Related Experiment Video
Updated: Dec 13, 2025
Author Spotlight: Capturing Infant-Caregiver Interactions Through Synchronized Multimodal Data Collection
Published on: May 31, 2024
Using multiple short epochs optimises the stability of infant EEG connectivity parameters
Rianne Haartsen1, Bauke van der Velde2,3,4, Emily J H Jones5
1Department of Psychological Sciences (BMA), Centre for Brain and Cognitive Development, Birkbeck College, University of London, Malet Street, London, WC1E 7HX, UK. rhaart01@mail.bbk.ac.uk.
Optimizing electroencephalography (EEG) analysis for infant brain connectivity is crucial for understanding neurodevelopmental disorders. Using many short epochs with the debiased weighted phase lag index (dbWPLI) offers the most reliable results with limited infant EEG data.
Area of Science:
- Neuroscience
- Developmental Neuroscience
- Computational Neuroscience
Background:
- Atypical brain connectivity is linked to neurocognitive disorders.
- Assessing stable individual differences in developing brain connectivity is challenging due to data limitations.
- Reliable metrics are needed for infant brain connectivity research.
Purpose of the Study:
- To optimize electroencephalography (EEG) processing parameters for reliable infant brain connectivity measures.
- To evaluate the impact of epoch length and number on test-retest reliability of EEG connectivity metrics.
- To determine the most suitable method for characterizing connectivity in infants with limited EEG data.
Main Methods:
- EEG data were collected twice from 10-month-old infants with a 1-week interval.
- Alpha band connectivity was analyzed using the phase lag index (PLI) and debiased weighted PLI (dbWPLI).
- Connectivity was assessed across varying epoch lengths and numbers, including whole-head and graph theory metrics.
Main Results:
- The debiased weighted PLI (dbWPLI) showed higher reliability across numerous short epochs.
- The phase lag index (PLI) demonstrated higher reliability with fewer, longer epochs, but was sensitive to segment count.
- Whole-brain connectivity measures were more reliable than graph theory metrics.
Conclusions:
- For infant populations with limited EEG data, segmenting data into many short epochs and employing the dbWPLI is recommended for robust connectivity characterization.
- This approach enhances the reliability of EEG connectivity measures in developmental studies.
- Optimized EEG analysis parameters are vital for advancing our understanding of neurodevelopmental trajectories and disorders.
