An improved morphological approach to background normalization of ECG signals

Area of Science:

• Biomedical Engineering
• Signal Processing
• Cardiology

Background:

• Neonatal electrocardiogram (ECG) signals are prone to baseline wander, which can obscure important diagnostic features.
• Existing methods for baseline wander removal may distort critical waveform components, impacting diagnostic accuracy.

Purpose of the Study:

• To develop and validate an improved morphological algorithm for accurate baseline wander removal in neonatal ECG signals.
• To enhance the preservation of the ST segment and overall waveform integrity during the baseline correction process.

Main Methods:

• A two-stage morphological processing approach is employed.
• Initial stage removes QRS complexes and impulsive noise.
• Second stage utilizes corrected QT (QTc) and RR intervals to determine a structuring element for removing baseline wander from the QRS-removed signal.

Main Results:

• The proposed algorithm demonstrates substantial improvement over existing morphological methods.
• Significant reduction in distortion of the baseline waveform within the PR and QT intervals was observed.
• Effective removal of baseline wander while preserving signal morphology.

Conclusions:

• The improved morphological algorithm offers a superior method for baseline wander removal in neonatal ECGs.
• This technique enhances the quality of neonatal ECG signals, aiding in more accurate interpretation.
• The method shows promise for clinical application in pediatric cardiology.