Poster No:
1402
Submission Type:
Abstract Submission
Authors:
Sida Chen1, Qianyuan Tang2, Werner Sommer3, Taro Toyoizumi4, Lianchun Yu5, Changsong Zhou6
Institutions:
1Hong Kong Baptist University, Kowloon City, Hong Kong, 2Hong Kong Baptist University, Kowloon, Hong Kong, 3Humboldt-Universität zu Berlin, Berlin, Berlin, 4Lab for Neural Computation and Adaptation, RIKEN Center for Brain Science, 2-1 Hirosawa, Wako, Saita, Saitama, Wako, 5Lanzhou University, Lanzhou, Gansu, 6Hong Kong Baptist University, Hong Kong, Hong Kong
First Author:
Sida Chen
Hong Kong Baptist University
Kowloon City, Hong Kong
Co-Author(s):
Taro Toyoizumi
Lab for Neural Computation and Adaptation, RIKEN Center for Brain Science, 2-1 Hirosawa, Wako, Saita
Saitama, Wako
Introduction:
Individual differences in brain function, influenced by genetic, neural, and environmental factors, significantly affect cognitive performance. Linking these differences to task outcomes is challenging due to the high dimensionality and dynamic fluctuations of brain networks. Traditional methods often fail to isolate functionally relevant variance in this complex space.
The concept of "sloppiness," derived from biological systems, describes systems robust to most parameter changes but sensitive to variations in a few key dimensions, known as "stiff" directions. Identifying these stiff dimensions offers a novel approach to understanding how neural variability contributes to task performance.
This study introduces sloppiness analysis as a framework for linking individual differences in brain network dynamics to cognitive outcomes. Focusing on the default mode network (DMN) and working memory network (WMN), we explore how stiff dimensions influence working memory, shedding light on integration and segregation dynamics shaping cognitive performance.
Methods:
Functional magnetic resonance imaging (fMRI) data from 991 participants performing a working memory task were analyzed. Using the pairwise maximum entropy model (PMEM), we modeled task-related brain dynamics for 21 brain regions within the DMN and WMN. Parameters representing activity levels (h) and effective connectivity (J) were estimated at both group and individual levels. Fisher Information Matrix (FIM) analysis was used to characterize sensitivity to parameter variations, identifying stiff and sloppy dimensions via eigendecomposition.
Key analyses included projecting individual parameters onto FIM eigenvectors, examining the relationship between parameter variations and functional connectivity (FC), and linking these variations to task performance. Predictive models of working memory performance were constructed using stiff dimensions, and parameter sensitivity was assessed to refine predictions further.
Results:
Sloppiness in Brain Networks:
Task-related brain networks exhibited stiff and sloppy dimensions, with FIM eigenvalues distributed in a power-law-like manner. Stiff dimensions reflected significant sensitivity in brain dynamics, primarily associated with global segregation and localized integration within and between the DMN and WMN.
Integration and Segregation:
Variations along the stiffest dimensions (η₁ and η₂) corresponded to distinct patterns. η₁ correlated with increased segregation between and within the DMN and WMN, while η₂ reflected enhanced integration within the WMN and slight segregation within the DMN.
Task Performance Prediction:
Stiff parameters robustly predicted working memory performance. Combining global segregation (η₁) and localized integration (η₂) achieved optimal predictive power. Sensitivity analysis revealed that excluding less sensitive parameters improved predictions, demonstrating the robustness of stiff dimensions in linking neural variability to task outcomes.
Functional Roles of Brain Regions:
Brain regions with high positive sensitivity (e.g., dlPFC, IPS) supported working memory integration, while those with high negative sensitivity (e.g., mPFC, PCC) reflected task-related segregation. Strengthened connectivity in regions with positive sensitivity and reduced connectivity in negative sensitivity regions were associated with better performance.
Conclusions:
Sloppiness analysis reveals that task-related brain dynamics are governed by a few stiff dimensions that significantly influence cognitive performance. These dimensions capture critical integration and segregation processes within large-scale brain networks, offering a framework to link neural variability to task outcomes. By robustly predicting working memory performance and identifying functionally relevant parameters, this approach advances personalized neuroscience and opens new avenues for therapeutic interventions.
Learning and Memory:
Working Memory
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI) 2
fMRI Connectivity and Network Modeling 1
Keywords:
Cognition
Computational Neuroscience
Data analysis
Memory
Modeling
MRI
1|2Indicates the priority used for review
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.
I accept
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:
I am submitting this abstract as an original work to be reproduced. I am available to be the “source party” in an upcoming team and consent to have this work listed on the OSSIG website. I agree to be contacted by OSSIG regarding the challenge and may share data used in this abstract with another team.
Please indicate below if your study was a "resting state" or "task-activation” study.
Task-activation
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Healthy subjects
Was this research conducted in the United States?
No
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.
Yes
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.
Not applicable
Please indicate which methods were used in your research:
Functional MRI
Neuropsychological testing
Computational modeling
For human MRI, what field strength scanner do you use?
3.0T
Which processing packages did you use for your study?
SPM
FSL
Provide references using APA citation style.
Ashourvan, A., P. Shah, A. Pines, S. Gu, C. W. Lynn, D. S. Bassett, K. A. Davis and B. Litt (2021). "Pairwise maximum entropy model explains the role of white matter structure in shaping emergent co-activation states." Commun Biol 4(1): 210.
Gutenkunst, R. N., J. J. Waterfall, F. P. Casey, K. S. Brown, C. R. Myers and J. P. Sethna (2007). "Universally sloppy parameter sensitivities in systems biology models." PLoS Comput Biol 3(10): 1871-1878.
No