Combining EEG Microstates with fMRI Structural Features for Modeling Brain Activity.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Computational Neuroscience

Background:

• Combining Electroencephalography (EEG) and Functional Magnetic Resonance Imaging (fMRI) offers complementary strengths for brain activity analysis.
• EEG provides high temporal resolution (milliseconds), while fMRI offers good spatial resolution.
• Integrating EEG and fMRI data aims to improve both spatial and temporal localization of neural activity.

Purpose of the Study:

• To present a novel technique for combining EEG and fMRI data for a more comprehensive understanding of brain activity.
• To enhance the representation and understanding of brain activities in time by integrating multimodal neuroimaging data.

Main Methods:

• EEG data modeled using microstates and Hidden Markov Models (HMMs) to capture temporal dynamics of Event Related Potentials (ERPs).
• Fisher score calculated to quantify sequence deviation from learned HMMs.
• Canonical Partial Least Squares (CPLS) used for decomposing and fusing EEG and fMRI datasets.

Main Results:

• The novel methodology successfully derived components that co-vary between EEG and fMRI signals.
• Significant differences in derived components were observed between different tasks.
• Comparison with CPLS using single-channel ERP showed the effectiveness of the proposed method.

Conclusions:

• The presented technique effectively integrates EEG and fMRI data, improving the spatiotemporal localization of brain activity.
• This multimodal approach offers a more nuanced understanding of neural dynamics compared to single-modality analyses.
• The method holds promise for advancing research in cognitive neuroscience and clinical applications.