Showing papers by "Bertrand Thirion published in 2015"

Connectivity-based parcellation: Critique and implications

Abstract: Regional specialization and functional integration are often viewed as two fundamental principles of human brain organization. They are closely intertwined because each functionally specialized brain region is probably characterized by a distinct set of long-range connections. This notion has prompted the quickly developing family of connectivity-based parcellation (CBP) methods in neuroi-maging research. CBP assumes that there is a latent structure of parcels in a region of interest (ROI). First, connectivity strengths are computed to other parts of the brain for each voxel/vertex within the ROI. These features are then used to identify functionally distinct groups of ROI voxels/vertices. CBP enjoys increasing popularity for the in-vivo mapping of regional specialization in the human brain. Due to the requirements of different applications and datasets, CBP has diverged into a heterogeneous family of methods. This broad overview critically discusses the current state as well as the commonal-ities and idiosyncrasies of the main CBP methods. We target frequent concerns faced by novices and veterans to provide a reference for the investigation and review of CBP studies.

Semi-supervised factored logistic regression for high-dimensional neuroimaging data

Abstract: Imaging neuroscience links human behavior to aspects of brain biology in ever-increasing datasets. Existing neuroimaging methods typically perform either discovery of unknown neural structure or testing of neural structure associated with mental tasks. However, testing hypotheses on the neural correlates underlying larger sets of mental tasks necessitates adequate representations for the observations. We therefore propose to blend representation modelling and task classification into a unified statistical learning problem. A multinomial logistic regression is introduced that is constrained by factored coefficients and coupled with an autoencoder. We show that this approach yields more accurate and interpretable neural models of psychological tasks in a reference dataset, as well as better generalization to other datasets.

Improving Sparse Recovery on Structured Images with Bagged Clustering

Abstract: The identification of image regions associated with external variables through discriminative approaches yields ill-posed estimation problems. This estimation challenge can be tackled by imposing sparse solutions. However, the sensitivity of sparse estimators to correlated variables leads to non-reproducible results, and only a subset of the important variables are selected. In this paper, we explore an approach based on bagging clustering-based data compression in order to alleviate the instability of sparse models. Specifically, we design a new framework in which the estimator is built by averaging multiple models estimated after feature clustering, to improve the conditioning of the model. We show that this combination of model averaging with spatially consistent compression can have the virtuous effect of increasing the stability of the weight maps, allowing a better interpretation of the results. Finally, we demonstrate the benefit of our approach on several predictive modeling problems.

Clustering the infant brain tissues based on microstructural properties and maturation assessment using multi-parametric MRI

Abstract: Microstructural and physiological changes are intense in the developing brain, thus considerably modifying parameters quantified by MRI (relaxation times, anisotropy and diffusivities). The latest advances in EPI enabled to non-invasively measure these parameters in infants in a reasonable acquisition time to follow brain maturation. To take advantage of the parameters complementarity, we first proposed to correct EPI distortions by a registration approach never applied on infants' data. We then clustered brain regions according to their different microstructural properties and maturation, without spatial priors. Results were in agreement with post-mortem studies : maturation proceeds from center to periphery; first maturing regions included primary regions, but also the amygdala and the medial cingular region. The most delayed regions were observed in superior frontal and parietal lobes. In future analyses, we will propose new partitions of the whole brain and of specific functional regions based on maturation patterns returned by this approach.

Integrating Multimodal Priors in Predictive Models for the Functional Characterization of Alzheimer's Disease

Abstract: Functional brain imaging provides key information to characterize neurodegenerative diseases, such as Alzheimer's disease AD. Specifically, the metabolic activity measured through fluorodeoxyglucose positron emission tomography FDG-PET and the connectivity extracted from resting-state functional magnetic resonance imaging fMRI, are promising biomarkers that can be used for early assessment and prognosis of the disease and to understand its mechanisms. FDG-PET is the best suited functional marker so far, as it gives a reliable quantitative measure, but is invasive. On the other hand, non-invasive fMRI acquisitions do not provide a straightforward quantification of brain functional activity. To analyze populations solely based on resting-state fMRI, we propose an approach that leverages a metabolic prior learned from FDG-PET. More formally, our classification framework embeds population priors learned from another modality at the voxel-level, which can be seen as a regularization term in the analysis. Experimental results show that our PET-informed approach increases classification accuracy compared to pure fMRI approaches and highlights regions known to be impacted by the disease.

Speeding-Up Model-Selection in Graphnet via Early-Stopping and Univariate Feature-Screening

Abstract: The Graph Net (aka S-Lasso), as well as other "spar-sity + structure" priors like TV-L1, are not easily applicable to brain data because of technical problems concerning the selection of the regularization parameters. Also, in their own right, such models lead to challenging high-dimensional optimization problems. In this manuscript, we present some heuristics for speeding up the overall optimization process: (a) Early-stopping, whereby one halts the optimization process when the test score(performance on left out data) for the internal cross validation for model-selection stops improving, and (b) univariate feature-screening, whereby irrelevant (non-predictive) voxels are detected and eliminated before the optimization problem is entered, thus reducing the size of the problem. Empirical results with Graph Net on real MRI (Magnetic Resonance Imaging) datasets indicate that these heuristics are a win-win strategy, as they add speed without sacrificing the quality of the predictions. We expect the proposed heuristics to work on other models like TV-L1, etc.

Grouping Total Variation and Sparsity: Statistical Learning with Segmenting Penalties

Abstract: Prediction from medical images is a valuable aid to diagnosis For instance, anatomical MR images can reveal certain disease conditions, while their functional counterparts can predict neuropsychiatric phenotypes However, a physician will not rely on predictions by black-box models: understanding the anatomical or functional features that underpin decision is critical Generally, the weight vectors of classifiers are not easily amenable to such an examination: Often there is no apparent structure Indeed, this is not only a prediction task, but also an inverse problem that calls for adequate regularization We address this challenge by introducing a convex region-selecting penalty Our penalty combines total-variation regularization, enforcing spatial contiguity, and l1 regularization, enforcing sparsity, into one group: Voxels are either active with non-zero spatial derivative or zero with inactive spatial derivative This leads to segmenting contiguous spatial regions inside which the signal can vary freely against a background of zeros Such segmentation of medical images in a target-informed manner is an important analysis tool On several prediction problems from brain MRI, the penalty shows good segmentation Given the size of medical images, computational efficiency is key Keeping this in mind, we contribute an efficient optimization scheme that brings significant computational gains

Correlations of correlations are not reliable statistics: implications for multivariate pattern analysis.

Abstract: Representational Similarity Analysis is a popular framework to flexibly represent the statistical dependencies between multi-voxel patterns on the one hand, and sensory or cognitive stimuli on the other hand. It has been used in an inferen-tial framework, whereby significance is given by a permutation test on the samples. In this paper , we outline an issue with this statistical procedure: namely that the so-called pattern similarity used can be influenced by various effects, such as noise variance, which can lead to inflated type I error rates. What we propose is to rely instead on proper linear models.

Total Variation meets Sparsity: statistical learning with segmenting penalties

Abstract: Prediction from medical images is a valuable aid to diagnosis For instance, anatomical MR images can reveal certain disease conditions, while their functional counterparts can predict neuropsychi-atric phenotypes However, a physician will not rely on predictions by black-box models: understanding the anatomical or functional features that underpin decision is critical Generally, the weight vectors of clas-sifiers are not easily amenable to such an examination: Often there is no apparent structure Indeed, this is not only a prediction task, but also an inverse problem that calls for adequate regularization We address this challenge by introducing a convex region-selecting penalty Our penalty combines total-variation regularization, enforcing spatial conti-guity, and 1 regularization, enforcing sparsity, into one group: Voxels are either active with non-zero spatial derivative or zero with inactive spatial derivative This leads to segmenting contiguous spatial regions (inside which the signal can vary freely) against a background of zeros Such segmentation of medical images in a target-informed manner is an important analysis tool On several prediction problems from brain MRI, the penalty shows good segmentation Given the size of medical images, computational efficiency is key Keeping this in mind, we contribute an efficient optimization scheme that brings significant computational gains

FAASTA: A fast solver for total-variation regularization of ill-conditioned problems with application to brain imaging

Abstract: The total variation (TV) penalty, as many other analysis-sparsity problems, does not lead to separable factors or a proximal operator with a closed-form expression, such as soft thresholding for the $\ell_1$ penalty. As a result, in a variational formulation of an inverse problem or statistical learning estimation, it leads to challenging non-smooth optimization problems that are often solved with elaborate single-step first-order methods. When the data-fit term arises from empirical measurements, as in brain imaging, it is often very ill-conditioned and without simple structure. In this situation, in proximal splitting methods, the computation cost of the gradient step can easily dominate each iteration. Thus it is beneficial to minimize the number of gradient steps. We present fAASTA, a variant of FISTA, that relies on an internal solver for the TV proximal operator, and refines its tolerance to balance computational cost of the gradient and the proximal steps. We give benchmarks and illustrations on ``brain decoding'': recovering brain maps from noisy measurements to predict observed behavior. The algorithm as well as the empirical study of convergence speed are valuable for any non-exact proximal operator, in particular analysis-sparsity problems.

Bootstrapped Permutation Test for Multiresponse Inference on Brain Behavior Associations

Abstract: Despite that diagnosis of neurological disorders commonly involves a collection of behavioral assessments, most neuroimaging studies investigating the associations between brain and behavior largely analyze each behavioral measure in isolation. To jointly model multiple behavioral scores, sparse multiresponse regression (SMR) is often used. However, directly applying SMR without statistically controlling for false positives could result in many spurious findings. For models, such as SMR, where the distribution of the model parameters is unknown, permutation test and stability selection are typically used to control for false positives. In this paper, we present another technique for inferring statistically significant features from models with unknown parameter distribution. We refer to this technique as bootstrapped permutation test (BPT), which uses Studentized statistics to exploit the intuition that the variability in parameter estimates associated with relevant features would likely be higher with responses permuted. On synthetic data, we show that BPT provides higher sensitivity in identifying relevant features from the SMR model than permutation test and stability selection, while retaining strong control on the false positive rate. We further apply BPT to study the associations between brain connectivity estimated from pseudo-rest fMRI data of 1139 fourteen year olds and behavioral measures related to ADHD. Significant connections are found between brain networks known to be implicated in the behavioral tasks involved. Moreover, we validate the identified connections by fitting a regression model on pseudo-rest data with only those connections and applying this model on resting state fMRI data of 337 left out subjects to predict their behavioral scores. The predicted scores significantly correlate with the actual scores, hence verifying the behavioral relevance of the found connections.

SpaceNet: Multivariate brain decoding and segmentation

Abstract: We present SpaceNet, a multivariate method for brain decoding and segmentation. SpaceNet uses priors like TV (Total Variation). SpaceNet uses priors like TV (Total Variation) [Michel et al. 2011], TV-L1 [Baldassarre et al. 2012, Gramfort et al. 2013], and GraphNet / Smooth-Lasso [Hebiri et al. 2011, Grosenick et al. 2013] to regularize / penalize classification and regression problems in brain imaging. The result are brain maps which are both sparse (i.e regression coefficients are zero everywhere, except at predictive voxels) and structured (blobby). The superiority of such priors over methods without structured priors like the Lasso, SVM, ANOVA, Ridge, etc. for yielding more interpretable maps and improved classification / prediction scores is now well-established [Baldassarre et al. 2012, Gramfort et al. 2013, Grosenick et al. 2013]. In addition, such priors lead to state-of-the-art methods for extracting brain atlases [Abraham et al. 2013].

Intra and inter subject analyses of brain functional Magnetic Resonance Images (fMRI)

Abstract: This chapter proposes a review of the most prominent issues in analysing brain functional Magnetic Resonance data. It introduces the domain for readers with no or little knowledge in the field. The introduction places the context and orients the reader in the many questions put to the data, and summarizes the currently most commonly applied approach. The second section deals with intra subject data analysis, emphasizing hemodynamic response estimation issues. The third section describes current approaches and advances in analysing group data in a standard coordinate system. The last section proposes new spatial models for group analyses. Overall, the chapter gives a brief overview of the field and details some specific advances that are important for application studies in cognitive neurosciences.

FAASTA: A fast solver for total-variation regularization of ill-conditioned problems with application to brain imaging

Abstract: The total variation (TV) penalty, as many other analysis-sparsity problems, does not lead to separable factors or a proximal operatorwith a closed-form expression, such as soft thresholding for the $\ell\_1$ penalty. As a result, in a variational formulation of an inverse problem or statisticallearning estimation, it leads to challenging non-smooth optimization problemsthat are often solved with elaborate single-step first-order methods. When thedata-fit term arises from empirical measurements, as in brain imaging, it isoften very ill-conditioned and without simple structure. In this situation, in proximal splitting methods, the computation cost of thegradient step can easily dominate each iteration. Thus it is beneficialto minimize the number of gradient steps.We present fAASTA, a variant of FISTA, that relies on an internal solver forthe TV proximal operator, and refines its tolerance to balance computationalcost of the gradient and the proximal steps. We give benchmarks andillustrations on "brain decoding": recovering brain maps from noisymeasurements to predict observed behavior. The algorithm as well as theempirical study of convergence speed are valuable for any non-exact proximaloperator, in particular analysis-sparsity problems.

Fast clustering for scalable statistical analysis on structured images

Abstract: The use of brain images as markers for diseases or behavioral differences is challenged by the small effects size and the ensuing lack of power, an issue that has incited researchers to rely more systematically on large cohorts. Coupled with resolution increases, this leads to very large datasets. A striking example in the case of brain imaging is that of the Human Connectome Project: 20 Terabytes of data and growing. The resulting data deluge poses severe challenges regarding the tractability of some processing steps (discriminant analysis, multivariate models) due to the memory demands posed by these data. In this work, we revisit dimension reduction approaches, such as random projections , with the aim of replacing costly function evaluations by cheaper ones while decreasing the memory requirements. Specifically, we investigate the use of alternate schemes, based on fast clustering, that are well suited for signals exhibiting a strong spatial structure, such as anatomical and functional brain images. Our contribution is twofold: i) we propose a linear-time clustering scheme that bypasses the percolation issues inherent in these algorithms and thus provides compressions nearly as good as traditional quadratic-complexity variance-minimizing clustering schemes; ii) we show that cluster-based compression can have the virtuous effect of removing high-frequency noise, actually improving subsequent estimations steps. As a consequence , the proposed approach yields very accurate models on several large-scale problems yet with impressive gains in computational efficiency , making it possible to analyze large datasets.

Fast clustering for scalable statistical analysis on structured images

Abstract: The use of brain images as markers for diseases or behavioral differences is challenged by the small effects size and the ensuing lack of power, an issue that has incited researchers to rely more systematically on large cohorts. Coupled with resolution increases, this leads to very large datasets. A striking example in the case of brain imaging is that of the Human Connectome Project: 20 Terabytes of data and growing. The resulting data deluge poses severe challenges regarding the tractability of some processing steps (discriminant analysis, multivariate models) due to the memory demands posed by these data. In this work, we revisit dimension reduction approaches, such as random projections, with the aim of replacing costly function evaluations by cheaper ones while decreasing the memory requirements. Specifically, we investigate the use of alternate schemes, based on fast clustering, that are well suited for signals exhibiting a strong spatial structure, such as anatomical and functional brain images. Our contribution is twofold: i) we propose a linear-time clustering scheme that bypasses the percolation issues inherent in these algorithms and thus provides compressions nearly as good as traditional quadratic-complexity variance-minimizing clustering schemes, ii) we show that cluster-based compression can have the virtuous effect of removing high-frequency noise, actually improving subsequent estimations steps. As a consequence, the proposed approach yields very accurate models on several large-scale problems yet with impressive gains in computational efficiency, making it possible to analyze large datasets.