Methods for ECG signal compression with reconstruction via cubic spline approximation

Abstract

In this article, two algorithms are described that are suited for real-time biomedical signal compression. These being, Amplitude Threshold compression and SQ segment compression. Comparison of these methods with well known methods such as lossy Discrete Cosine Transform (DCT) and lossless Turning Point (TP) is shown. The compression method outputs were reconstructed using a cubic spline approximation and compared. The values of compression ratio (CR), percent mean square difference (PRD) and area criteria were chosen for method comparison. Here it is shown that the method presented here (Threshold, SQ segment) provide considerably lower CR values than the DCT method and slightly higher CR values than the TP method. However, the PRD value for both proposed methods is lower than the PRD values of reference methods DCT and TP.