Improving the transform domain ECG denoising performance by applying interbeat and intra-beat decorrelating transforms

A method for suppression of electromyogram (EMG) interference in electrocardiogram (EGG) recordings is presented, It combines inter-beats and intra-beats decorrelation techniques with transform domain Wiener filtering. The interbeats relations are kept by a short window DCT over multichannel data representing the successive beats, and the Wiener filtering is performed in translation-invariant wavelet domain. The method results in improved signal-to-noise ratio of the processed signals and reduced ripple artifacts around the signal transients, thus preserving the true waveform in diagnostically important signal segments. The algorithm's structure allows an on-line parallel processing and can be built in the ECG registration systems.