Time-varying synergy and redundancy dominance in the human cerebral cortex
Poster No:
1609
Submission Type:
Abstract Submission
Authors:
Maria Pope1, Thomas Varley2, Maria Grazia Puxeddu3, Joshua Faskowitz4, Olaf Sporns1
Institutions:
1Indiana University, Bloomington, IN, 2University of Vermont, Burlington, VT, 3University of Rome Sapienza, Rome, Rome, 4National Institute of Mental Health, Bethesda, MD
First Author:
Co-Author(s):
Introduction:
There has been substantial recent interest in higher-order interactions in brain activity [Petri et al., 2014, Santoro et al., 2023, Hindriks et al., 2024], and in particular, in distinguishing between redundant and synergistic higher order interactions using information theory [Varley et al., 2023a, Luppi et al., 2022, Varley et al., 2023b, Gatica et al., 2021]. Several studies have revealed that traditional pairwise methods of studying interactions between brain regions, such as functional connectivity analyses, almost exclusively capture redundancy and are blind to synergistic interactions [Varley et al., 2023a, Varley et al., 2023b]. However, synergistic interactions are thought to be highly relevant to information modification and computation in complex systems[Lizier, 2013] and have begun to be fruitfully identified in neural systems. In humans, synergistic interactions are clinically relevant and implicated in Alzheimer's disease [Arbabyazd et al., 2023] and stroke recovery [Pirovano et al., 2023]. They also change characteristically during healthy aging [Gatica et al., 2021]. Whether synergistic and redundant higher-order interactions in the cortex are time-varying has not yet been established, and, as a consequence, their temporal structure is largely unknown. Here we provide a thorough analysis of ongoing higher-order dynamics during resting state fMRI.
Methods:
We studied the time-varying synergy/redundancy dominance of functional magnetic resonance imaging (fMRI) data taken from 100 unrelated subjects of the Human Connectome Project and parcellated according to the Schaefer 200 cortical parcellation. The BOLD time series were analyzed using the local O-information, which measures whether an interaction is synergy or redundancy dominated at each point in time. We studied several sizes of interaction. First, we treated the whole brain as a single large interaction and calculated a single local O-information time series for all time points. Second, we randomly sampled subsets of size three to 25. 10,000 subsets were sampled for each size, and a local O-information time series was calculated for each subset. Third, we found the maximally synergistic and maximally redundant triad independently for each time point. This was done by exhaustively calculating the local O-information time series for every possible triad (1,313,400 triads, 418,000 time points per triad) and retaining the triad with the most synergistic and most redundant local O-information at each time point. Finally, for larger subsets, calculation of all possible subsets is prohibitive, so we performed an optimization using a simulated annealing algorithm. The algorithm was run independently on every time point and for subset sizes five to seventy-five.
Further analyses were performed to analyze temporal autocorrelation, recurrence, and correspondence to nodal activity of synergistic and redundant states.
Further analyses were performed to analyze temporal autocorrelation, recurrence, and correspondence to nodal activity of synergistic and redundant states.
·Schematic of Approach
Results:
Focusing on momentary synergy and redundancy dominance reveals that it is redundancy-dominated subsets that experience the most strongly synergy-dominated moments. We have further shown that these synergistic moments often occur when the subset tends to be evenly split across the bipartition created by the signs of the nodal activity, which, for small subsets, occurs at those moments when typically highly coherent systems (such as the Yeo systems) disintegrate. In addition, we have demonstrated that strongly synergistic and strongly redundant interactions of all sizes have temporal structure. They change smoothly in time and recur at significantly higher rates than expected by chance (see second Figure included in abstract).
·Redundant and Synergistic Subsets Show Recurrence
Conclusions:
We have shown that local O-info is effective in identifying recurrent interactions with specific interregion relationships. Recurrence indicates that interactions are not random, but are dynamic states the cortex regularly returns to. Both of these traits make it a promising measure for future application to studies of ongoing cognition.
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural)
fMRI Connectivity and Network Modeling 2
Multivariate Approaches 1
Keywords:
Computational Neuroscience
Other - Information Theory; Cortical Dynamics; Higher Order Interactions
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?
Are you Internal Review Board (IRB) certified? Please note: Failure to have IRB, if applicable will lead to automatic rejection of abstract.
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:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
tiotemporal organization of dynamic functional connectivity are disrupted by
Alzheimer’s disease. Network Neuroscience, 7(4):1420–1451.
Gatica, M.(2021). High-Order Interdepen-
dencies in the Aging Brain. Brain Connectivity, 11(9):734–744.
Hindriks, R. (2024). Higher-order functional connectivity analysis of
resting-state functional magnetic resonance imaging data using multi-
variate cumulants. Human Brain Mapping, 45(5):e26663.
Lizier, J. T. (2013). The Local Information Dynamics of Dis-
tributed Computation in Complex Systems. Springer Theses. Springer Berlin
Heidelberg, Berlin, Heidelberg.
Luppi, A. I. (2022). A Synergistic Core for human brain evolution and cognition. Nature Neuroscience 25 (6), 771-782
Petri, G. (2014). Homological scaffolds of brain functional networks. Journal of The Royal Society Interface,
11(101):20140873.
Pirovano, I. (2023). Rehabilitation modulates high-order interactions among
large-scale brain networks in subacute stroke. IEEE Transactions on Neural
Systems and Rehabilitation Engineering, 31:4549–4560.
Santoro, A. (2023). Higher-order organization of multivariate time series. Nat. Phys.,
pages 1–9.
Varley, T. F. (2023a). Multivariate information theory uncovers synergistic subsystems of
the human cerebral cortex. Commun Biol, 6(1):1–12.
Varley, T. F., Pope, M., Grazia, M., Joshua, and Sporns,
O. (2023b). Partial entropy decomposition reveals higher-order information
structures in human brain activity. Proceedings of the National Academy
of Sciences, 120(30):e2300888120.