A study on removing baseline drift of ECG signal in motion state based on morphological contour algorithm

Abstract

At present, wearable devices' motion state baseline drift brings some interference to electrocardiogram (ECG) signal analysis. In this paper, a kind of morphological contour algorithm, which is cascaded by the opened-close operation and the closed-open operation, is used to filter the original ECG signal in two stages. Among them, the first-stage filter uses a structural element with a time width of 0.11s to remove the QRS complex and the P wave from the ECG signal, and the second-stage filter adopts a structural element with a time width of 0.25s to remove the T wave of the ECG signal, and obtains the baseline of the raw ECG signal, and then the drifting noise is eliminated after subtracting the baseline from the original ECG signal. The algorithm was verified on the MIT/BIH database and the collected ECG signals in the moving state, and the results show that the algorithm can improve the accuracy of R wave detection of MIT/BIH ECG signals by 10.21% and improve the accuracy of R wave detection of actual collected moving ECG signals by 41.67%. The morphometric contour-based algorithm proposed in this paper can effectively remove the baseline drift of moving ECG signals on wearable devices and improve the accuracy of R-wave detection.