Beyond functional connectivity: Exploring high-order lenses in brain networks
Poster No:
1459
Submission Type:
Abstract Submission
Authors:
Marilyn Gatica1, Andrea Santoro2, Simone Poetto3, Davide Orsenigo3, Matteo Neri4, Giovanni Petri1
Institutions:
1Northeastern University London, London, United Kingdom, 2CENTAI Institute, Torino, Italy, 3NPLab, Torino, Italy, 4Institut de Neurosciences de la Timone, Aix Marseille Université, Marseille , France
First Author:
Co-Author(s):
Introduction:
Brain interdependencies are typically studied through structural connectivity, which focuses on pairwise interactions via white-matter fibers, and functional connectivity (FC), which examines statistical relationships between brain regions. While FC typically focuses on pairwise interactions, recent methods have expanded this approach to infer higher-order interactions (HOIs) involving three or more regions [1]. These HOI methods have gained attention for their potential to enhance our understanding of brain function in areas such as health [2, 3], aging [4, 5], cognition [6], neuromodulation [7], and consciousness [8]. This study investigates when these complex methods provide advantages over traditional approaches and how different HOI techniques relate to one another, using topological and informationtheoretic methods to offer a more comprehensive view of brain dynamics.
Methods:
We analyzed resting-state fMRI data from 100 subjects in the Human Connectome Project (HCP), computing traditional pairwise and higher-order metrics, categorized into information-theoretic and topological approaches. Measures included: Functional Connectivity (FC), synergy (Syn) and redundancy (Red) from O-Information [9], PhiID [10], and PED [3]. Topological measures included the homological scaffold (Pscaff, Fscaff) [11] and a dynamic approach [2, 12] for extracting hypercoherent triangles and the scaffold. These measures provided insights into both traditional and complex brain functional organization. By systematically applying these measures, we captured both traditional and complex aspects of the brain's functional organization. To ensure comparability, all measures were either directly used or edge-projected and then averaged across the population.
Results:
We compared the HOI approaches against FC when representing rfMRI to identify key differences among HOI methods. For each measure, we analyzed direct or edge-projected activations averaged across subjects. In Figure 1a, we report the Pearson's ρ between all measures. Here, redundancy and triangle measures closely resemble FC (ρ ≈ 0.9), while synergistic and scaffold measures are less correlated. Figure 1b reports the hierarchical clustering obtained from the similarity matrix, revealing distinct clusters that align with topological and information-theoretic methods, further split into synergy and redundancy categories. In a second analysis, we examined HOI measures against the principal cortical gradient of the human connectome [13], which ranges from unimodal sensory regions to transmodal higher-order areas. Synergistic measures showed the highest correlation with this gradient, followed by scaffold, redundancies, and FC, suggesting that measures beyond FC better capture cortical organization (Figure 2a). Moreover, differences in synergy and redundancy in PhiID and PED, as well as triangles and scaffold measures, were highly correlated with the principal gradient (Figure 2b). While most of the HOI approaches were linked to the princial gradient, we also identified associations between different HOI measures and distinct receptor maps [14] (not shown here), suggesting that HOI measures capture diverse aspects of cortical organization.
Conclusions:
Our study demonstrates that HOI measures provide diverse insights into brain organization beyond classical functional connectivity (FC). While FC primarily relates to redundancy and triangle measures and shows only minimal correlation with synergy and scaffold measures, HOI measures reveal cortical maps aligned with established cognitive gradients from low- to high-order tasks [13]. Consistent with prior research [3, 6, 12], our systematic analysis and benchmarking of HOI versus FC measures within a single dataset highlight the distinct contributions of topological and information-theoretic methodologies. Additionally, we identified associations between HOI measures and distinct receptor maps, suggesting that HOI measures capture multiple dimensions of cortical organization.
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural)
fMRI Connectivity and Network Modeling 1
Multivariate Approaches 2
Novel Imaging Acquisition Methods:
BOLD fMRI
Keywords:
Computational Neuroscience
FUNCTIONAL MRI
Machine Learning
Multivariate
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:
Provide references using APA citation style.
2. Santoro, A., et al. (2023). Higher-order interactions uncover synergies in the human brain. Nature Physics, 19(2), 221–228. https://doi.org/10.1038/s41567-022-01847-7
3. Varley, T. F., et al. (2023). Higher-order interactions in brain dynamics. arXiv. https://arxiv.org/abs/2301.05307
4. Gatica, M., et al. (2021). High-order interdependencies in the aging brain. Brain Connectivity, 11(6), 438–448. https://doi.org/10.1089/brain.2020.0922
5. Gatica, M., et al. (2022). Integrated information and interdependencies in the human brain. PLoS Computational Biology, 18(7), e1010301. https://doi.org/10.1371/journal.pcbi.1010301
6. Luppi, A. I., et al. (2022). A synergistic core for human brain evolution and cognition. Nature Neuroscience, 25(6), 771–782. https://doi.org/10.1038/s41593-022-01092-x
7. Gatica, M., et al. (2024). Transcranial ultrasound stimulation effect in the macaque brain. Network Neuroscience. https://doi.org/10.1162/netn_a_00388
8. Luppi, A. I., et al. (2023). Consciousness and integrated information in severe brain injury. eLife, 12, e88173. https://doi.org/10.7554/eLife.88173
9. Rosas, F., et al. (2019). Quantifying high-order interdependencies via multivariate extensions of the mutual information. Physical Review E, 100(3), 032305. https://doi.org/10.1103/PhysRevE.100.032305
10. Mediano, P. A. M., et al. (2021). Towards a theory of integrated information decomposition. arXiv. https://arxiv.org/abs/2109.13186
11. Petri, G., et al. (2014). Homological scaffolds of brain functional networks. Journal of the Royal Society Interface, 11(101), 20140873. https://doi.org/10.1098/rsif.2014.0873
12. Santoro, A., et al. (2024). Topological indicators of higher-order interactions in human brain networks. Nature Communications, 15(1), 10244. https://doi.org/10.1038/s41467-024-54472-y
13. Margulies, D. S., et al. (2016). Situating the default-mode network along a principal gradient of macroscale cortical organization. Proceedings of the National Academy of Sciences, 113(44), 12574–12579. https://doi.org/10.1073/pnas.1608282113
14. Markello, R. D., et al. (2022). Neuromaps: Structural and functional interpretation of brain maps. Nature Methods, 19(12), 1472–1484. https://doi.org/10.1038/s41592-022-01628-8