A novel approach for real-time ECG signal denoising using Fourier decomposition method

Object: The heart is a vital organ in human anatomy, and proper care is essential. Electrocardiogram (ECG) is a non-invasive and popular tool to monitor cardiac health. ECG signal provides enormous information about the heart. ECG signals get contaminated with various kinds of noises. The presence of noises in the ECG signal may deceive cardiac health. A healthy heart may be diagnosed as unhealthy or suffering from arrhythmia. To get the appropriate information on cardiac health, noise removal from the ECG signal is essential. Method: This paper proposes a method based on the Fourier decomposition method (FDM) to suppress the commonly occurring noises such as power line interference, white Gaussian noise, baseline wander, muscle contraction, motion artifacts, and electromyographic noise in an ECG signal. The proposed method utilizes a discrete cosine transform (DCT) to process the ECG signal. The DCT coefficients corresponding to various noises are recognized and suppressed by FDM-based zero-phase filters. The proposed method is validated on the MIT-BIH arrhythmia database, noise stress test database, and real-time ECG data. Result: The performance of the method is evaluated in terms of the output signal-to-noise ratio, mean squared error, and percent root means square difference. The proposed method performs very well at a low input signal-to-noise ratio. With muscle contraction, motion artifact, and electromyographic noise, the proposed method provides an output signal-to-noise ratio of 27.42 dB, 29.39 dB, and 27.94 dB at a −10 dB input signal-to-noise ratio, respectively. Conclusion: Experimental results show that the proposed denoising method is superior to the existing techniques at a different level of the input signal-to-noise ratio. Furthermore, the fast Fourier transform to implement the proposed method offers low computational complexity and makes it suitable for real-time ECG signal processing. © 2022, The Author(s), under exclusive licence to The Brazilian Society of Biomedical Engineering.