Ming Meng, Assistant Professor of Psychological and Brain Sciences, Dartmouth College
A "predictive coding" hypothesis proposes that the brain dynamically anticipates and generates predictions about upcoming stimuli to guide perception efficiently. Here, we combined functional magnetic resonance imaging (fMRI), multi-voxel pattern decoding, and an innovative time-resolved psychophysical paradigm to assess the temporal profile and spatial distributions of this prediction process.
Strikingly, we demonstrate rhythmic population activity in several task-related brain areas. Specifically, multi-voxel activity patterns in the fusiform face area (FFA) and the parahippocampal place area (PPA) show temporal fluctuations at a theta-band (~5 Hz) rhythm that accompany effects in visual object priming. These results provide novel and essential constraints to understand the neuronal dynamics of predictive coding. Moreover, these results suggest a feasible fMRI strategy to measure temporal fluctuations of multi-voxel activity patterns in the human brain, providing a critical link between fMRI measurements and neurophysiological recordings to understand fine-scale spatiotemporal dynamics of attention and consciousness.