Global and Local Functional Gradients of the Human Cortex
Poster No:
1641
Submission Type:
Abstract Submission
Authors:
Hazel Milla1, Khoi Huynh1, Kim-Han Thung1, Hoyt Patrick Taylor1, Guoye Lin1, Sahar Ahmad1, Pew-Thian Yap1
Institutions:
1University of North Carolina at Chapel Hill, Chapel Hill, NC
First Author:
Co-Author(s):
Introduction:
This study aims to stratify the cortices of young adults based on resting-state functional connectivity (RSFC), using global and local gradients computed via a unified framework with diffusion embedding (Margulies et al., 2016). Global gradients represent the major axes of cortical organization, outlining the brain's overall structure, while local gradients capture finer details, highlighting localized transitions. Characterizing the brain at multiple levels, and integrating these insights, is particularly important for understanding complex neurodevelopment (Sydnor et al. 2021).
We introduce a framework designed to quantify gradients across multiple scales, unifying global gradients (Margulies et al., 2016) and local gradients (Glasser et al., 2016). This approach enables the characterization of functional organizational principles from diverse perspectives, offering a comprehensive understanding of both global and local patterns within the young adult cortex.
We introduce a framework designed to quantify gradients across multiple scales, unifying global gradients (Margulies et al., 2016) and local gradients (Glasser et al., 2016). This approach enables the characterization of functional organizational principles from diverse perspectives, offering a comprehensive understanding of both global and local patterns within the young adult cortex.
Methods:
Our study utilized resting-state fMRI (rs-fMRI) from 1,094 young adult subjects from the Human Connectome Project Young Adult (HCP-YA). rs-fMRI data were mapped on individual surfaces consisting of 10,242 vertices for each hemisphere. We calculated vertex-wise functional connectivity (FC) as Pearson correlations of vertex time series. Then we applied diffusion embedding to decompose FC into multiple components, represented as eigenvectors of the diffusion matrix (derived from the FC). Among these, the first three nontrivial components-referred to as global gradients-captured major axes of variation: the somatosensory-association (S-A), visual-somatosensory (V-S), and modulation-representation (M-R) axes (Taylor et al., 2024).
Rather than limiting our analysis to the first few components, we computed gradients using up to 1,000 components, collectively capturing approximately 90% of the information in the functional connectivity (FC) matrix. To identify local gradients, we focused on components 900 to 1,000, balancing spatial detail with noise. The FC matrix was projected onto this component space, yielding a 20,484 × 20,484 matrix. We then squared this matrix and averaged its rows to produce a 20,484-length vector, representing the local gradient across both hemispheres.
Arealization was identified using the local gradient by selecting vertices above the 90th percentile threshold. These vertices were classified as boundaries delineate distinct functional regions. Boundaries that were more consistent across subjects were assigned higher edge probabilities.
Rather than limiting our analysis to the first few components, we computed gradients using up to 1,000 components, collectively capturing approximately 90% of the information in the functional connectivity (FC) matrix. To identify local gradients, we focused on components 900 to 1,000, balancing spatial detail with noise. The FC matrix was projected onto this component space, yielding a 20,484 × 20,484 matrix. We then squared this matrix and averaged its rows to produce a 20,484-length vector, representing the local gradient across both hemispheres.
Arealization was identified using the local gradient by selecting vertices above the 90th percentile threshold. These vertices were classified as boundaries delineate distinct functional regions. Boundaries that were more consistent across subjects were assigned higher edge probabilities.
Results:
When mapped onto a smooth cortical surface, large-scale gradients align visually with established gradients (S-A, V-S, and M-R). Local gradients provide additional detail, revealing fine-grained boundaries and variability both within and between networks. Both global and local gradients demonstrate strong left-right symmetry. Averaging across subjects reduces individual noise, improving the identification of consistent network boundaries and transitions across participants (Fig. 1).
Figure 2 shows that the Ventral Attention B and Control A networks exhibit a wide range of edge probabilities, indicating varying degrees of arealization. In contrast, the Default Mode C network displays a narrower distribution. Additionally, the Somatomotor B and Dorsal Attention B networks generally have lower edge probabilities compared to networks such as Control A.
Figure 2 shows that the Ventral Attention B and Control A networks exhibit a wide range of edge probabilities, indicating varying degrees of arealization. In contrast, the Default Mode C network displays a narrower distribution. Additionally, the Somatomotor B and Dorsal Attention B networks generally have lower edge probabilities compared to networks such as Control A.
·Figure 1. Global and local gradients for an individual and a population of 1,094 healthy young adults. The edge probability map illustrates the likelihood of identifying a boundary at each vertex.
·Figure 2. Region-based edge probabilities calculated based on the Schaefer parcellation. ROIs are colored and grouped according to Yeo’s 17 functional networks (Thomas Yeo et al., 2011).
Conclusions:
We demonstrated that diffusion embedding can serve as a unifying framework for computing both global and local gradients, helping to understand the different organizational principles of the human cortex based on functional connectivity.
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural)
fMRI Connectivity and Network Modeling 2
Methods Development
Segmentation and Parcellation 1
Task-Independent and Resting-State Analysis
Keywords:
Computational Neuroscience
Computing
Cortex
Data analysis
Design and Analysis
Development
FUNCTIONAL MRI
Segmentation
Other - Parcellation
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.
2. Margulies, D. S., Ghosh, S. S., Goulas, A., Falkiewicz, M., Huntenburg, J. M., Langs, G., Bezgin, G., Eickhoff, S. B., Castellanos, F. X., Petrides, M., Jefferies, E., & Smallwood, J. (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
3. Sydnor, V. J., Larsen, B., Bassett, D. S., Alexander-Bloch, A., Fair, D. A., Liston, C., Mackey, A. P., Milham, M. P., Pines, A., Roalf, D. R., Seidlitz, J., Xu, T., Raznahan, A., & Satterthwaite, T. D. (2021). Neurodevelopment of the association cortices: Patterns, mechanisms, and implications for psychopathology. Neuron, 109(18), 2820–2846. https://doi.org/10.1016/j.neuron.2021.06.016
4. Taylor, H.P., Thung, K.H., Huynh, K.M., Lin, W., Ahmad, S. and Yap, P.T., 2024. Functional Hierarchy of the Human Neocortex from Cradle to Grave. bioRxiv.
5. Thomas Yeo, B. T., Krienen, F. M., Sepulcre, J., Sabuncu, M. R., Lashkari, D., Hollinshead, M., Roffman, J. L., Smoller, J. W., Zöllei, L., Polimeni, J. R., Fischl, B., Liu, H., & Buckner, R. L. (2011). The organization of the human cerebral cortex estimated by intrinsic functional connectivity. Journal of Neurophysiology, 106(3), 1125–1165. https://doi.org/10.1152/jn.00338.2011