Title : Multi - Connection Pattern Analysis : Decoding the Representational 1 Content of Neural Communication

33 34 The lack of multivariate methods for decoding the representational content of 35 interregional neural communication has left it difficult to know what information is 36 represented in distributed brain circuit interactions. Here we present Multi-Connection 37 Pattern Analysis (MCPA), which works by learning mappings between the activity 38 patterns of the populations as a factor of the information being processed. These maps are 39 used to predict the activity from one neural population based on the activity from the 40 other population. Successful MCPA-based decoding indicates the involvement of 41 distributed computational processing and provides a framework for probing the 42 representational structure of the interaction. Simulations demonstrate the efficacy of 43 MCPA in realistic circumstances. Applying MCPA to fMRI data shows that interactions 44 between visual cortex regions are sensitive to information that distinguishes individual 45 natural images, suggesting that image individuation occurs through interactive 46 computation across the visual processing network. MCPA-based representational 47 similarity analyses (RSA) results support models of error coding in interactions among 48 regions of the network. Further RSA analyses relate the non-linear information 49 transformation operations between layers of a computational model (HMAX) of visual 50 processing to the information transformation between regions of the visual processing 51 network. Additionally, applying MCPA to human intracranial electrophysiological data 52 demonstrates that the interaction between occipital face area and fusiform face area 53 contains information about individual faces. Thus, MCPA can be used to assess the 54 information represented in the coupled activity of interacting neural circuits and probe the 55 underlying principles of information transformation between regions. 56 . CC-BY-NC-ND 4.0 International license peer-reviewed) is the author/funder. It is made available under a The copyright holder for this preprint (which was not . http://dx.doi.org/10.1101/046441 doi: bioRxiv preprint first posted online Mar. 31, 2016;