Foetal phonocardiographic signal denoising based on non-negative matrix factorization

Area of Science:

• Biomedical Engineering
• Signal Processing
• Maternal-Fetal Medicine

Background:

• Foetal phonocardiography (fPCG) is a valuable non-invasive tool for antenatal care.
• fPCG signals offer crucial diagnostic information on foetal health.
• Noise contamination in fPCG signals necessitates advanced signal processing for accurate analysis.

Purpose of the Study:

• To develop and evaluate a novel methodology for effective noise removal from fPCG signals.
• To enhance the diagnostic quality of fPCG data through signal denoising.
• To improve the assessment of foetal well-being using processed fPCG signals.

Main Methods:

• Utilization of the non-negative matrix factorization (NMF) algorithm for signal denoising.
• Testing the developed NMF methodology on both simulated and real-time fPCG datasets.
• Performance evaluation based on the achieved gain in signal-to-noise ratio (SNR).

Main Results:

• The NMF-based methodology demonstrated significant noise reduction capabilities.
• Substantial improvements in SNR, ranging from 12-30 dB, were achieved.
• The denoising process resulted in high-quality fPCG signals suitable for clinical assessment.

Conclusions:

• The proposed NMF methodology offers an effective solution for denoising fPCG signals.
• Improved SNR in fPCG data leads to more reliable foetal health monitoring.
• This technique enhances the utility of fPCG in routine antenatal care and diagnostics.