A special application of functional MRI is real-time fMRI neurofeedback (rt-fMRI-NF). In this modality, fMRI is used to access brain activity while the subject is in the scanner, and a certain measure of brain activity is presented back to the subject. The idea is that if a subject is aware of his/her ongoing brain activity, he/she can be trained to influence it. Although not in routine clinical practice, encouraging results have been obtained with rt-fMRI-NF. For example, certain cases of depression are characterized by reduced left amygdala activity. Using rt-fMRI-NF, depressed subjects can be trained to upregulate left amygdala activity, and their clinical symptoms improve.
Within the Neu3ca program we are investigating rt-fMRI-NF as a way to mitigate cognitive impairment in epilepsy. Since complex brain functions such as cognition are typically mediated by brain networks (and not single regions), this includes implementing rt-fMRI-NF for networks rather than stand-alone regions. Furthermore, we are looking for biomarkers of functional network impairment that may be exploit for rt-fMRI-NF. Having identified these, they may be used to provide neurofeedback and to study the resulting changes in functional network organization and clinical outcome.
Another important issue is the sustainability of the effect; how many training sessions are needed in order to have an effect that extends the rt-fMRI-NF session (i.e. also outside the scanner)? And how can these changes be maintained; do subjects need to a new session of rt-fMRI-NF trainings every now and then? At what interval? Or can we maybe prolong the effect using ambulatory and/or less expensive modalities that assess brain activity (EEG neurofeedback, fNIRS neurofeedback)?