Spatial (mis)match between EEG and fMRI signal patterns revealed by spatio-spectral source-space EEG decomposition

Area of Science:

• Neuroscience
• Cognitive Science
• Biophysics

Background:

• Simultaneous electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) offer complementary insights into neural dynamics.
• Reconciling findings across EEG and fMRI remains a challenge due to differing spatial and temporal resolutions.

Purpose of the Study:

• To directly compare whole-brain EEG neural dynamics patterns with concurrently measured fMRI BOLD data.
• To derive and validate spatio-spectral EEG patterns in source-reconstructed space and their corresponding BOLD signatures.

Main Methods:

• Utilized a large dataset of 72 subjects undergoing resting-state high-density EEG-fMRI (hdEEG-fMRI).
• Applied spatio-spectral decomposition to band-limited EEG power in source-reconstructed space.
• Assessed reliability and reproducibility of extracted EEG patterns and their spatial BOLD signatures.

Main Results:

• Identified five robust EEG spatio-spectral patterns, including known occipital alpha dynamics and novel patterns.
• Observed weak but statistically significant spatial similarity between EEG patterns and fMRI BOLD signatures, especially with stronger temporal synchronization.
• Found no significant relationship between EEG patterns and classical fMRI BOLD resting-state networks.

Conclusions:

• Both EEG (frequency-specific) power and fMRI BOLD signal capture reproducible spatio-temporal neural dynamics.
• EEG and fMRI provide largely complementary, rather than redundant, information about low-frequency brain dynamics.
• The derived EEG patterns offer new insights into brain activity beyond established findings.