Learning to Model the Relationship Between Brain Structural and Functional Connectomes

Li, Yang; Mateos, Gonzalo González; Zhang, Zhengwu

doi:10.48550/arxiv.2112.09906

Cited by 2 publications

(2 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Structural and functional connectomes have further been jointly modeled using independent component analysis (ICA) combining structural and functional connections as features for the ICA (Amico and Goñi, 2018 ). Recently, deep learning autoencoders (Banka et al, 2020 ) and graph neural networks (see also Bessadok et al, 2021 for an overview) have been proposed for multimodal integration providing non-linear mappings between structural and functional connectivity (Li et al, 2021 ) as well as joint learning of connectivity fingerprints predictive of phenotypic traits (Filip et al, 2020 ; Dsouza et al, 2021 ). In the context of connectivity based parcellations the most prominent approaches have been to use k-means, hierarchical, or spectral clustering (Eickhoff et al, 2015 ; Reuter et al, 2020 ) to parcellate functional and structural connectivity.…”

Section: Introductionmentioning

confidence: 99%

Uncovering Cortical Units of Processing From Multi-Layered Connectomes

et al. 2022

View full text Add to dashboard Cite

Modern diffusion and functional magnetic resonance imaging (dMRI/fMRI) provide non-invasive high-resolution images from which multi-layered networks of whole-brain structural and functional connectivity can be derived. Unfortunately, the lack of observed correspondence between the connectivity profiles of the two modalities challenges the understanding of the relationship between the functional and structural connectome. Rather than focusing on correspondence at the level of connections we presently investigate correspondence in terms of modular organization according to shared canonical processing units. We use a stochastic block-model (SBM) as a data-driven approach for clustering high-resolution multi-layer whole-brain connectivity networks and use prediction to quantify the extent to which a given clustering accounts for the connectome within a modality. The employed SBM assumes a single underlying parcellation exists across modalities whilst permitting each modality to possess an independent connectivity structure between parcels thereby imposing concurrent functional and structural units but different structural and functional connectivity profiles. We contrast the joint processing units to their modality specific counterparts and find that even though data-driven structural and functional parcellations exhibit substantial differences, attributed to modality specific biases, the joint model is able to achieve a consensus representation that well accounts for both the functional and structural connectome providing improved representations of functional connectivity compared to using functional data alone. This implies that a representation persists in the consensus model that is shared by the individual modalities. We find additional support for this viewpoint when the anatomical correspondence between modalities is removed from the joint modeling. The resultant drop in predictive performance is in general substantial, confirming that the anatomical correspondence of processing units is indeed present between the two modalities. Our findings illustrate how multi-modal integration admits consensus representations well-characterizing each individual modality despite their biases and points to the importance of multi-layered connectomes as providing supplementary information regarding the brain's canonical processing units.

show abstract

Section: Introductionmentioning

confidence: 99%

Uncovering Cortical Units of Processing From Multi-Layered Connectomes

et al. 2022

View full text Add to dashboard Cite

show abstract

“…That cognition is supported by large-scale interactions between different brain regions [4] suggests that the connectomics approach, the study of anatomical and functional connections across the whole brain [4, 5], may shed new light on the elusive brain structure–and–function relationship. Several models have been proposed to bridge the link between brain function and connectome, including statistical [6, 7, 8, 9, 10, 11, 12], network communication [13, 14], and machine learning model [15, 16, 17, 18, 19, 20].…”

Section: Introductionmentioning

confidence: 99%

Predicting brain functions from structural connectome using graph neural network

Hui

Sun

et al. 2022

Preprint

View full text Add to dashboard Cite

The relationship between brain structure and function remains elusive, amidst the tremendous advances in brain mapping techniques. In this work, we attempt to partially disentangle this relationship by connecting task--evoked functional MRI responses with the underlying structural connections using graph neural network (GNN). We have successfully demonstrated that our GNN--based model could predict brain functions with good fidelity, and that it is possible to infer the subset of neighboring cortical regions whose structural connections are important for the prediction of the functions of individual cortical regions using a graph attention mechanism. In particular, this subset of neighboring cortical regions is much smaller than the underlying structural connectome, and that the majority of these regions were ipsilateral to the cortical regions whose functions our GNN--based model predicts. We found that the higher cognitive functions subserved by the cingulo--opercular, dorsal attention, frontoparietal and default mode clusters may depend on neighboring cortical regions across a wide range of functional brain clusters in the ipsilateral hemisphere, whilst the sensory functions subserved by the visual1 and auditory clusters on cortical regions across much fewer functional brain clusters.

show abstract

Learning to Model the Relationship Between Brain Structural and Functional Connectomes

Cited by 2 publications

References 45 publications

Uncovering Cortical Units of Processing From Multi-Layered Connectomes

Uncovering Cortical Units of Processing From Multi-Layered Connectomes

Predicting brain functions from structural connectome using graph neural network

Contact Info

Product

Resources

About