Scholarlink

<p>Electrocardiogram (ECG), is a vital biological signal that is used to diagnose cardiovascular diseases. There is often noise present when ECG data is being acquired in a healthcare facility. Baseline wander (BWR) is a noise that can distort the ECG signal and cause erroneous interpretations. BWR describes low-frequency variations in ECG signal baseline that can conceal the underlying physiological data. As a result, great attention has been paid to ECG denoising for accurate diagnosis and analysis. In this research, we have used the filter bank-based technique to remove BWR from the ECG. We present a dyadic boundary point-based empirical wavelet transform (DPET) filter bank for multi-scale ECG decomposition into distinct sub-bands (SBND). MIT-BIH Arrhythmia dataset, which is openly accessible, is used to support the proposed methodology. When compared using performance metrics such as the signal-to-noise ratio of the output (STNRo) and the correlation coefficient (<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(C_e\)</EquationSource> </InlineEquation>) at various intensities of the signal-to-noise ratio of the input (STNRi), our simulation results show that the suggested technique performs better than other state-of-the-art ECG filtering procedures.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Expunging Baseline Wander from ECG Signals by Empirical Wavelet Transform Based Dyadic Boundary Points

  • 期刊论文

  • 发表日期：

    2025年7月23日

查看全文
  • Diksha Sharma,
  • M. Krishna Chaitanya,
  • Jammisetty Yedukondalu,
  • Jagdeep Rahul,
  • Achintya Kumar Sarkar,
  • Lakhan Dev Sharma

摘要

Electrocardiogram (ECG), is a vital biological signal that is used to diagnose cardiovascular diseases. There is often noise present when ECG data is being acquired in a healthcare facility. Baseline wander (BWR) is a noise that can distort the ECG signal and cause erroneous interpretations. BWR describes low-frequency variations in ECG signal baseline that can conceal the underlying physiological data. As a result, great attention has been paid to ECG denoising for accurate diagnosis and analysis. In this research, we have used the filter bank-based technique to remove BWR from the ECG. We present a dyadic boundary point-based empirical wavelet transform (DPET) filter bank for multi-scale ECG decomposition into distinct sub-bands (SBND). MIT-BIH Arrhythmia dataset, which is openly accessible, is used to support the proposed methodology. When compared using performance metrics such as the signal-to-noise ratio of the output (STNRo) and the correlation coefficient ( \(C_e\) ) at various intensities of the signal-to-noise ratio of the input (STNRi), our simulation results show that the suggested technique performs better than other state-of-the-art ECG filtering procedures.