Shared neural activities for schema

Chen, J., Leong, Y. C., Honey, C. J., Yong, C. H., Norman, K. A., & Hasson, U. (2017). Shared memories reveal shared structure in neural activity across individuals. Nat Neurosci, 20(1), 115-125. doi:10.1038/nn.4450

How memory is formed and retrieved is one of the most "hard-core" questions in cognitive neuroscience. Recently, a paper published in Nature Neuroscience found that the pattern of neural activity when recalling memorized information across individuals is more similar than the neural pattern between encoding-recall. 

In this study, participants are required to view two video clips that were extracted from TV series "Sherlock", then they recall the episode of the video without any cues. That's, free-viewing, free-recalling. Like we view a TV and then tell it to a friend, but in a fMRI scanner. 

The authors found that the neural activities in certain brain regions during viewing and recalling are highly correlated, including the default mode network (DMN), medial-temporal, and high-level visual areas. Also, they found that for the individual event, the neural activity pattern is highly discriminable from one another, at the same time, these neural activity patterns are similar across individuals. More interestingly, the similarity of activation pattern during recall certain events across individuals is higher than encoding-recall similarity.

The main methods used in this study, as I arbitrarily called it, event-locked spatial pattern similarity analysis (PSA). But it is totally different from traditional event-related fMRI design because this study didn't put any constraints on participants. The "event-locked" part of the analysis was content-based, i.e., based on the narrative of the story (as presented by the video and recalling). So it is more ecological than many previous studies.

Four PSAs were used (notes in brackets are added by me):

(1) Movie-recall within participant (--> memory consolidation); 

(2) Movie-recall between participants ( --> shared memory pattern); 

(3) Recall-recall between-participant (--> shared memory pattern);

(4) movie-movie between participants (--> shared encoding pattern)

Of course, we should not forget the alternation analysis: comparison between movie-recall neural activity similarity and recall-recall pattern similarity.

From my perspective, the most interesting result is that the systematic alternations across individuals from perception (encoding) and recall. When I was reading the paper, I thought the authors would mention about the schema, which is an old concept in psychology, in the discussion part. Indeed, they mentioned it. So these systematic alternations may reflect the meaningful integration of memory, which extracted meaningful and abstract information, leave the perceptual feature.

A fantastic study! I have two ideas after reading this paper: (1), what if the video is not narrative; (2) what about the individual differences; (3) what about the errors in the recall? (maybe I need to read more details in this article, it is not an easy task to read it)

Asl, it is interesting to associate this study and PTSD, for whom the intrusive memory is often perceptual details and salient. For those suffering from PTSD, is the memory integration is as good as normal people? Where is the part that makes the differences? Those may require me to read more in the relevant literature.