Functional Parcellation of White Matter Tracks Using Track Dynamic Functional Connectivity
Poster No:
1197
Submission Type:
Abstract Submission
Authors:
Yifei Sun1, James Shine1, Sharon Naismith1, Fernando Calamante1, Jinglei Lv1
Institutions:
1The University of Sydney, Camperdown, New South Wales, Australia
First Author:
Co-Author(s):
Introduction:
The human brain's functional organisation is complex and not yet fully understood. Current studies commonly focus on the functional role of grey matter (GM), with functional atlases providing GM parcellation, particularly for cortical regions. However, white matter (WM) structural connections are also critical for supporting brain function. While prior work mapped the functional role of WM tracks during tasks (Lv, 2024), their dynamic roles at the resting state remain underexplored. In this work, we combined diffusion MRI (dMRI) and functional MRI (fMRI) to explore WM track parcellation based on the resting state activities their endpoints support. We calculated dynamic functional connectivity for each WM track (Track DFC) (Lv, 2024) with a tissue-unbiased multimodal tractogram template for 90k whole-brain WM tracks (Lv, 2023). Independent component analysis (ICA) followed by clustering helps decompose the Track DFC and group tracks into 100 functional-related subgroups. In this abstract, we focus on the default mode network (DMN) (Greicius, 2003; Raichle, 2001), which is the most active during rest and linked to various cognitive functions. Our clustering results reveal 5 DMN subnetworks among 90k WM tracts.
Methods:
We used the tissue-unbiased tractogram template with 90k streamlines (Lv, 2023) generated from minimally pre-processed T1, T2, and dMRI data of 50 randomly selected subjects from the HCP dataset (Glasser, 2013; Van Essen, 2013). This template was further registered to the MNI152 space to align with the fMRI data. DFC was calculated using minimally preprocessed resting state fMRI data from 100 unrelated HCP subjects (mean age=29.1, 54 females & 46 males). fMRI data were further spatial smoothed (FWHM=4mm) and temporal filtered (0.01-0.1Hz). Track DFC maps the time series at the two endpoints of each streamline and uses their element-wise product to reflect signal co-fluctuation (Faskowitz, 2020; Lv, 2024).
Group melodic ICA was applied to the 90k Track DFC, extracting 150 independent components (ICs), each representing a spatial pattern of correlated resting-state brain activity. This step reduced the dimensionality and identified independent patterns of DFC. We then applied K-medoids clustering to group the 90k tracks into 100 clusters based on their contributions to ICs. Among these clusters, we identified 5 clusters with more than 20% of the tracks in that cluster having both ends in the DMN defined by Yeo17 atlas (Schaefer, 2018). The DMN mask (Fig.1) was solely used to select the DMN-related clusters from the 100 parcels obtained from K-medoids clustering.
Group melodic ICA was applied to the 90k Track DFC, extracting 150 independent components (ICs), each representing a spatial pattern of correlated resting-state brain activity. This step reduced the dimensionality and identified independent patterns of DFC. We then applied K-medoids clustering to group the 90k tracks into 100 clusters based on their contributions to ICs. Among these clusters, we identified 5 clusters with more than 20% of the tracks in that cluster having both ends in the DMN defined by Yeo17 atlas (Schaefer, 2018). The DMN mask (Fig.1) was solely used to select the DMN-related clusters from the 100 parcels obtained from K-medoids clustering.
Results:
The 100 clusters for the whole-brain tractogram show a WM parcellation (Fig.2(A)). Among the 100 clusters, we identified 5 clusters with a high concentration of tracks in the DMN (Fig.2(B)). These clusters exhibit distinct spatial distributions corresponding to specific DMN subregions. The prefrontal cortex (PFC) tracts are parcellated into the medial, dorsal, and ventral PFC, represented by the purple, green, and blue clusters, respectively in Fig.2(C-E). The orange cluster predominantly focuses on the temporal lobe connections (Fig.2(F)), while the red cluster is strongly associated with the precuneus, posterior cingulate cortex and retrosplenial cortex (Fig.2(G)). These findings reveal the functional subdivision of the DMN into regions supported by distinct WM tracts, demonstrating our methods' ability to effectively capture network-specific functional features from Track DFC. This parcellation of WM tracts into functionally meaningful subnetworks enhances our understanding of the structural-functional organization of the DMN.
Conclusions:
In this study, we combined dMRI and fMRI to parcellate WM tracts, resulting in a clear subdivision of the DMN using whole brain Track DFC. Our method provides a new way to parcellate WM tracks into functionally meaningful subdivisions. In the future, we will look into more functional networks and subcortical regions.
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural) 1
fMRI Connectivity and Network Modeling
Segmentation and Parcellation
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
White Matter Anatomy, Fiber Pathways and Connectivity 2
Keywords:
FUNCTIONAL MRI
Machine Learning
NORMAL HUMAN
Tractography
White Matter
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
Other - Dynamic functional connectivity
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.
Glasser, M. F. (2013). The minimal preprocessing pipelines for the Human Connectome Project. Neuroimage, 80, 105-124.
Greicius, M. D. (2003). Functional connectivity in the resting brain: a network analysis of the default mode hypothesis. Proceedings of the National Academy of Sciences, 100(1), 253-258.
Lv, J. (2024). Mapping the Functional Role of White Matter Tracks by fusing Diffusion and Functional MRI. Proceedings of the international society for magnetic resonance in medicine.
Lv, J. (2023). Building a tissue‐unbiased brain template of fiber orientation distribution and tractography with multimodal registration. Magnetic resonance in medicine, 89(3), 1207-1220.
Raichle, M. E. (2001). A default mode of brain function. Proceedings of the National Academy of Sciences, 98(2), 676-682.
Schaefer, A. (2018). Local-global parcellation of the human cerebral cortex from intrinsic functional connectivity MRI. Cerebral cortex, 28(9), 3095-3114.
Van Essen, D. C. (2013). The WU-Minn human connectome project: an overview. Neuroimage, 80, 62-79.