A Novel Technique for Fetal ECG Extraction Using Single-Channel Abdominal Recording

Area of Science:

• Biomedical Engineering
• Signal Processing
• Maternal-Fetal Medicine

Background:

• Non-invasive fetal electrocardiograms (FECGs) offer continuous monitoring but suffer from weak signal amplitude and overlap with maternal ECG (MECG) in abdominal recordings.
• Existing adaptive noise cancellation (ANC) methods struggle with nonlinear signal propagation, where maternal thoracic and abdominal ECGs lack waveform similarity.
• Research in non-invasive FECG separation is emerging, highlighting the need for advanced signal processing techniques.

Purpose of the Study:

• To develop and validate a novel method for accurately separating FECG from abdominal ECG (AECG) recordings.
• To address the limitations of traditional ANC by creating a robust reference signal for MECG cancellation.
• To improve the detection of fetal QRS complexes in non-invasive monitoring.

Main Methods:

• A hybrid approach combining singular value decomposition (SVD) and smooth window (SW) techniques was employed to construct a reference signal for ANC.
• The proposed method aims to overcome the waveform similarity limitation of standard ANC by generating a reference signal independent of separate maternal thoracic ECG recordings.
• Validation was performed using public datasets (r01, r07) and a private self-recorded dataset.

Main Results:

• The proposed method achieved high F1 scores for fetal QRS detection: 99.61% (r01), 99.28% (r07), and 98.58% (private dataset).
• Compared to four other single-channel methods, the proposed approach demonstrated superior accuracy: 99.22% (r01), 98.57% (r07), and 97.21% (private dataset).
• The SVD-SW based ANC effectively extracted FECG signals, outperforming existing techniques.

Conclusions:

• The developed SVD-SW combined ANC method shows significant potential for accurate FECG extraction from MECG recordings.
• This technique offers a promising solution for enhancing non-invasive fetal monitoring in both clinical and commercial applications.
• The findings suggest a pathway towards more reliable and accessible fetal health assessment.