Compression of ECG signal based on its deviation from a reference signal using discrete cosine transform

In this paper a discrete cosine transform (DCT) based ECG signal compression method is proposed. DCT is applied on the residual beats of ECG signal obtained after subtracting the period normalized and dc removed beats from a reference beat. The quantization and the threshold vectors are optimized to minimize the entropy of the quantized coefficients for a target distortion. The quantized coefficients are lossless encoded for further compression. Simulation result shows that the proposed method ensures higher compression ratio than the conventional methods published in the literature.