A New Adaptive Wavelet-Based ECG Compression Technique

In this paper, an adaptive compression approach targeting ECG signals is proposed. The proposed approach utilizes the published literature in designing an efficient compression technique that provides higher compression ratios subjected to high quality reconstructed signals. It is based on integrating a simple thresholding mechanism for the significant wavelet coefficients and an easy to implement coding strategy for their locations. The approach is tested using five different databases. The results show that the considered approach is able to provide compression ratios up to 20 with Signal to Noise Ratio (SNR) above 18 dB. Performance evaluation of the considered approach is conducted by several objective and subjective criteria. Furthermore, several issues regarding implementation, feasibility, and validation are also considered.