Denoising ECG by a New Wavelet Threshold Function

Dai, Bingze; Yang, Dixiong; Feng, Dongbo

doi:10.1109/cisp-bmei53629.2021.9624454

Cited by 4 publications

(3 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Conducted the study by implementing a one-dimensional analysis approach (duration of the heart cycle) of heart rate variability to multidimensional analysis (shape parameters and peak location of the heart cycle) analysis of heart parameters [9]. Conducted an ECG signal study with a wavelet, where the threshold function is improved, and resulted in the best SNR improvement using the MIT-BIH arrhythmia database [10].…”

Section: Related Workmentioning

confidence: 99%

Image Analysis of 12 Lead Electrocardiogram Using Wavelet Transformation

Darwan¹,

Anwar²,

Indra³

et al. 2023

ETJ

View full text Add to dashboard Cite

The heart is one of the most important human organs. One of instruments to detect cardiac abnormalities is the electrocardiogram (ECG). This research tries to analyze ECG image in normal heart condition from ECG machine. The previous research related to the pre-processing process is the same, only at the feature extraction process look for peaks P, Q, R, S, T, heart rate, and Deviation-ST. While this research is the characteristic extraction process using wavelet transformation. The image of lead ECG 12 is processed using discrete wavelet transforms with decomposition up to ten levels, by searching for mean square error (MSE). The type of mother wavelet and the wavelet order used are Daubechies (db) with 1 (db1 (Haar)). The smallest MSE value decomposition results are obtained at the level 5, which are lead I, II, III, aVR, aVF, V4 and V5, lead V1 & V2 on level 4, for aVL (level 9), V3 (level 7) and V4 (level 6). It is expected that such research can be followed up to the identification model of cardiac abnormalities using wavelets.

show abstract

Section: Related Workmentioning

confidence: 99%

Image Analysis of 12 Lead Electrocardiogram Using Wavelet Transformation

Darwan¹,

Anwar²,

Indra³

et al. 2023

ETJ

View full text Add to dashboard Cite

show abstract

“…However, because a noise reference is necessary in the adaptive filter, it is not the best practice. In another technique, the time-frequency analysis tool of the discrete wavelet transform (DWT) with a threshold is also adopted in noise filtering [12]- [15]. The DWT decomposes the noisy ECG into different frequency components.…”

Section: Introductionmentioning

confidence: 99%

“…However, the hard threshold leads to oscillation of the reconstructed ECG, and the soft threshold may reduce the amplitude of the ECG features. The approaches in [12], [14], [15] proposed an improved threshold to eliminate noise from ECG. The improved threshold not only avoided the reconstructed ECG oscillation but also effectively preserved the original features of the ECG.…”

Section: Introductionmentioning

confidence: 99%

Lightweight Denoising Autoencoder Design for Noise Removal in Electrocardiography

et al. 2022

View full text Add to dashboard Cite

This study proposes two denoising autoencoder (DAE) models with discrete cosine transform (DCT) and discrete wavelet transform (DWT), namely DCT-DAE and DWT-DAE, to remove electrode motion artifacts in noisy electrocardiography (ECG). Initially, the discrete cosine transform and discrete wavelet transform efficiently removed the high-frequency noise. The six encoder layers then retain important ECG features while the six decoder layers reconstruct the clean ECG. To improve the denoising performance, two network layers, the residual block and pixel adjustment, are added to the encoder and decoder layers to solve the vanishing gradient and improve subtle feature extraction. The proposed methods were applied to 66,000 real-recorded noisy ECG fragments. The experimental result indicates that DWT-DAE and DCT-DAE can improve the signal-to-noise ratio by 25.29 and 25.13 dB on average when the input signal-to-noise ratio is -6 dB.

show abstract