Multimodal integration of high-resolution EEG and functional magnetic resonance imaging data: a simulation study

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biophysics

Background:

• Cortical current density estimation is crucial for understanding brain activity.
• Functional magnetic resonance imaging (fMRI) priors can enhance these estimations.
• Previous studies lacked systematic analysis of signal-to-noise ratio (SNR) and electrode count effects.

Purpose of the Study:

• To investigate the utility of fMRI priors in cortical current density estimation.
• To systematically evaluate the impact of SNR, electrode number, and fMRI prior strength.
• To compare standard and fMRI-based inverse operators under varying conditions.

Main Methods:

• Utilized a realistic head and cortical surface model for simulations.
• Varied SNR, inverse operators, fMRI prior strengths, and number of electrodes.
• Assessed estimation accuracy using correlation coefficient and relative error at cortical regions of interest (ROIs).

Main Results:

• All simulated factors (SNR, inverse operators, fMRI priors, electrode count) significantly affected estimation accuracy.
• fMRI-based inverse operators yielded superior cortical source current estimation when ROIs had simulated fMRI hot spots.
• No significant difference in accuracy was found between standard and fMRI-based operators for ROIs without fMRI hot spots.

Conclusions:

• fMRI priors are valuable for improving cortical current density estimation, particularly in brain regions with detectable fMRI activity.
• The choice of inverse operator and data quality (SNR, electrode number) are critical for accurate source localization.
• The benefit of fMRI priors is context-dependent, being most pronounced in the presence of fMRI hot spots.