Training spatiotemporally-defined brain activity patterns with EEG neurofeedback

A host of mental disorders have been shown to exhibit abnormal activity in electroencephalographic (EEG) recordings. Among these abnormalities, a number of brain rhythms showed have been implicated in differences between healthy and non-healthy clinical populations. Based on these results, different studies have shown that subjects can be trained to self-regulate specific brain rhythms (e.g. alpha and beta oscillations) using a novel technique called neurofeedback. In parallel, studies on large-scale EEG synchronization events (termed “microstates”) have revealed new spatio-temporal biomarkers in multiple psychiatric disorders. Therefore, a question of clinical significance is whether these “microstate-based” biomarkers may also be self-regulated using real-time neurofeedback, derived from combining such spatial and/or temporal indicators.

The objective of this thesis will be to determine whether it is possible to translate the offline EEG biomarkers from microstates analysis in order to develop more spatio-temporally specific neurofeedback.

Initially, the project will focus on comparing electroencephalography (EEG) data in a population with a psychiatric disorders (in this case, ADHD) and healthy population in order to extract biomarkers that diagnostically differentiate the groups. Particular attention will be paid to spatio-temporally stable functional EEG states (so called “microstates”) and related dynamical measures of EEG activity. Then, healthy volunteers will be selected in order to test whether self-regulation of these neuromarkers is feasible as well as statistically significant. From this perspective, a new neurofeedback framework will be developed to test and demonstrate the ability of adult patients with ADHD to train and control spatio-temporally defined brain states. This will include real-time EEG data processing analysis through using microstates and/or spatial filter(s), together with evaluations of cognitive performance relevant to impulsivity and inattention. Should our hypotheses be confirmed, this approach could give rise to new personalized brain-computer interface applications for psychiatric disorders and thereby improve the quality of life of patients.