What is the role of gradients in the prediction of functional connectivity?

Supervised learning experiments indicated that structural gradients can serve as sensitive, low-dimensional predictors of the functional organization of the temporal lobe.

What is the significance of the gradients in predicting functional connectivity?

structural connectivity gradients also related to established markers of functional reconfiguration and areal expansion between humans and non-human primates, supporting the potential of connectome gradients in shaping evolutionary changes.

How were the qT1 values interpolated to the surface?

qT1 values were interpolated to the surface by taking the average of seven trilinear interpolations evenly interspersed between the 20th and 80th percentile distances from the pial to white matter surfaces using Freesurfer’s mri_vol2surf command.a) HCP.

What is the morphological significance of the temporal lobe?

The temporal lobe hosts a diverse array of functional processes implicated in sensory processing, memory, and language abilities, and is among the macroscopic structures most frequently compromised in neurological and neuropsychiatric disorders, including Alzheimer’s disease (Braak & Braak, 1991) and drug-resistant epilepsy (Dam, 1982).

What was the important factor in the computation of connectivity distance?

Connectivity distance, a measure that characterizes the relationship between physical distance and connectivity (Larivière et al., 2020), was computed by thresholding the structural connectivity matrix at the 80th percentile, multiplying each connection by the geodesic distance between their nodes, and averaging all connections for each node.

What is the common method used to describe the largest axes of variance in connectivity?

To describe the largest axes of variance in connectivity the authors used diffusion map embedding (Coifman & Lafon, 2006), a non-linear dimensionality reduction techniques technique used previously to identify neocortical, hippocampal, and cerebellar functional gradients (Guell et al., 2018; Margulies et al., 2016; Vos de Wael et al., 2018).

How was the warp to the freesurfer image calculated?

A warp to the Freesurfer T1w image was computed by averaging volumetric timeseries across the time dimension and registering this image using boundary-based registration.

What is the purpose of this study?

To provide a comprehensive account of its substructural organization in humans, their study harnessed manifold learning operating on high resolution diffusion MRI tractography data of the temporal lobe to identify separate, yet partially overlapping axes of its structural connectome embedding.

How was the gradient algorithm implemented in BrainSpace?

the authors generated random datasets with equivalent spatial autocorrelation as the response variable using Moran spectral randomization with the singleton procedure (Wagner & Dray, 2015) as implemented in BrainSpace (Vos de Wael et al., 2020).

What is the relationship between the sulco-gyral fold and the temporal?

The relationship with cortical thickness and curvature was weaker, suggesting that their cortical connectivity gradients closely reflect intracortical factors and only to a lesser extent macroscopic morphological variations and/or potential biases from sulco-gyral folding (Schilling et al., 2018).

What is the description of the temporal lobe?

Recapitulating prior work, the authors found that the temporal lobe is best described by three gradients (Bajada et al., 2019; Blazquez Freches et al., 2020), spanning medio-lateral (G1), anterior-posterior (G2) and anterolateralposteromedial (G3) axes.

How did the authors assess the relationship between structural gradients and edgewise functional connectivity?

To further assess the relationship between structural gradients and edgewise functional connectivity, the authors used a decision tree with binary splits for regression.

Open AccessPosted Content10.1101/2020.11.26.400382

Microstructural underpinnings and macroscale functional implications of temporal lobe connectivity gradients

Reinder Vos de Wael, +11 more

- 27 Nov 2020

- bioRxiv

TL;DR: These results provide robust evidence for three axes of structural connectivity in human temporal cortex with consistent microstructural underpinnings and contributions to large-scale brain network function.

Abstract: The temporal lobe is implicated in higher cognitive processes and is one of the regions that underwent substantial reorganization during primate evolution. Its functions are instantiated, in part, by its complex layout of structural connections. This study identified low-dimensional representations of structural connectivity variations in human temporal cortex and explored their microstructural underpinnings and associations to macroscale function. We identified three eigenmodes which described gradients in structural connectivity. These gradients reflected inter-regional variations in cortical microstructure derived from quantitative MRI and post-mortem histology. Gradient-informed models accurately predicted macroscale measures of temporal lobe function. Gradients aligned closely with established measures of functional reconfiguration and areal expansion between macaques and humans, highlighting the important role evolution has played in shaping temporal lobe function. Our results provide robust evidence for three axes of structural connectivity in human temporal cortex with consistent microstructural underpinnings and contributions to large-scale brain network function.

AI Agents for this Paper

Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps

Most frequently asked questions

1. What have the authors contributed in "Microstructural underpinnings and macroscale functional implications of temporal lobe connectivity gradients" ?

Vos de Wael et al. this paper proposed a method to detect the brain tumor using a brain MRI.

2. What was used to test for associations between the structural gradients?

Moran spectral randomization (Wagner & Dray, 2015) was used to test for associations between BigBrain gradient 1 and the structural gradients.

3. How was the HCP-Discovery split into five folds?

HCP-Discovery was split into five folds of 15 subjects each; for each fold the authors performed a multinomial logistic regression with the first three gradients as predictor variables and networks as outcome variables.

4. What is the method to model temporal lobe connectivity?

By propagating each streamline to cortical surface points, rather than to macroscopic parcels, the authors were able to resolve fine grained changes in temporal lobe connectivity and thus account for heterogeneity of subregional connectivity.

Fig. 2 Microstructural basis of the structural gradients. (A) We tested for linear relations between eccentricity and curvature, cortical thickness, as well as T1w/T2w intensity. Stars denote significant correlations. (B) We tested for an association between microstructural profile covariance derived previously from the BigBrain atlas (Amunts et al., 2013; Paquola et al., 2019), and eccentricity. Eccentricity was projected to the same parcellation scheme as microstructural profile covariance by taking the mean within each parcel. Next, we evaluated the association between gradients and cortical cytoarchitecture (Error! Reference source not found.B), capitalizing on the BigBrain dataset, an ultra-high resolution 3D histological reconstruction of a post-mortem human brain (Amunts et al., 2013). We adopted a previously established approach to identify cytoarchitectural gradients (Paquola et al., 2019) and compared the principal cytoarchitectural gradient, which runs from primary sensory to limbic

Fig 4. Relationship between the structural gradients and phylogenetic markers. Both functional homology and areal expansion are significantly associated with eccentricity of the first three structural gradients.

Fig. 3 Functional markers of the structural gradients. (A) Based on a canonical network parcellation (Yeo et al., 2011), we attempted to predict the functional networks. Performance was high with a Cohen’s kappa of 0.77+0.01(left) and 0.81+0.01 (right). Predicted networks shown here are the results of one of the five folds. (B) Histograms of the prediction accuracy per subject, as measured by the Pearson correlation between empirical and predicted data, of a decision tree regression estimating functional connectivity from structural gradients. HCPDiscovery predictions were trained with a 5-fold crossvalidation, the predictions of the other datasets were trained on HCP-Discovery (C) Pearson’s correlation between the predicted and empirical functional connectivity for every vertex across subjects. Predictions were especially accurate in lateral temporal regions, and less robust in the medial temporal lobe.

Fig. 1 Diffusion embedding of structural connectivity. (A) Streamlines were generated throughout the entire brain and systematically mapped to the cortical surface using nearest neighbor interpolation. We computed the affinity matrix of the connectivity matrix using a cosine similarity kernel and constructed gradients of structural connectivity of the temporal lobe to ipsilateral hemisphere with diffusion embedding. The first three gradients of diffusion embedding, sorted by variance explained of each vector, were selected for further analyses. Eccentricity in this manifold space was high in posterior medial temporal lobe and the superior temporal gyrus. (B) The relationship of eccentricity with connectivity distance (left) and degree centrality (right). The star denotes a significant correlation.

Citations

The International Society for Magnetic Resonance in Medicine (ISMRM)

Paul Kennedy, +14 more

- 01 Jan 2012

•Journal Article

Evolutionary specializations of human association cortex

Rogier B. Mars, +4 more

- 01 Jan 2017

- Evolution of Nervous Systems

TL;DR: In this article, the human association cortex is larger than would be expected for an equivalent ape brain, suggesting that the human brain adaptation is a unique adaptation, and the internal organization of the human Association cortex shows modifications of the ape plan in terms of expanded or novel areas and connections.

...read moreread less

•Journal Article•10.1016/j.cortex.2022.01.019

Individual differences in gradients of intrinsic connectivity within the semantic network relate to distinct aspects of semantic cognition

Ximing Shao, +8 more

- 01 Feb 2022

- Cortex

TL;DR: In this paper , a nonlinear dimensionality reduction technique was applied to restingstate functional magnetic resonance imaging from 176 participants to characterise the strength of two key connectivity gradients in each individual: the principal gradient captured the separation between unimodal and heteromodal cortex, while the second gradient corresponded to the distinction between motor and visual cortex.

...read moreread less

References

Journal Article•10.1007/BF00308809

Neuropathological stageing of Alzheimer-related changes.

Heiko Braak, +1 more

- 01 Jan 1991

- Acta Neuropathologica

TL;DR: The investigation showed that recognition of the six stages required qualitative evaluation of only a few key preparations, permitting the differentiation of six stages.

...read moreread less

15.7K

Journal Article•10.1016/J.NEUROIMAGE.2006.01.021

An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest.

Rahul S. Desikan, +13 more

- 01 Jul 2006

- NeuroImage

TL;DR: An automated labeling system for subdividing the human cerebral cortex into standard gyral-based neuroanatomical regions is both anatomically valid and reliable and may be useful for both morphometric and functional studies of the cerebral cortex.

...read moreread less

12.7K

Journal Article•10.1006/NIMG.1998.0395

Cortical surface-based analysis. I. Segmentation and surface reconstruction

Anders M. Dale, +2 more

- 01 Feb 1999

- NeuroImage

TL;DR: A set of automated procedures for obtaining accurate reconstructions of the cortical surface are described, which have been applied to data from more than 100 subjects, requiring little or no manual intervention.

...read moreread less

11.1K

•Journal Article•10.1371/JOURNAL.PMED.0020124

Why Most Published Research Findings Are False

John P. A. Ioannidis

- 30 Aug 2005

- PLOS Medicine

TL;DR: Simulations show that for most study designs and settings, it is more likely for a research claim to be false than true.

...read moreread less

9.7K

...

Expand

Microstructural underpinnings and macroscale functional implications of temporal lobe connectivity gradients

Chat with Paper

AI Agents for this Paper

Most frequently asked questions

1. What have the authors contributed in "Microstructural underpinnings and macroscale functional implications of temporal lobe connectivity gradients" ?

2. What was used to test for associations between the structural gradients?

3. How was the HCP-Discovery split into five folds?

4. What is the method to model temporal lobe connectivity?

5. What is the role of gradients in the prediction of functional connectivity?

6. What is the significance of the gradients in predicting functional connectivity?

7. How were the qT1 values interpolated to the surface?

8. What is the morphological significance of the temporal lobe?

9. What was the important factor in the computation of connectivity distance?

10. What is the common method used to describe the largest axes of variance in connectivity?

11. How was the warp to the freesurfer image calculated?

12. What is the purpose of this study?

13. How was the gradient algorithm implemented in BrainSpace?

14. What is the relationship between the sulco-gyral fold and the temporal?

15. What is the description of the temporal lobe?

16. How did the authors assess the relationship between structural gradients and edgewise functional connectivity?

Figures

Citations

The International Society for Magnetic Resonance in Medicine (ISMRM)

Evolutionary specializations of human association cortex

Individual differences in gradients of intrinsic connectivity within the semantic network relate to distinct aspects of semantic cognition

References

Neuropathological stageing of Alzheimer-related changes.

Advances in functional and structural MR image analysis and implementation as FSL.

An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest.

Cortical surface-based analysis. I. Segmentation and surface reconstruction

Why Most Published Research Findings Are False

Related Papers (5)

Principles of temporal association cortex organisation as revealed by connectivity gradients

Topographic gradients of intrinsic dynamics across neocortex.

Microstructural and functional gradients are increasingly dissociated in transmodal cortices

BrainSpace: a toolbox for the analysis of macroscale gradients in neuroimaging and connectomics datasets

Spectral mapping of brain functional connectivity from diffusion imaging.