Structurally-informed Bayesian functional connectivity analysis

Hinne, Max; Ambrogioni, Luca; Janssen, Ronald J.; Heskes, Tom; Gerven, Marcel A. J. van

doi:10.1016/j.neuroimage.2013.09.075

Cited by 52 publications

(51 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Generative models of neuroimaging (Friston et al, 2003;Harrison et al, 2015;Havlicek et al, 2017;Hinne et al, 2014;Langs et al, 2014) or behavioral (Behrens et al, 2007;Friston et al, 2017;Mathys et al, 2014) data have become important pillars of computational and cognitive neuroscience. This type of analysis has the advantage of inferring putative mechanisms underlying neurophysiological and cognitive processes from neuroimaging and behavioral measurements.…”

Section: Introductionmentioning

confidence: 99%

Variational Bayesian inversion for hierarchical unsupervised generative embedding (HUGE)

et al. 2018

View full text Add to dashboard Cite

A B S T R A C TA recently introduced hierarchical generative model unified the inference of effective connectivity in individual subjects and the unsupervised identification of subgroups defined by connectivity patterns. This hierarchical unsupervised generative embedding (HUGE) approach combined a hierarchical formulation of dynamic causal modelling (DCM) for fMRI with Gaussian mixture models and relied on Markov chain Monte Carlo (MCMC) sampling for inference. While well suited for the inversion of complex hierarchical models, MCMC-based sampling suffers from a computational burden that is prohibitive for many applications.To address this problem, this paper derives an efficient variational Bayesian (VB) inversion scheme for HUGE that simultaneously provides approximations to the posterior distribution over model parameters and to the log model evidence. The face validity of the VB scheme was tested using two synthetic fMRI datasets with known ground truth. Additionally, an empirical fMRI dataset of stroke patients and healthy controls was used to evaluate the practical utility of the method in application to real-world problems.Our analyses demonstrate good performance of our VB scheme, with a marked speed-up of model inversion by two orders of magnitude compared to MCMC, while maintaining a similar level of accuracy. Notably, additional acceleration would be possible if parallel computing techniques were applied. Generally, our VB implementation of HUGE is fast enough to support multi-start procedures for whole-group analyses, a useful strategy to ameliorate problems with local extrema. HUGE thus represents a potentially useful practical solution for an important problem in clinical neuromodeling and computational psychiatry, i.e., the unsupervised detection of subgroups in heterogeneous populations that are defined by effective connectivity.

show abstract

Section: Introductionmentioning

confidence: 99%

Variational Bayesian inversion for hierarchical unsupervised generative embedding (HUGE)

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Various methods have been proposed (for a comprehensive review, see Karahanoglu and Van De Ville, 2017), ranging from conventional correlation or coherence analyses which assume stationarity (Fox et al, 2005) to sliding-window correlation analyses that can capture dynamic fluctuations in functional connectivity (Chang and Glover, 2010 NeuroImage 179 (2018) 505-529 coupled) Hidden Markov models (HMM;Bolton et al, 2018;Karahanoglu and Van De Ville, 2015;Vidaurre et al, 2017), and approaches from statistical mechanics that rely on entropy maximization (Ashourvan et al, 2017). Other functional connectivity methods have focused on sparsity (Bielczyk et al, 2018;Eavani et al, 2015;Ryali et al, 2012), including generative models that can exploit anatomical information (Hinne et al, 2014). While functional connectivity affords valuable insights into the dynamics of brain networks both in health and disease (Buckner et al, 2013;Bullmore and Sporns, 2009;Fornito et al, 2015), it provides undirected measures of coupling (Friston, 2011), without an explicit model of the system of interest (Stephan, 2004).…”

mentioning

confidence: 99%

A generative model of whole-brain effective connectivity

et al. 2018

View full text Add to dashboard Cite

A B S T R A C TThe development of whole-brain models that can infer effective (directed) connection strengths from fMRI data represents a central challenge for computational neuroimaging. A recently introduced generative model of fMRI data, regression dynamic causal modeling (rDCM), moves towards this goal as it scales gracefully to very large networks. However, large-scale networks with thousands of connections are difficult to interpret; additionally, one typically lacks information (data points per free parameter) for precise estimation of all model parameters.This paper introduces sparsity constraints to the variational Bayesian framework of rDCM as a solution to these problems in the domain of task-based fMRI. This sparse rDCM approach enables highly efficient effective connectivity analyses in whole-brain networks and does not require a priori assumptions about the network's connectivity structure but prunes fully (all-to-all) connected networks as part of model inversion. Following the derivation of the variational Bayesian update equations for sparse rDCM, we use both simulated and empirical data to assess the face validity of the model. In particular, we show that it is feasible to infer effective connection strengths from fMRI data using a network with more than 100 regions and 10,000 connections. This demonstrates the feasibility of whole-brain inference on effective connectivity from fMRI data -in single subjects and with a run-time below 1 min when using parallelized code. We anticipate that sparse rDCM may find useful application in connectomics and clinical neuromodeling -for example, for phenotyping individual patients in terms of wholebrain network structure. IntroductionThe human brain comprises multiple levels of organization, with cognitive functions arising from the interplay of functional specialization and integration (Sporns, 2013;Tononi et al., 1994). With the advent of non-invasive neuroimaging techniques, such as functional magnetic resonance imaging (fMRI), researchers have begun to systematically study these fundamental properties of human brain organization. While early neuroimaging studies focused on localizing cognitive processes to specific brain regions, contemporary studies are commonly concerned with functional integration of these regions; this requires analyzing brain connectivity (Smith, 2012). In addition to analyses of structural (anatomical) connections, assessments of functional connectivity (statistical dependencies between network nodes) play a major role. Particularly, functional connectivity has been used frequently for studying the functional organization of large (whole-brain) networks both in tasks and in the "resting state" (i.e., unconstrained cognition in the absence of external perturbations). Various methods have been proposed (for a comprehensive review, see Karahanoglu and Van De Ville, 2017), ranging from conventional correlation or coherence analyses which assume stationarity (Fox et al., 2005) to sliding-window correlation analyses that can capture dyna...

show abstract

“…Since the true precision matrix is dense for the ICA generated data, we consider two alternate measures of performance under this scenario—the inverse error (Padmanabhan et al, 2016) and Kullback-Leibler divergence (Hinne et al, 2014). The inverse error captures the accuracy of the estimated precision matrix, and is defined as || Ω − 1 Ωˆ − I ||F where Ωˆ is the estimated precision matrix, I is a p×p identity matrix, and ||.|| F is the frobenius norm.…”

Section: Resultsmentioning

confidence: 99%

“…They provide measures of functional co-activation deviating from standard measures of FC such as Pearson or partial correlations. Hinne et al (2014) proposed a Bayesian approach which uses fMRI data to model the distribution of partial correlations for edges determined by the given SC information. The assumption that FC only exists between anatomically connected regions ignores the contributions of indirect anatomical pathways and does not capture the complexity of the relationship between brain structure and function (Honey et al, 2009, 2010; Messé et al, 2014).…”

Section: Introductionmentioning

confidence: 99%

Integrative Bayesian analysis of brain functional networks incorporating anatomical knowledge

2018

View full text Add to dashboard Cite

Recently, there has been increased interest in fusing multimodal imaging to better understand brain organization by integrating information on both brain structure and function. In particular, incorporating anatomical knowledge leads to desirable outcomes such as increased accuracy in brain network estimates and greater reproducibility of topological features across scanning sessions. Despite the clear advantages, major challenges persist in integrative analyses including an incomplete understanding of the structure-function relationship and inaccuracies in mapping anatomical structures due to inherent deficiencies in existing imaging technology. This calls for the development of advanced network modeling tools that appropriately incorporate anatomical structure in constructing brain functional networks. We propose a hierarchical Bayesian Gaussian graphical modeling approach which models the brain functional networks via sparse precision matrices whose degree of edge specific shrinkage is a random variable that is modeled using both anatomical structure and an independent baseline component. The proposed approach adaptively shrinks functional connections and flexibly identifies functional connections supported by structural connectivity knowledge. This enables robust brain network estimation even in the presence of misspecified anatomical knowledge, while accommodating heterogeneity in the structure-function relationship. We implement the approach via an efficient optimization algorithm which yields maximum a posteriori estimates. Extensive numerical studies involving multiple functional network structures reveal the clear advantages of the proposed approach over competing methods in accurately estimating brain functional connectivity, even when the anatomical knowledge is misspecified up to a certain degree. An application of the approach to data from the Philadelphia Neurodevelopmental Cohort (PNC) study reveals gender based connectivity differences across multiple age groups, and higher reproducibility in the estimation of network metrics compared to alternative methods.

show abstract

Structurally-informed Bayesian functional connectivity analysis

Cited by 52 publications

References 63 publications

Variational Bayesian inversion for hierarchical unsupervised generative embedding (HUGE)

Variational Bayesian inversion for hierarchical unsupervised generative embedding (HUGE)

A generative model of whole-brain effective connectivity

Integrative Bayesian analysis of brain functional networks incorporating anatomical knowledge

Contact Info

Product

Resources

About