Tracking Information Flow Across the Brain Using Representational Connectivity Analyses
Poster No:
1199
Submission Type:
Abstract Submission
Authors:
Hamid Karimi-Rouzbahani1, Jason Mattingley2
Institutions:
1Queensland Brain Institute, The University of Queensland, Brisbane, Queensland, 2University of Queensland, St Lucia, Queensland
First Author:
Hamid Karimi-Rouzbahani
Queensland Brain Institute, The University of Queensland
Brisbane, Queensland
Queensland Brain Institute, The University of Queensland
Brisbane, Queensland
Co-Author:
Introduction:
Functional connectivity (FC) methods, which quantify co-activity between brain areas, have limited utility in understanding information flow (Anzelotti & Coutanche, 2018). Instead, in cognitive neuroscience, we are interested in how and if "information" - the neural representations distinguishing cognitive states - is transferred. Examples of cognitive states can be the differentiation between different orientations of grating stimuli (Fig 1). To address this, we introduce Representational Connectivity Analyses (RCAs), a novel method that tracks information flow across the brain (Fig 1).
Unlike traditional FC, which summarise activity patterns within brain areas losing information (Basti et al., 2021), RCAs compare information representation in multi-dimensional activation patterns across brain areas (Fig 1). This allows us to track various types of information, such as stimulus features, task, and response patterns. Moreover, RCAs can detect information flow even when neural representations are transformed between brain areas (Karimi-Rouzbahani et al., 2022).
We demonstrate the versatility and power of RCAs in revealing millisecond scale information flow, by applying them on diverse neural data, including monkey and mouse electrophysiology in visual perception experiments. Our findings reveal the dynamic nature of information flow during visual recognition, highlighting the role of both feed-forward and feedback mechanisms in shaping visual perception.
Unlike traditional FC, which summarise activity patterns within brain areas losing information (Basti et al., 2021), RCAs compare information representation in multi-dimensional activation patterns across brain areas (Fig 1). This allows us to track various types of information, such as stimulus features, task, and response patterns. Moreover, RCAs can detect information flow even when neural representations are transformed between brain areas (Karimi-Rouzbahani et al., 2022).
We demonstrate the versatility and power of RCAs in revealing millisecond scale information flow, by applying them on diverse neural data, including monkey and mouse electrophysiology in visual perception experiments. Our findings reveal the dynamic nature of information flow during visual recognition, highlighting the role of both feed-forward and feedback mechanisms in shaping visual perception.
·Figure 1
Methods:
We previously used RCAs to track visual stimulus and task information between sensory and cognitive brain areas in human EEG, MEG and fMRI (Karimi-Rouzbahani et al., 2021; Karimi-Rouzbahani et al., 2024). To investigate neural interactions at a finer spatial scale, we applied RCAs to monkey electrophysiology data. We analysed a dataset where monkeys categorized morphed cat and dog images based on task cues (Roy et al., 2010).
Monkey studies typically record fewer than 1000 neurons, often non-simultaneously, introducing inter-session variability. To gain a comprehensive understanding of neural interactions at a finer level, we analysed the Allen Institute's Mouse Brain Observatory dataset (Siegle et al., 2021). This dataset provides simultaneous recordings from millions of neurons during passive viewing of static gratings.
We applied a novel model-based RCA method to track information flow in the monkey and mouse datasets which were built on our previous work (Karimi-Rouzbahani et al., 2021; Fig 2).
Monkey studies typically record fewer than 1000 neurons, often non-simultaneously, introducing inter-session variability. To gain a comprehensive understanding of neural interactions at a finer level, we analysed the Allen Institute's Mouse Brain Observatory dataset (Siegle et al., 2021). This dataset provides simultaneous recordings from millions of neurons during passive viewing of static gratings.
We applied a novel model-based RCA method to track information flow in the monkey and mouse datasets which were built on our previous work (Karimi-Rouzbahani et al., 2021; Fig 2).
Results:
In the monkey dataset, we found earlier feedback flow of task information (i.e., the difference between the two categorisation tasks; Fig 2B, left) from cognitive (prefrontal) to sensory (inferior temporal) areas to potentially enhance the task-relevant categories. This led to an earlier and stronger feed-forward flow of relevant (i.e. difference between categories across the relevant boundary) vs. irrelevant category information (Fig 2B, right). These results revealed, for the first time, that feedback flows carried multiple information contents including tasks and categories. Moreover, feedback flows were dynamic being early for task and late for category information.
In the mouse dataset, we found two clear components of information flow across the visual cortex 1) transient and dominantly feed-forward, 2) sustained and recurrent (Fig 2C). This supported previous human studies showing earlier feed-forward followed by recurrent flows (Kietzmann et al., 2019) and reveals the previously unknown information content: both components carried more information about spatial frequency than orientation and phase of the grating stimuli (Fig 2D).
In the mouse dataset, we found two clear components of information flow across the visual cortex 1) transient and dominantly feed-forward, 2) sustained and recurrent (Fig 2C). This supported previous human studies showing earlier feed-forward followed by recurrent flows (Kietzmann et al., 2019) and reveals the previously unknown information content: both components carried more information about spatial frequency than orientation and phase of the grating stimuli (Fig 2D).
·Figure 2
Conclusions:
We used RCA to provide a more nuanced understanding of information flow across the brain in different modalities and species during visual perception. These establish the versatility and power of RCA as a novel tool for tracking information in the brain with great potential for answering a wide range of questions in cognitive neuroscience.
Higher Cognitive Functions:
Executive Function, Cognitive Control and Decision Making
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural) 1
Perception, Attention and Motor Behavior:
Perception: Visual 2
Keywords:
Cognition
Computational Neuroscience
Cortex
Vision
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
Provide references using APA citation style.
Anzellotti, S., & Coutanche, M. N. (2018). Beyond functional connectivity: investigating networks of multivariate representations. Trends in cognitive sciences, 22(3), 258-269.
Basti, A., Nili, H., Hauk, O., Marzetti, L., & Henson, R. N. (2020). Multi-dimensional connectivity: a conceptual and mathematical review. NeuroImage, 221, 117179.
Karimi-Rouzbahani, H., Ramezani, F., Woolgar, A., Rich, A., & Ghodrati, M. (2021). Perceptual difficulty modulates the direction of information flow in familiar face recognition. NeuroImage, 233, 117896.
Karimi-Rouzbahani, H., Rich, A. N., & Woolgar, A. (2024). Spatiotemporal characterisation of information coding and exchange in the multiple demand network. bioRxiv, 2024-10.
Karimi-Rouzbahani, H., Woolgar, A., Henson, R., & Nili, H. (2022). Caveats and nuances of model-based and model-free representational connectivity analysis. Frontiers in Neuroscience, 16, 755988.
Kietzmann, T. C., Spoerer, C. J., Sörensen, L. K., Cichy, R. M., Hauk, O., & Kriegeskorte, N. (2019). Recurrence is required to capture the representational dynamics of the human visual system. Proceedings of the National Academy of Sciences, 116(43), 21854-21863.
Roy, J. E., Riesenhuber, M., Poggio, T., & Miller, E. K. (2010). Prefrontal cortex activity during flexible categorization. Journal of Neuroscience, 30(25), 8519-8528.
Siegle, J. H., Jia, X., Durand, S., Gale, S., Bennett, C., Graddis, N., ... & Koch, C. (2021). Survey of spiking in the mouse visual system reveals functional hierarchy. Nature, 592(7852), 86-92.