Showing papers by "Bertrand Thirion published in 2016"

Exploring the Early Organization and Maturation of Linguistic Pathways in the Human Infant Brain

Abstract: Linguistic processing is based on a close collaboration between temporal and frontal regions connected by two pathways: the “dorsal” and “ventral pathways” (assumed to support phonological and semantic processing, respectively, in adults). We investigated here the development of these pathways at the onset of language acquisition, during the first post-natal weeks, using cross-sectional diffusion imaging in 21 healthy infants (6–22 weeks of age) and 17 young adults. We compared the bundle organization and microstructure at these two ages using tractography and original clustering analyses of diffusion tensor imaging parameters. We observed structural similarities between both groups, especially concerning the dorsal/ventral pathwaysegregation and the arcuate fasciculus asymmetry. We further highlighted the developmental tempos of the linguistic bundles: The ventral pathway maturation was more advanced than the dorsal pathway maturation, but the latter catches up during the first post-natal months. Its fast development during this period might relate to the learning of speech cross-modal representations and to the first combinatorial analyses of the speech input.

Formal Models of the Network Co-occurrence Underlying Mental Operations.

Abstract: Systems neuroscience has identified a set of canonical large-scale networks in humans. These have predominantly been characterized by resting-state analyses of the task-unconstrained, mind-wandering brain. Their explicit relationship to defined task performance is largely unknown and remains challenging. The present work contributes a multivariate statistical learning approach that can extract the major brain networks and quantify their configuration during various psychological tasks. The method is validated in two extensive datasets (n = 500 and n = 81) by model-based generation of synthetic activity maps from recombination of shared network topographies. To study a use case, we formally revisited the poorly understood difference between neural activity underlying idling versus goal-directed behavior. We demonstrate that task-specific neural activity patterns can be explained by plausible combinations of resting-state networks. The possibility of decomposing a mental task into the relative contributions of major brain networks, the "network co-occurrence architecture" of a given task, opens an alternative access to the neural substrates of human cognition.

Best Practices in Data Analysis and Sharing in Neuroimaging using MRI

Abstract: Neuroimaging enables rich noninvasive measurements of human brain activity, but translating such data into neuroscientific insights and clinical applications requires complex analyses and collaboration among a diverse array of researchers. The open science movement is reshaping scientific culture and addressing the challenges of transparency and reproducibility of research. To advance open science in neuroimaging the Organization for Human Brain Mapping created the Committee on Best Practice in Data Analysis and Sharing (COBIDAS), charged with creating a report that collects best practice recommendations from experts and the entire brain imaging community. The purpose of this work is to elaborate the principles of open and reproducible research for neuroimaging using Magnetic Resonance Imaging (MRI), and then distill these principles to specific research practices. Many elements of a study are so varied that practice cannot be prescribed, but for these areas we detail the information that must be reported to fully understand and potentially replicate a study. For other elements of a study, like statistical modelling where specific poor practices can be identified, and the emerging areas of data sharing and reproducibility, we detail both good practice and reporting standards. For each of seven areas of a study we provide tabular listing of over 100 items to help plan, execute, report and share research in the most transparent fashion. Whether for individual scientists, or for editors and reviewers, we hope these guidelines serve as a benchmark, to raise the standards of practice and reporting in neuroimaging using MRI.

Dictionary learning for massive matrix factorization

Abstract: Sparse matrix factorization is a popular tool to obtain interpretable data decompositions, which are also effective to perform data completion or denoising. Its applicability to large datasets has been addressed with online and randomized methods, that reduce the complexity in one of the matrix dimension, but not in both of them. In this paper, we tackle very large matrices in both dimensions. We propose a new factorization method that scales gracefully to terabyte-scale datasets. Those could not be processed by previous algorithms in a reasonable amount of time. We demonstrate the efficiency of our approach on massive functional Magnetic Resonance Imaging (fMRI) data, and on matrix completion problems for recommender systems, where we obtain significant speed-ups compared to state-of-the art coordinate descent methods.

Dictionary Learning for Massive Matrix Factorization

Abstract: Sparse matrix factorization is a popular tool to obtain interpretable data decompositions, which are also effective to perform data completion or denoising. Its applicability to large datasets has been addressed with online and randomized methods, that reduce the complexity in one of the matrix dimension, but not in both of them. In this paper, we tackle very large matrices in both dimensions. We propose a new factoriza-tion method that scales gracefully to terabyte-scale datasets, that could not be processed by previous algorithms in a reasonable amount of time. We demonstrate the efficiency of our approach on massive functional Magnetic Resonance Imaging (fMRI) data, and on matrix completion problems for recommender systems, where we obtain significant speed-ups compared to state-of-the art coordinate descent methods.

Identification of Mood-Relevant Brain Connections Using a Continuous, Subject-Driven Rumination Paradigm

Abstract: Rumination, an internal cognitive state characterized by recursive thinking of current self-distress and past negative events, has been found to correlate with the development of depressive disorders. Here, we investigated the feasibility of using connectivity for distinguishing different emotional states induced by a novel free-streaming, subject-driven experimental paradigm. Connectivity between 78 functional regions of interest (ROIs) within 14 large-scale networks and 6 structural ROIs particularly relevant to emotional processing were used for classifying 4 mental states in 19 healthy controls. The 4 mental states comprised: An unconstrained period of mind wandering; a ruminative mental state self-induced by recalling a time of personal disappointment; a euphoric mental state self-induced by recalling what brings the subject joy; and a sequential episodic recollection of the events of the day. A support vector machine achieved accuracies ranging from 89% to 94% in classifying pairs of different mental states. We reported the most significant brain connections that best discriminated these mental states. In particular, connectivity changes involving the amygdala were found to be important for distinguishing the rumination condition from the other mental states. Our results demonstrated that connectivity-based classification of subject-driven emotional states constitutes a novel and effective approach for studying ruminative behavior.

Learning brain regions via large-scale online structured sparse dictionary-learning

Abstract: We propose a multivariate online dictionary-learning method for obtaining decompositions of brain images with structured and sparse components (aka atoms). Sparsity is to be understood in the usual sense: the dictionary atoms are constrained to contain mostly zeros. This is imposed via an $\ell_1$-norm constraint. By "structured", we mean that the atoms are piece-wise smooth and compact, thus making up blobs, as opposed to scattered patterns of activation. We propose to use a Sobolev (Laplacian) penalty to impose this type of structure. Combining the two penalties, we obtain decompositions that properly delineate brain structures from functional images. This non-trivially extends the online dictionary-learning work of Mairal et al. (2010), at the price of only a factor of 2 or 3 on the overall running time. Just like the Mairal et al. (2010) reference method, the online nature of our proposed algorithm allows it to scale to arbitrarily sized datasets. Experiments on brain data show that our proposed method extracts structured and denoised dictionaries that are more intepretable and better capture inter-subject variability in small medium, and large-scale regimes alike, compared to state-of-the-art models.

Transmodal Learning of Functional Networks for Alzheimer's Disease Prediction

Abstract: Functional connectivity describes neural activity from resting-state functional magnetic resonance imaging (rs-fMRI). This noninvasive modality is a promising imaging biomark-er of neurodegenerative diseases, such as Alzheimer's disease (AD), where the connectome can be an indicator to assess and to understand the pathology. However, it only provides noisy measurements of brain activity. As a consequence, it has shown fairly limited discrimination power on clinical groups. So far, the reference functional marker of AD is the fluorodeoxyglucose positron emission tomography (FDG-PET). It gives a reliable quantification of metabolic activity, but it is costly and invasive. Here, our goal is to analyze AD populations solely based on rs-fMRI, as functional connectivity is correlated to metabolism. We introduce transmodal learning : leveraging a prior from one modality to improve results of another modality on different subjects. A metabolic prior is learned from an independent FDG-PET dataset to improve functional connectivity-based prediction of AD. The prior acts as a regularization of connectivity learning and improves the estimation of discriminative patterns from distinct rs-fMRI datasets. Our approach is a two-stage classification strategy that combines several seed-based connectivity maps to cover a large number of functional networks that identify AD physiopathology. Experimental results show that our transmodal approach increases classification accuracy compared to pure rs-fMRI approaches, without resorting to additional invasive acquisitions. The method successfully recovers brain regions known to be impacted by the disease.

Transport on Riemannian Manifold for Connectivity-Based Brain Decoding

Abstract: There is a recent interest in using functional magnetic resonance imaging (fMRI) for decoding more naturalistic, cognitive states, in which subjects perform various tasks in a continuous, self-directed manner. In this setting, the set of brain volumes over the entire task duration is usually taken as a single sample with connectivity estimates, such as Pearson's correlation, employed as features. Since covariance matrices live on the positive semidefinite cone, their elements are inherently inter-related. The assumption of uncorrelated features implicit in most classifier learning algorithms is thus violated. Coupled with the usual small sample sizes, the generalizability of the learned classifiers is limited, and the identification of significant brain connections from the classifier weights is nontrivial. In this paper, we present a Riemannian approach for connectivity-based brain decoding. The core idea is to project the covariance estimates onto a common tangent space to reduce the statistical dependencies between their elements. For this, we propose a matrix whitening transport, and compare it against parallel transport implemented via the Schild's ladder algorithm. To validate our classification approach, we apply it to fMRI data acquired from twenty four subjects during four continuous, self-driven tasks. We show that our approach provides significantly higher classification accuracy than directly using Pearson's correlation and its regularized variants as features. To facilitate result interpretation, we further propose a non-parametric scheme that combines bootstrapping and permutation testing for identifying significantly discriminative brain connections from the classifier weights. Using this scheme, a number of neuro-anatomically meaningful connections are detected, whereas no significant connections are found with pure permutation testing.

Compressed online dictionary learning for fast resting-state fMRI decomposition

Abstract: We present a method for fast resting-state fMRI spatial decompositions of very large datasets, based on the reduction of the temporal dimension before applying dictionary learning on concatenated individual records from groups of subjects. Introducing a measure of correspondence between spatial decompositions of rest fMRI, we demonstrates that time-reduced dictionary learning produces result as reliable as non-reduced decompositions. We also show that this reduction significantly improves computational scalability.

Compressed Online Dictionary Learning for Fast fMRI Decomposition

Abstract: We present a method for fast resting-state fMRI spatial decomposi-tions of very large datasets, based on the reduction of the temporal dimension before applying dictionary learning on concatenated individual records from groups of subjects. Introducing a measure of correspondence between spatial decompositions of rest fMRI, we demonstrates that time-reduced dictionary learning produces result as reliable as non-reduced decompositions. We also show that this reduction significantly improves computational scalability.

Deriving robust biomarkers from multi-site resting-state data: An Autism-based example

Abstract: Resting-state functional Magnetic Resonance Imaging (R-fMRI) holds the promise to reveal functional biomarkers of neuropsychiatric disorders. However, extracting such biomarkers is challenging for complex multi-faceted neuropathologies, such as autism spectrum disorders. Large multi-site datasets increase sample sizes to compensate for this complexity, at the cost of uncontrolled heterogeneity. This heterogeneity raises new challenges, akin to those face in realistic diagnostic applications. Here, we demonstrate the feasibility of inter-site classification of neuropsychiatric status, with an application to the Autism Brain Imaging Data Exchange (ABIDE) database, a large (N=871) multi-site autism dataset. For this purpose, we investigate pipelines that extract the most predictive biomarkers from the data. These R-fMRI pipelines build participant-specific connectomes from functionally-defined brain areas. Connectomes are then compared across participants to learn patterns of connectivity that differentiate typical controls from individuals with autism. We predict this neuropsychiatric status for participants from the same acquisition sites or different, unseen, ones. Good choices of methods for the various steps of the pipeline lead to 67% prediction accuracy on the full ABIDE data, which is significantly better than previously reported results. We perform extensive validation on multiple subsets of the data defined by different inclusion criteria. These enables detailed analysis of the factors contributing to successful connectome-based prediction. First, prediction accuracy improves as we include more subjects, up to the maximum amount of subjects available. Second, the definition of functional brain areas is of paramount importance for biomarker discovery: brain areas extracted from large R-fMRI datasets outperform reference atlases in the classification tasks.

Comparing functional connectivity based predictive models across datasets

Abstract: Resting-state functional Magnetic Resonance Imaging (rs-fMRI) holds the promise of easy-to-acquire and widespectrum biomarkers. However, there are few predictivemodeling studies on resting state, and processing pipelines all vary. Here, we systematically study resting state functionalconnectivity (FC)-based prediction across three different cohorts. Analysis pipelines consist of four steps: Delineation of brain regions of interest (ROIs), ROI-level rs-fMRI time series signal extraction, FC estimation and linear model classification analysis of FC features. For each step, we explore various methodological choices: ROI set selection, FC metrics, and linear classifiers to compare and evaluate the dominant strategies for the sake of prediction accuracy. We achieve good prediction results on the three different targets. With regard to pipeline selection, we obtain consistent results in two pipeline steps -FC metrics and linear classifiers- that are vital in the diagnosis of rs-fMRI based disease biomarkers. Regarding brain ROIs selection, we observe that the effects of different diseases are best characterized by different strategies: Schizophrenia discrimination is best performed in dataset-specific ROIs, which is not clearly the case for other pathologies. Overall, we outline some dominant strategies, in spite of the specificity of each brain disease in term of FC pattern disruption.

Social-sparsity brain decoders: faster spatial sparsity

Abstract: Spatially-sparse predictors are good models for brain decoding: they give accurate predictions and their weight maps are interpretable as they focus on a small number of regions. However, the state of the art, based on total variation or graph-net, is computationally costly. Here we introduce sparsity in the local neighborhood of each voxel with social-sparsity, a structured shrinkage operator. We find that, on brain imaging classification problems, social-sparsity performs almost as well as total-variation models and better than graph-net, for a fraction of the computational cost. It also very clearly outlines predictive regions. We give details of the model and the algorithm.

Functional Neuroimaging Group Studies

Abstract: Multi-subject statistical analysis is an essential step of neuroimaging studies, as it makes it possible to draw conclusions that hold with a prescribed confidence level for the population under study. The use of the linear assumption to model activation signals in brain images and their modulation by various factors has opened the possibility to rely on relatively simple estimation and statistical testing procedures. Specifically, the analysis of functional neuroimaging signals is typically carried out on a per-voxel basis, in the so-called mass univariate framework. However, the lack of power in neuroimaging studies has incited neuroscientists to develop new procedures to improve this framework: various solutions have been set up to take into account the spatial context in statistical inference or to deal with violations of distributional assumptions of the data. In this chapter, we review the general framework for group inference, the ensuing mixed-effects model design and its simplifications, together with the various solutions that have been considered to improve the standard mass-univariate testing framework.

Subsampled online matrix factorization with convergence guarantees

Abstract: We present a matrix factorization algorithm that scales to input matrices that are large in both dimensions (i.e., that contains more than 1TB of data). The algorithm streams the matrix columns while subsampling them, resulting in low complexity per iteration and reasonable memory footprint. In contrast to previous online matrix factorization methods, our approach relies on low-dimensional statistics from past iterates to control the extra variance introduced by subsampling. We present a convergence analysis that guarantees us to reach a stationary point of the problem. Large speed-ups can be obtained compared to previous online algorithms that do not perform subsampling, thanks to the feature redundancy that often exists in high-dimensional settings.

Local Q-linear convergence and finite-time active set identification of ADMM on a class of penalized regression problems

Abstract: We study the convergence of the ADMM (Alternating Direction Method of Multipliers) algorithm on a broad range of penalized regression problems including the Lasso, Group-Lasso and Graph-Lasso,(isotropic) TV-L1, Sparse Variation, and others. First, we establish a fixed-point iterationvia a nonlinear operator-which is equivalent to the ADMM iterates. We then show that this nonlinear operator is Frechet-differentiable almost everywhere and that around each fixed point, Q-linear convergence is guaranteed, provided the spectral radius of the Jacobian of the operator at the fixed point is less than 1 (a classical result on stability). Moreover, this spectral radius is then a rate of convergence for the ADMM algorithm. Also, we show that the support of the split variable can be identified after finitely many iterations. In the anisotropic cases, we show that for sufficiently large values of the tuning parameter, we recover the optimal rates in terms of Friedrichs angles, that have appeared recently in the literature. Empirical results on various problems are also presented and discussed.

Social-sparsity brain decoders: faster spatial sparsity

Abstract: Spatially-sparse predictors are good models for brain decoding: they give accurate predictions and their weight maps are interpretable as they focus on a small number of regions. However, the state of the art, based on total variation or graph-net, is computationally costly. Here we introduce sparsity in the local neighborhood of each voxel with social-sparsity, a structured shrinkage operator. We find that, on brain imaging classification problems, social-sparsity performs almost as well as total-variation models and better than graph-net, for a fraction of the computational cost. It also very clearly outlines predictive regions. We give details of the model and the algorithm.

Fast brain decoding with random sampling and random projections

Abstract: Machine learning from brain images is a central tool for image-based diagnosis and diseases characterization. Predicting behavior from functional imaging, brain decoding, analyzes brain activity in terms of the behavior that it implies. While these multivariate techniques are becoming standard brain mapping tools, like mass-univariate analysis, they entail much larger computational costs. In an time of growing data sizes, with larger cohorts and higher-resolutions imaging, this cost is increasingly a burden. Here we consider the use of random sampling and projections as fast data approximation techniques for brain images. We evaluate their prediction accuracy and computation time on various datasets and discrimination tasks. We show that the weight maps obtained after random sampling are highly consistent with those obtained with the whole feature space, while having a fair prediction performance. Altogether, we present the practical advantage of random sampling methods in neuroimaging, showing a simple way to embed back the reduced coefficients, with only a small loss of information.

Compressed Online Dictionary Learning for Fast fMRI Decomposition

Abstract: We present a method for fast resting-state fMRI spatial decomposi-tions of very large datasets, based on the reduction of the temporal dimension before applying dictionary learning on concatenated individual records from groups of subjects. Introducing a measure of correspondence between spatial decompositions of rest fMRI, we demonstrates that time-reduced dictionary learning produces result as reliable as non-reduced decompositions. We also show that this reduction significantly improves computational scalability.

Subsampled online matrix factorization with convergence guarantees

Abstract: We present a matrix factorization algorithm that scales to input matrices that are large in both dimensions (i.e., that contains morethan 1TB of data). The algorithm streams the matrix columns while subsampling them, resulting in low complexity per iteration andreasonable memory footprint. In contrast to previous online matrix factorization methods, our approach relies on low-dimensional statistics from past iterates to control the extra variance introduced by subsampling. We present a convergence analysis that guarantees us to reach a stationary point of the problem. Large speed-ups can be obtained compared to previous online algorithms that do not perform subsampling, thanks to the feature redundancy that often exists in high-dimensional settings.

Multimodal Connectivity-Based Parcellation of the Brain

Abstract: Bertrand Thirion – Neurospin Multimodal connectivity-based parcellation (CBP) delineates distinct sub-regions within a larger region of interest (ROI) based on multiple imaging modalities. First, connectivity between ROI voxels and other parts of the whole brain are computed. These features are then used to identify distinct groups of voxels. These represent functional modules of the brain that provide a priori information for modelling and pathophysiological investigations.