Related Experiment Video
Updated: Apr 17, 2026
Author Spotlight: Advancing the Study of Brain-Heart Interplay with a Comprehensive EEGLAB Plugin for Multimodal Signal Analysis
Published on: April 26, 2024
Denoising preterm EEG by signal decomposition and adaptive filtering: a comparative study.
X Navarro1, F Porée2, A Beuchée3
1INSERM, U1099, Rennes, F-35000, France; Université de Rennes 1, Laboratoire Traitement du Signal et de l'Image, Rennes, F-35000, France; Sorbonne Universités, UPMC Univ Paris 06, UMRS-1158, Neurophysiologie Respiratoire Expérimentale et Clinique, Paris, F-75005, France.
Noise in preterm infant electroencephalography (EEG) hinders analysis. Combining EEG decomposition techniques with adaptive filters (AF) significantly improves signal denoising, reducing errors by up to 30% for more reliable preterm infant monitoring.
Area of Science:
- Biomedical Engineering
- Neuroscience
- Signal Processing
Background:
- Electroencephalography (EEG) monitoring in preterm infants is crucial for neurological assessment.
- Noise contamination in preterm EEG signals compromises accurate interpretation and automated analysis.
- Conventional filtering methods like band-pass and adaptive filters (AF) may be insufficient for complex preterm EEG patterns.
Purpose of the Study:
- To enhance the denoising process for preterm infant EEG signals.
- To evaluate the efficacy of combining EEG decomposition techniques with adaptive filters.
- To improve the reliability of automated analysis of preterm EEG.
Main Methods:
- Artificially contaminated real preterm EEG signals were used for simulations.
- Compared denoising performance of discrete wavelet transform, empirical mode decomposition (EMD), and complete ensemble EMD with adaptive noise (CEEMDAN).
- Assessed the impact of applying decomposition techniques prior to adaptive filtering.
Main Results:
- EMD-based decomposition techniques, particularly CEEMDAN, showed superior performance.
- The combination of EEG decomposition and AF reduced root mean squared errors by up to 30%.
- Improved signal quality was observed for challenging preterm EEG patterns like tracé alternant and slow delta-waves.
Conclusions:
- Integrating EMD-based decomposition methods with adaptive filtering offers a significant improvement in preterm EEG denoising.
- This combined approach enhances the reliability of preterm infant EEG signal analysis.
- The findings suggest a more robust method for processing noisy neonatal EEG data.
