Optimal thresholds for two-state ECG compressing algorithm

ECG (Electrocardiogram) data compression is one of the important requirements of the remote system of monitoring and diagnosis of cardiovascular diseases based on ECG signal. Recently, our team has proposed a two-state ECG compressing algorithm which not only has a high compression ratio and a low distortion, but also is quite simple and flexible for a real-time system. In this algorithm, the selection of the threshold amplitude value for the classification of two compression states is a key factor. This paper proposed an experimental approach to provide optimal threshold values for the two-state algorithm based on 10 MIT-BIH records and 10 CU records. After the process of experimenting with different threshold values, the threshold value generating the most suitable PRD (RMS Percentage difference) and CR (compression ratio) is Thr1 = 10 and Thr2= 0.3 * Thr1=3.