RT Journal Article SR Electronic T1 Multi-Connection Pattern Analysis: Decoding the Representational Content of Neural Communication JF bioRxiv FD Cold Spring Harbor Laboratory SP 046441 DO 10.1101/046441 A1 Yuanning Li A1 Avniel Singh Ghuman YR 2016 UL http://biorxiv.org/content/early/2016/03/31/046441.abstract AB It is unknown what information is represented in distributed brain circuit interactions because we lack methods for decoding the representational content of interregional neural communication. Here we present Multi-Connection Pattern Analysis (MCPA), which is designed to probe the nature of the representational space contained in the multivariate functional connectivity pattern between neural populations. MCPA works by learning mappings between the activity patterns of the populations separately for each condition, stimulus, or brain state in training data. These maps are used to predict the activity from one neural population based on the activity from the other population as a factor of the information being processed in test data. Successful MCPA-based decoding indicates the involvement of distributed computational processing and provides a framework for probing the representational structure of the interaction. Simulations demonstrate the efficacy of MCPA for decoding distributed information processing across a set of realistic circumstances and show that MCPA is insensitive to local information processing. Furthermore, applying MCPA to human intracranial electrophysiological data demonstrates that the interaction between occipital face area and fusiform face area contains information about individual faces. Representational analysis indicates that the OFA-FFA interaction codes face information differently than either the OFA or FFA individually. These results support the hypothesis that face individuation occurs not in a single region, but through interactive computation and distributed information representation across the face processing network. Thus, MCPA, can be used to assess the information processed the coupled activity of distributed, interacting neural circuits.Significance Statement Information is represented in the brain by the coordinated activity of neurons both at the regional level and the level of large-scale, distributed networks. Multivariate methods from machine learning have advanced our understanding of the representational structure of local information coding, but the nature of distributed information representation remains unknown. Here we present a novel method that integrates multivariate connectivity analysis with machine learning classification techniques that can be used to decode the representational structure of neural interactions. This method is then used to provide a novel neuroscientific insight, that information about individual faces is distributed across at least two critical nodes of the face-processing network. Thus, this work provides a framework to assess the representational content of circuit-level processing.