Combined coding and adaptive thresholding algorithms for ECG compression

This paper proposes a new wavelet-based ECG compression technique. An adaptive thresholding, based on the required signal-to-noise-ratio (SNR) and the concept of energy packing efficiency, is used to classify the wavelet coefficients into significant and insignificant ones. Both coefficients are encoded using an efficient coding algorithm. The proposed technique is tested using several records taken from the MIT-BIH Arrhythmia database. The results show that high compression ratios with SNR´s better than 18dB can be achieved.