Manipulating Social Percepts During fMRI

During each fMRI session, we will monitor subjects' brain activity in real time and use univariate and/or multivariate activity patterns to trigger trial presentation. For each individual, we will select three regions: one key region early in the cortical hierarchy (e.g., MT), one key region late in the cortical hierarchy (e.g., anterior temporal), and one control region that does not show strong tuning to social information (e.g., primary auditory). These regions will be functionally defined for each individual using data from their characterization sessions. Subjects will report percepts behaviorally after each trial. For each region for each individual, we will then test the correlation between pre-stimulus activity and the likelihood of reporting certain percepts across trials.

Neurofeedback is a type of biofeedback that uses real-time displays of brain activity in an attempt to teach subjects to self-regulate their brain function. Here we will use implicit neurofeedback, meaning that subjects will see a feedback signal, but will not be told what that feedback signal represents (i.e., which brain region) or what the expected effects of learning to control that brain region might be. This type of neurofeedback is less prone to suggestibility and other confounding effects. Each of the two scan sessions dedicated to this intervention will be split into four equal blocks in a 2x2 design with region (early or late) and direction of modulation (up or down). Subjects will see a continuous feedback signal representing the region and direction, and will be asked to try to modulate the signal using any strategy that occurs to them.