Categorized EEG neurofeedback performance unveils simultaneous fMRI deep brain activation

Decades of Electroencephalogram-NeuroFeedback (EEG-NF) practice have proven that people can be effectively trained to selectively regulate their brain activity, thus potentially improving performance. A common protocol of EEG-NF training aims to guide people via a closed-loop operation shifting from high-amplitude of alpha (8-14Hz) to high-amplitude of theta (4-7 Hz) oscillations resulting in greater theta/alpha ratio (T/A). The induction of such a shift in EEG oscillations has been shown to be useful in reaching a state of relaxation in psychiatric conditions of anxiety and mood disorders. However, the clinical implication of this practice remains elusive and is considered to have relatively low therapeutic yield, possibly due to its poor specificity to a unique brain mechanism. The current project aims to use simultaneous acquisition of Functional Magnetic Resonance Imaging (fMRI) and EEG in order to unfold in high spatial and temporal resolutions, respectively the neural modulations induced via T/A EEG-NF. We used real time EEG preprocessing and analysis during the simultaneous T/A EEG-NF/fMRI. A data driven algorithm was implemented off-line to categorize individual scans into responders and non-responders to the EEG-NF practice via a temporal signature of T/A continuous modulation. Comparing the two groups along with their parasympathetic Heart-Rate reactivity profile verified the relaxed state of the responders. Projection of responders variations in the T/A power to the fMRI whole brain maps revealed networks of correlated and inversely correlated activity reflecting induced relaxation, uniquely among responders.