Abstract
Two techniques for investigating the neural coding of stimulus- specific visual information are fMRI adaptation (fMRIa) and multi-voxel pattern analysis (MVPA). To understand and compare how fMRIa and MVPA representations relate to behavior, we scanned eight participants while they viewed repeated presentations of scenes, and tested their subsequent memory for each scene. We observed robust adaptation within the parahippocampal place area (PPA) and greater pattern similarity across repetitions of the same scene relative to non-identical scenes, consistent with previous studies (Epstein & Morgan, 2011; Morgan et al., 2011). Furthermore, we observed that scenes showing greater pattern similarity were more likely to be remembered (Xue et al., 2010), while adaptation measures were not predictive of memory in this study. Interestingly, the amount of adaptation and the degree of pattern similarity was not significantly correlated for either remembered or forgotten scenes. These results inform hypotheses about the mechanisms for visual representation across repeated viewings and how these representations relate to what we later remember.
Meeting abstract presented at VSS 2012