基于误差可控的转换变换心电信号去噪方法OA北大核心CSTPCD

In view of the amplitude and shape deformation caused by uncontrollable error in removing noise from electrocardiogram(ECG)signals,an ECG signal denoising method based on shift transformation with controllable error is presented.The low-frequency and high-frequency components of the ECG signals can be obtained by the shift transformation.By setting different error bounds on the characteristic waveform and high-frequency noise,the high-frequency noise can be filtered out effectively and the reconstruction error at the characteristic waveform can be kept within a small range.Smoother baseline drift can be obtained by implementing cubic interpolation for low-frequency components.On the basis of the above,the morphological changes caused by denoising can be improved.Experiments on the ECG of the database MIT-BIH show that the proposed method can remove baseline drift and high-frequency noise effectively,while maintaining the amplitude and shape of the characteristic waveform well.In comparison with the contrast methods,the signal-to-noise ratio(SNR),the root mean square error(RMSE)and the normalized correlation coefficient(NCC)of the ECG signals after denoising reach satisfactory values.