Phase Amplitude Coupling workflow for mapping EEG signals to standard brain networks

Abstract

Human brain mapping or neuroimaging plays a pivotal role in understanding the intricacies of the human brain and paving the way for potential therapeutic interventions. Studying the standard brain networks, typically obtained from fMRI, provide valuable insights into the fundamental organization of the human brain. In this work we present a workflow for processing electroencephalography (EEG) signals to determine the correlations of the phase-amplitude coupling (PAC) of the standard brain networks during a given timewindow. We validate this pipeline with synthetic signals on realistic head models of two subjects with the ultimate goal of studying the changes of these networks during different sleep stages. The proposed workflow consists of: mapping the signals to the source space, averaging per Brodmann Area (BA), low and high pass filtering, computing the modulation index per low-high frequency pair, generating surrogate data to obtain significance thresholds, obtaining the significant PAC signals, computing the signal and noise covariance matrices, removing the model bias, and applying confirmatory factor analysis (CEA) to determine the relevance of each standard brain network. We included the novelty of using CEA instead of principal component analysis as done in previous studies. We tested the workflow with synthetic signals, and it performed as expected. Next steps will be fine-tuning it and improving its robustness before processing real signals during sleep that we already have collected for the two subjects of the head models used here.