Dunsmoor et al., 2014 Cerebral Cortex
Averisve Learning Modulates Cortical Representations of Object Cortex
Results
- Analyses include RSA and PPI following category conditioning.
- Found increased overall neural activity in CS+ related visual cortex
- increased CS+ vs CS- representational dissimilarity in respective category cortices
Notes
- Excluded voxels with mean CS+ > CS- activity in RSA to reduce SNR
Ritchey et al., 2013 Cerebral Cortex
Neural Similarity Between Encoding and Retrieval is Related to Memory via Hippocampal Interactions
Methods
- viewed emotionally negative, positive, or neutral images, 24hr delayed recognition memory test
- computed RSA on item level, or grouped by valence, task (E vs. R), and memory (R vs. F) across items
- set level groupings exclude item pairs (no diagonal)
- trial betas were Z-scored according to mean activity within a trial, something that I’m not doing (yet)
- logistic regression analysis - trial mean ROI activation during either encoding retrieval or with pattern similarity to to predict binary memory outcome
- significant non-zero coefficient implies mean activation estiamtes and encoding retrieval similarity make sepreateble contributions to memory
- restricted subsequent memory analyses to regions whihc ER similarity reminaed significant predictor of memory
- also ran a mediation analysis testing if HC activity mediates relationship between ER similarity and 5 choice memory responses
Results
- In ROIs with main effect of match level also looked at effect of emotion
- middle occipital gyrus showed strongest match X emotion interaction
- Found lots of ROIs that vary representations based on match , i.e. (item or set level)
- Also found lots of ROIs that vary representations based on memory, regardless of match
- Interaction Match X Memory mostly in MTL and VTC
- Hippocampal mediation analysis found sig. mediation of ER similarity and memory in inferior frontal, inferior parietal, and occipital ROIs
- MOG ER similarity correlated with amygdala activity for negative remembered items more than negaitve forgotten items
- In the end though, HC was more important than Amygdala in predicting memory, suggesting amygdala responses are driven by “recovery and porcessing of item-specific detials”
Tambini et al., 2017 Nature Neuroscience
Emotional brain states carry over and enhance future memory formation
Then relevence here is in the methods, as it might not be fair to compare the emotional –> neutral viewing group the same as our conditioning –> extinction.
Methods
- low frequency connectivity of amygdala and anterior hippocampus
- filter out task frequency to essentially look at “resting state”
- RSA Approach included permutation testing