EEG spatiospectral patterns and their link to fMRI BOLD signal via variable hemodynamic response functions

Area of Science:

• Neuroscience
• Brain Imaging
• Systems Neuroscience

Background:

• Combining functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) can improve the spatial and temporal resolution of brain network activity analysis.
• fMRI offers whole-brain coverage but limited temporal resolution, while EEG provides high temporal resolution of cortical activity.
• Integrating these modalities offers potential for enhanced brain network characterization.

Purpose of the Study:

• To examine the relationships between EEG spatiospectral patterns and concurrent fMRI BOLD signals.
• To investigate the utility of different hemodynamic response function (HRF) shapes in modeling EEG-fMRI data.
• To characterize large-scale brain networks (LSBNs) using combined EEG-fMRI data across different experimental paradigms.

Main Methods:

• Utilized voxel-wise general linear models (GLMs) to analyze relationships between EEG spatiospectral pattern timecourses and fMRI BOLD signals.
• Incorporated the canonical HRF and its temporal derivatives to model the hemodynamic response.
• Derived HRF shapes from EEG-fMRI data acquired during resting-state, visual oddball, and semantic decision tasks.

Main Results:

• GLM F-maps revealed self-organized large-scale brain networks (LSBNs).
• Differences in GLM-derived HRF shapes indicated varying timing between EEG and fMRI signals, with some EEG-fMRI derived HRF peaks appearing earlier than the canonical HRF.
• Found significant correlations between the timecourses of 14 independent EEG spatiospectral patterns and fMRI dynamics within LSBN structures.
• Demonstrated that some EEG spatiospectral patterns are weakly task-modulated.

Conclusions:

• This study is the first to investigate EEG-fMRI relationships across independent EEG spatiospectral patterns and multiple paradigms.
• The findings underscore the limitations of assuming a canonical HRF shape in EEG-fMRI analyses.
• Emphasizes the importance of considering diverse HRF shapes for accurate spatiotemporal characterization of brain networks using combined EEG and fMRI.