Md. Maniruzzaman - Academia.edu (original) (raw)

Uploads

Papers by Md. Maniruzzaman

Research paper thumbnail of REMOVING BASELINE WANDER FROM ECG SIGNAL USING WAVELET TRANSFORM

REMOVING BASELINE WANDER FROM ECG SIGNAL USING WAVELET TRANSFORM, Sep 4, 2019

Electrocardiogram (ECG) signal is the representation of electrical activity generated by heart mu... more Electrocardiogram (ECG) signal is the representation of electrical activity generated by heart muscles, which is primarily utilized to detect cardiac abnormalities. Due to the sensitive nature of ECG, its important features are affected by different noises and create problems for diagnosis. This study proposes biorthogonal wavelet family by investigating different wavelet families to reduce baseline wander from the ECG signal. The proposed approach performance is compared to adaptive normalized least-mean-square (NLMS) and notch filters. Different performance parameters, such as amplitude spectrum, magnitude squared coherence (MSC), and power spectral density (PSD) has been evaluated. Signal-to-noise ratio (SNR), percentage root-mean-square difference (PRD), mean-square-error (MSE), normalized mean-square-error (NMSE), root mean-square-error (RMSE), and normalized root mean-square-error (NRMSE) performance parameters are calculated as well. The SNR values of the reconstructed ECG signal are-0.0046 dB and 1.6122 dB for notch and adaptive NLMS filters, respectively, which are lower than that of 8.0464 dB for the biorthogonal wavelet transform. Similarly, the MSC values are 0.091903 and 0.44522 after notch and adaptive NLMS filtrations, respectively, which are lower than those of 0.8913 after wavelet filtration. Also, the PSD value for the wavelet transform is-9.317 dB/Hz, which is better than that of adaptive NLMS (-6.788 dB/Hz) and notch (-6.669 dB/Hz) filters. Therefore, the analysis based on performance parameters has justified that proposed biorthogonal wavelet family represent better performance for reducing baseline wander from the ECG signal than adaptive NLMS and notch filters.

Research paper thumbnail of REMOVING BASELINE WANDER FROM ECG SIGNAL USING WAVELET TRANSFORM

REMOVING BASELINE WANDER FROM ECG SIGNAL USING WAVELET TRANSFORM, Sep 4, 2019

Electrocardiogram (ECG) signal is the representation of electrical activity generated by heart mu... more Electrocardiogram (ECG) signal is the representation of electrical activity generated by heart muscles, which is primarily utilized to detect cardiac abnormalities. Due to the sensitive nature of ECG, its important features are affected by different noises and create problems for diagnosis. This study proposes biorthogonal wavelet family by investigating different wavelet families to reduce baseline wander from the ECG signal. The proposed approach performance is compared to adaptive normalized least-mean-square (NLMS) and notch filters. Different performance parameters, such as amplitude spectrum, magnitude squared coherence (MSC), and power spectral density (PSD) has been evaluated. Signal-to-noise ratio (SNR), percentage root-mean-square difference (PRD), mean-square-error (MSE), normalized mean-square-error (NMSE), root mean-square-error (RMSE), and normalized root mean-square-error (NRMSE) performance parameters are calculated as well. The SNR values of the reconstructed ECG signal are-0.0046 dB and 1.6122 dB for notch and adaptive NLMS filters, respectively, which are lower than that of 8.0464 dB for the biorthogonal wavelet transform. Similarly, the MSC values are 0.091903 and 0.44522 after notch and adaptive NLMS filtrations, respectively, which are lower than those of 0.8913 after wavelet filtration. Also, the PSD value for the wavelet transform is-9.317 dB/Hz, which is better than that of adaptive NLMS (-6.788 dB/Hz) and notch (-6.669 dB/Hz) filters. Therefore, the analysis based on performance parameters has justified that proposed biorthogonal wavelet family represent better performance for reducing baseline wander from the ECG signal than adaptive NLMS and notch filters.