Inversion of a large-scale circuit model reveals a cortical hierarchy in the dynamic resting human brain

Wang, Peng; Kong, Ru; Kong, Xiangwei; Liégeois, Raphaël; Orban, Csaba; Deco, Gustavo; Heuvel, Martijn P. van den; Yeo, B.T. Thomas

doi:10.1126/sciadv.aat7854

Cited by 226 publications

(383 citation statements)

References 44 publications

Supporting

Mentioning

361

Contrasting

Order By: Relevance

“…The results are strengthened by their consistency across species and scales, and are consistent with interareal connectivity playing a direct causal role in shaping intrinsic timescales, as recent computational models have hinted at [30,49]. However, our results are also consistent with gradients of cortical microstructure driving differences in spontaneous dynamics [18,26], with interareal connectivity simply varying along this gradient. In the absence of experiments that can legion or manipulate structural connectivity, computational modeling will play a crucial role in providing mechanistic explanations of the statistical relationships characterized here, towards a causal understanding of the mechanisms through which intrinsic timescales are shaped.…”

Section: Discussionsupporting

confidence: 88%

Timescales of spontaneous fMRI fluctuations relate to structural connectivity in the brain

Fallon

Ward

Parkes

et al. 2019

Preprint

View full text Add to dashboard Cite

2 FALLON ET AL. motifs. Our findings demonstrate a conserved property of mouse and human brain organization, in which a brain region's spontaneous activity fluctuations are closely related to their surrounding structural scaffold. K E Y W O R D S structure-function relationship, time-series analysis, structural connectivity, resting-state fMRI, interspecies comparison 1 | INTRODUCTION The brain's complex spatiotemporal dynamics unfold on an intricate web of axonal connections: the connectome [1, 2]. These pathways facilitate information transfer between brain regions, manifesting in a complex relationship between connectome structure and neural dynamics. Reflecting the pairwise nature of structural connectivity (regionregion), existing studies have overwhelmingly compared pairwise measurements of anatomical connectivity to pairwise statistical relationships between neural activity time series, or functional connectivity, often using simulations of network dynamics to better understand how the observed relationships may arise [3-18]. Structural connectivity is highly informative of functional connectivity, consistent with the connectome as a physical substrate constraining inter-regional communication dynamics.Our understanding of pairwise structure-function relationships in the brain remains disconnected from our understanding of how a brain area's structural connectivity properties shape its activity dynamics. Indeed, the structural connectivity profile of a region's incoming and outgoing axonal connections is thought to define its function [19]. Furthermore, the activity dynamics of brain areas follow a functional hierarchy, with rapid dynamics in 'lower' sensory regions, and slower fluctuations in 'higher' regions associated with integrative processes [20][21][22][23][24]. The spatial variation of intrinsic timescales has been measured using ECOG [21], MEG [25,26], TMS-EEG [27], and fMRI [24,[28][29][30][31][32][33], and may a function variation in temporal receptive windows: timescales over which new information can be actively integrated with recently received information [20,33,34]. Spatial variation in intrinsic activity fluctuations may form a key basis for the brain's functional hierarchical organization, shaped by structural variation in the brain's microcircuitry [35][36][37]. This organization is thought to be important for behavior and cognition [20,[38][39][40], and its disruption has clinical implications: e.g., differences in intrinsic timescales are associated with symptom severity in autism [33]. While much is known about the structure-function relationship at the level of brain-region pairs, and how structural and functional connectivity architecture shape cognitive function and are affected in disease [41][42][43][44], relatively little is known about how it affects the information processing dynamics of individual brain areas. In particular, we do not yet understand the role structural connections play in the cortical organization of intrinsic timescales.Recent work has provided statistical e...

show abstract

Section: Discussionsupporting

confidence: 88%

Timescales of spontaneous fMRI fluctuations relate to structural connectivity in the brain

Fallon

Ward

Parkes

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…There are a few potential reasons why DNNs did not outperform kernel regression in our experiments on RSFC-based behavioral prediction. First, while the human brain is nonlinear and hierarchically organized (Deco et al, 2011;Breakspear, 2017;Wang et al, 2019), such a structure might not be reflected in the RSFC matrix in a way that was exploitable by the DNNs we considered. This could be due to the measurements themselves (Pearson's correlations of rs-fMRI timeseries), the particular representation (N x N connectivity matrices) or particular choices of DNNs, although we again note that BrainNetCNN and GCNN were specifically developed for connectome data.…”

Section: Potential Reasons Why Dnns Did Not Outperform Kernel Regressmentioning

confidence: 99%

Deep Neural Networks and Kernel Regression Achieve Comparable Accuracies for Functional Connectivity Prediction of Behavior and Demographics

Kong

Holmes

et al. 2018

Preprint

Self Cite

116

View full text Add to dashboard Cite

There is significant interest in the development and application of deep neural networks (DNNs) to neuroimaging data. A growing literature suggests that DNNs outperform their classical counterparts in a variety of neuroimaging applications, yet there are few direct comparisons of relative utility. Here, we compared the performance of three DNN architectures and a classical machine learning algorithm (kernel regression) in predicting individual phenotypes from whole-brain resting-state functional connectivity (RSFC) patterns. One of the DNNs was a generic fully-connected feedforward neural network, while the other two DNNs were recently published approaches specifically designed to exploit the structure of connectome data. By using a combined sample of almost 10,000 participants from the Human Connectome Project (HCP) and UK Biobank, we showed that the three DNNs and kernel regression achieved similar performance across a wide range of behavioral and demographic measures. Furthermore, the generic feedforward neural network exhibited similar performance to the two state-of-the-art connectome-specific DNNs. When predicting fluid intelligence in the UK Biobank, performance of all algorithms dramatically improved when sample size increased from 100 to 1000 subjects. Improvement was smaller, but still significant, when sample size increased from 1000 to 5000 subjects. Importantly, kernel regression was competitive across all sample sizes. Overall, our study suggests that kernel regression is as effective as DNNs for RSFC-based behavioral prediction, while incurring significantly lower computational costs. Therefore, kernel regression might serve as a useful baseline algorithm for future studies.

show abstract

“…[5]). 72 Likewise, with neural mass modeling approaches, their ability to quantitatively recreate key 73 features of individual-level functional connectivity has also been limited ( [10] [7] [8]). This may 74 be because the most common approach is to parameterize connectivity from estimates of white 75 matter integrity from diffusion imaging, which also can lead to potential misinference, since these 76 connectivity estimates are constrained to be symmetric and positive ( [11]).…”

mentioning

confidence: 99%

“…Even in state-of-the-art techniques (e.g [7]. 647[8]), most parameters are fixed a priori (local neural-mass parameters), or determined from 648 diffusion imaging data, with only a limited subset taken from fMRI functional connectivity 649 estimates, and then only at the group-average level. Thus, the vast majority of parameters are not 650 sufficiently constrained by the relevant individual-level data, and instead are adapted from 651 measurement of proxy variables, which is likely to limit the accuracy of model predictions.652 Diffusion imaging data, for instance is inherently unsigned and undirected, so the resultant 653 models are unable to consider hierarchical connection schemes or long-distance connections that 654 depress activity in the post-synaptic targets.…”

mentioning

confidence: 99%

“…Although we emphasize the ability to generate individualized brain models, previous 709 studies using neural-mass models with weights seeded by diffusion imaging have emphasized 710 predicting group-level data ( [10][7] [8]). Two recent studies fit free parameters with the explicit 711 optimization objective of predicting the group-average rsFC matrix ( [7][8]). By contrast,712 MINDy seeks to predict the short-term evolution of the neural activity time series for single713 subjects, which often results in the simulated individual rsFC correlating highly with the714 empirical rsFC (Fig.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Estimation and Validation of Individualized Dynamic Brain Models with Resting State fMRI

Singh

Braver

Cole

et al. 2019

Preprint

View full text Add to dashboard Cite

20A key challenge for neuroscience is to develop generative, causal models of the human 21 nervous system in an individualized, data-driven manner. Previous initiatives have either 22 33 34 MAIN TEXT 35 36 45 Generative models are then formed by integrating these cellular-level observations with known 46 neuronal biophysics at the spatial scale of individual neurons or small populations (e.g. [1]). 47 48 In contrast, another set of large initiatives has instead focused on modeling individual 49 human brain function using an approach often referred to as "connectomics" (e.g., Human 50 Page 2 of 38 Connectome Project, [2]). This approach relies on descriptive statistics, typically correlation 51 between fluctuating activity signals in brain regions assessed during the resting state (``resting 52 state functional connectivity" or rsFC; [3]). As a result, it is sometimes difficult to make 53 mechanistic inferences based upon functional connectivity correlations ( [4]). Moreover, neural 54 processes are notoriously nonlinear and inherently dynamic, meaning that stationary descriptions, 55 such as correlation/functional connectivity, may be unable to fully capture brain mechanisms. 56 Nevertheless, rsFC remains the dominant framework for describing connectivity patterns in 57 individual human brains. 59Despite the promise of human connectomics, there have been only a few attempts to equip 60 human fMRI studies with the sorts of generative neural population models that have powered 61 insights into non-human nervous systems. Notable advances have occurred in data-driven 62 approaches, with methods being developed to identify directed, causal influences between brain 63 regions (e.g. [5]). Conversely, neural mass modeling approaches have also been extended to study 64 human brain activity in a generative fashion, and these have provided new insights into the 65 computational mechanisms underlying fMRI and MEG/EEG activity dynamics ( [6] [7] [8]). 66 Neural mass modeling approaches have also been applied to individuals by incorporating single 67 subject diffusion data and anatomy ( [6]), but, unlike (linear) data-driven approaches (e.g. 68 Dynamic Causal Modeling; [5]), these models have not been directly fit to the individual brain 69 activity dynamics they seek to capture. However, both of these approaches have important 70 limitations. In particular, the data-driven approaches are subject to potential misinferences due to 71 assumptions of linearity ( [9]) or limitation to a relatively small number of brain regions (e.g. [5]). 72 Likewise, with neural mass modeling approaches, their ability to quantitatively recreate key 73 features of individual-level functional connectivity has also been limited ( [10] [7] [8]). This may 74 be because the most common approach is to parameterize connectivity from estimates of white 75 matter integrity from diffusion imaging, which also can lead to potential misinference, since these 76 connectivity estimates are constrained to be symmetric and positive ( [11]). ...

show abstract

Inversion of a large-scale circuit model reveals a cortical hierarchy in the dynamic resting human brain

Abstract: Converging evidence from biophysical modeling, magnetic resonance imaging, and histology reveals a large-scale cortical gradient.

Cited by 226 publications

References 44 publications

Timescales of spontaneous fMRI fluctuations relate to structural connectivity in the brain

Timescales of spontaneous fMRI fluctuations relate to structural connectivity in the brain

Deep Neural Networks and Kernel Regression Achieve Comparable Accuracies for Functional Connectivity Prediction of Behavior and Demographics

Estimation and Validation of Individualized Dynamic Brain Models with Resting State fMRI

Contact Info

Product

Resources

About