Robust fetal QRS detection from noninvasive abdominal electrocardiogram based on channel selection and simultaneous ...

Area of Science:

• Biomedical Engineering
• Signal Processing
• Maternal-Fetal Medicine

Background:

• Accurate fetal heart rate monitoring is crucial for maternal and fetal well-being.
• Current methods for fetal electrocardiogram (fECG) analysis often require complex separation of fetal and maternal signals (fECG/mECG).
• Existing techniques may necessitate specialized thoracic channels, increasing computational load and limiting device applicability.

Purpose of the Study:

• To introduce a novel, computationally efficient method for detecting fetal QRS complexes and extracting fetal heart rate (FHR) from non-invasive fECG.
• To achieve high-accuracy FHR extraction without the need for fECG/mECG signal separation.
• To eliminate the requirement for thoracic channels, thereby reducing computational complexity and enhancing suitability for real-time applications.

Main Methods:

• A new channel scoring system prioritizing noise distribution and fetal heartbeat visibility was developed.
• Simultaneous and combinatorial use of selected fECG channels based on their scores.
• Extraction of FHR using a combination of geometric features and wavelet-based techniques.
• Categorization of fECG signals based on fetal geometric features for tailored analysis.

Main Results:

• The proposed method achieves high accuracy in FHR extraction without fECG/mECG separation.
• Reduced computational requirements due to the elimination of signal separation and thoracic channels.
• Validation across multiple datasets (Noninvasive fetal ECG database, DaISy, PhysioNet/Computing in Cardiology Challenge 2013) demonstrates robust performance.
• The method's effectiveness was confirmed through comparison with existing studies.

Conclusions:

• The novel approach offers a simplified and efficient method for non-invasive fetal heart rate monitoring.
• Eliminating the need for fECG/mECG separation and thoracic channels significantly enhances applicability in real-time healthcare devices.
• Intelligent channel selection and simultaneous utilization contribute to the method's promising performance and potential for widespread adoption.