stage the signal was reduced to about one-ninth of the original size while in the encoding stage it was reduced to one third. When applied sequentially, the compression ratios got multiplied.