Modeling BOLD-fMRI Hemodynamics via Multidimensional Decomposition of Electrophysiology Data: A Simulation Study

Area of Science:

• Neuroscience
• Biophysics
• Data Science

Background:

• Simultaneous electroencephalography (EEG)/local field potential (LFP) and BOLD-fMRI recordings offer complementary insights into brain function.
• Analyzing the complex interactions between electrophysiological signals and hemodynamic responses remains a challenge.

Purpose of the Study:

• To introduce a novel framework for investigating the electrophysiological correlates of BOLD-fMRI data.
• To demonstrate the utility of structured coupled matrix-tensor factorization (sCMTF) for joint analysis of LFP/EEG and BOLD signals.

Main Methods:

• Implementation of a structured coupled matrix-tensor factorization (sCMTF) framework for joint multidimensional decomposition.
• Application of sCMTF to resting-state whole-brain modeling data.
• Validation using permuted datasets to assess the significance of extracted EEG/LFP temporal patterns and hemodynamic response functions (HRFs).

Main Results:

• The sCMTF framework successfully reveals dynamical interactions between LFP/EEG oscillatory features and BOLD-fMRI data.
• Significant correlations were found between extracted EEG/LFP temporal patterns and BOLD signal fluctuations.
• The framework accurately estimates hemodynamic response functions (HRFs) that embody simulated hemodynamics.

Conclusions:

• The proposed sCMTF framework provides a powerful tool for studying electrophysiological correlates of BOLD-fMRI.
• This approach enables the estimation of LFP/EEG-BOLD co-fluctuations and regional HRFs.
• Further research may benefit from careful consideration of sCMTF algorithm initialization for optimal performance.