hide
Free keywords:
-
Abstract:
Real-time fMRI Neurofeedback (rtfMRI NF) is the technique in which blood oxygen level dependent (BOLD) response of specific Region of Interest (ROI) is presented in real time to the participant with the goal of enabling the subjects to volitionally regulate their brain signals. Neurofeedback is considered a promising alternative or supplementary
treatment (Arns et al., 2012, Alegria et al., 2017). Because many neuropsychiatric disorders, MDD, ADHD, etc. are often associated with pathological changes in dynamic interactions between brain areas (that is, functional brain networks), the ability to modulate neural dynamics on a network level with neurofeedback may be a more effective than neurofeedback involving a single area (Sitaram et.al., 2017). We present an Independent Component Analysis (ICA) - based approach which finds the individualized ROI that can be used as the target of NF regulation within real-time fMRI neurofeedback. First, individual resting-state fMRI data are decomposed into 40 Independent Components (IC) using GIFT toolbox, next the Independent Component (IC) that corresponds to a specific network is identified and, on the last step, voxels that pass a certain intensity threshold within specific brain region are chosen as Region of Interest.The algorithm was validated on resting state 3T fMRI recordings of 37 subjects. It shows highly consistent results across subjects, different software packages used to pre-process the data: SPM (The FIL Methods group), CONN (Whitfield-Gabrieli, and Nieto-Castanon, 2012), Turbo-BrainVoyager (www.brainvoyager.com). We also used a special recording with extremely inclined and rotated head and the algorithm showed high accuracy. The execution time of our algorithm is in the range between 1.5 to 5 minutes, depending on the available computational power, which makes it possible to use this algorithm for the rtfMRI NF experiments.
