Adaptive Fourier decomposition based ECG denoising.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Cardiology

Background:

• Electrocardiogram (ECG) signals are crucial for diagnosing heart conditions.
• Noise in ECG signals can obscure important diagnostic features, leading to misdiagnosis.
• Existing denoising methods struggle with noise that overlaps in frequency with the ECG signal.

Purpose of the Study:

• To propose a novel ECG denoising method using adaptive Fourier decomposition (AFD).
• To evaluate the effectiveness of the AFD-based method in separating ECG signals from noise with overlapping frequency ranges.
• To compare the performance of the AFD method against other established ECG denoising techniques.

Main Methods:

• Adaptive Fourier Decomposition (AFD) was employed to decompose ECG signals based on energy distribution.
• A novel stop criterion for the AFD iterative process was developed using estimated signal-to-noise ratio (SNR).
• The method was validated using synthetic ECG signals and real-world data from the MIT-BIH Arrhythmia Database with added Gaussian white noise.

Main Results:

• The AFD-based method demonstrated superior performance in ECG denoising compared to wavelet transform, Stockwell transform, empirical mode decomposition, and ensemble empirical mode decomposition.
• The proposed method also showed improved accuracy in QRS complex detection.
• AFD effectively separated ECG signals from noise even when frequency ranges overlapped, leveraging differences in energy distribution.

Conclusions:

• The adaptive Fourier decomposition (AFD) presents a highly effective and novel approach for ECG signal denoising.
• This method offers significant advantages over existing techniques, particularly for signals with overlapping noise frequencies.
• The AFD-based denoising method holds promise for enhancing the accuracy of automated ECG analysis and diagnosis.