A novel controllable energy constraints-variational mode decomposition denoising algorithm.

Area of Science:

• Biomedical Engineering
• Signal Processing

Background:

• Electrocardiogram (ECG) signals are vital for diagnosing heart conditions.
• Noise in ECG signals significantly compromises diagnostic accuracy.
• Existing denoising methods often struggle to balance noise removal with feature preservation.

Purpose of the Study:

• To develop a novel algorithm for denoising ECG signals.
• To enhance the accuracy of ECG-based heart disease diagnosis.
• To improve upon existing signal processing techniques for ECG noise reduction.

Main Methods:

• The proposed Controlled Energy Constraint-Variational Mode Decomposition (CEC-VMD) algorithm is employed.
• Noisy ECG signals are decomposed into intrinsic mode functions (IMFs) and a residual.
• A modulation factor and ADMM-based updates are utilized to refine modal properties and minimize residual information.

Main Results:

• CEC-VMD achieved an average SNR of 22.5139, RMSE of 0.1128, and CC of 0.9882 on simulated and MIT-BIH signals.
• The algorithm significantly improved classification accuracy to 99.0% (SVM) and 99.9% (KNN).
• CEC-VMD outperformed traditional methods like EMD, VMD, and SWT in denoising and feature preservation.

Conclusions:

• The CEC-VMD technique effectively removes noise from ECG signals.
• This method enhances the preservation of essential ECG signal features.
• The improved signal quality leads to more accurate heart disease classification.