Distributed Patterns of Brain Activity Underlying Real-Time fMRI Neurofeedback Training

摘要: Neurofeedback (NF) based on real-time functional magnetic resonance imaging (rt-fMRI) is an exciting neuroimaging application. In most rt-fMRI NF studies, the activity level of a single region interest (ROI) provided as feedback signal and participants are trained to up or down regulate signal. training effects typically analyzed using confirmatory univariate approach, i.e., changes in target ROI explained by linear modulation. However, learning self-regulate through mediated distributed across brain. Here, we deploy multivariate decoding model for assessing whole Specifically, first explain effect posthoc that leads pattern coactivation 90 atlas regions. We then use cross validation reveal set brain regions with best fit. This novel approach was applied data from study where learned auditory cortex. found optimal consisted 16 whose patterns described over days. Cross showed it generalized participants. Interestingly, could be clustered into two groups distinct coactivation, potentially reflecting different strategies. Overall, our findings revealed multiple involved ROI, thus leading better understanding mechanisms underlying training.