A Novel SVD Noise Cancellation Algorithm for ICG Signal

The impedance cardiography (ICG) is a noninvasive technique, simple, easy, and cheaper for measuring systolic time intervals, it provides information about the cardiovascular disorder diagnosis and monitoring. The blood volume variation involved the thoracic electrical impedance changes where the ICG waveform range is between 0.8 and 20 Hz cycle may be altered by artefacts' cause distortions in the signal wave due to several causes thus the deterministic component estimation will be more and more complicated, in this context, the purpose of this study is noise removal without destroying the characteristics information carried on the signal. Our research presents a new noise-reduction technique of filtering tool to estimate a high-resolution spectrum has been proposed named singular value decomposition (SVD) based method for impedance cardiography (ICG) denoising, and compared it with LMS based adaptive filter. The SVD is already used as a tool to analyze a physiological signal as electrocardiogram (ECG) to improve the filtering accuracy. This study used specific performance settings for evaluation as SE, RMSE, SNR and SNR improvement are measured. The results show that the SVD technique is more performant.