Research on ECG signal reconstruction based on improved weighted nuclear norm minimization and approximate message...

Area of Science:

• Biomedical Engineering
• Signal Processing
• Wearable Technology

Background:

• Improving energy efficiency in wearable devices requires efficient compression and reconstruction of electrocardiogram (ECG) data.
• Transmission of compressed ECG data can introduce noise, necessitating robust denoising techniques for accurate signal reconstruction.

Purpose of the Study:

• To introduce a denoising-based approximate message passing (AMP) algorithm for ECG signal reconstruction.
• To enhance the weighted nuclear norm minimization (WNNM) algorithm for improved denoising performance in noisy ECG signals.

Main Methods:

• The study integrates a denoising-based approximate message passing (AMP) algorithm into ECG signal reconstruction.
• An improved weighted nuclear norm minimization (WNNM) algorithm, termed IWNNM, is proposed, utilizing weighted averaging for signal blocks post-low-rank decomposition.
• The effectiveness of the IWNNM-AMP algorithm is validated on electrocardiogram signals under various compression ratios and noise levels.

Main Results:

• The proposed IWNNM-AMP algorithm demonstrates superior reconstruction performance across different compression ratios and noise conditions.
• IWNNM-AMP achieves the lowest values for Percentage Root Mean Square Difference (PRD) and Root Mean Square Error (RMSE).
• Compared to the WNNM-AMP algorithm, IWNNM-AMP shows a reduction in PRD by 0.17–4.56 and an improvement in Peak Signal-to-Noise Ratio (P-SNR) by 0.12–2.70.

Conclusions:

• The enhanced IWNNM-AMP algorithm significantly improves the reconstruction accuracy of noisy ECG signals from wearable devices.
• Weighted averaging in the WNNM algorithm effectively addresses noise-induced errors in similar block searching, enhancing denoising.
• The IWNNM-AMP algorithm presents a promising solution for reliable ECG data reconstruction in energy-efficient wearable systems.