Arrow Research search

Author name cluster

Bertrand Thirion

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

55 papers
2 author rows

Possible papers

55

ICML Conference 2025 Conference Paper

False Coverage Proportion Control for Conformal Prediction

  • Alexandre Blain
  • Bertrand Thirion
  • Pierre Neuvial

Split Conformal Prediction (SCP) provides a computationally efficient way to construct confidence intervals in prediction problems. Notably, most of the theory built around SCP is focused on the single test point setting. In real-life, inference sets consist of multiple points, which raises the question of coverage guarantees for many points simultaneously. While on average, the False Coverage Proportion (FCP) remains controlled, it can fluctuate strongly around its mean, the False Coverage Rate (FCR). We observe that when a dataset is split multiple times, classical SCP may not control the FCP in a majority of the splits. We propose CoJER, a novel method that achieves sharp FCP control in probability for conformal prediction, based on a recent characterization of the distribution of conformal $p$-values in a transductive setting. This procedure incorporates an aggregation scheme which provides robustness with respect to modeling choices. We show through extensive real data experiments that CoJER provides FCP control while standard SCP does not. Furthermore, CoJER yields shorter intervals than the state-of-the-art method for FCP control and only slightly larger intervals than standard SCP.

Details

ICML Conference 2025 Conference Paper

Measuring Variable Importance in Heterogeneous Treatment Effects with Confidence

  • Joseph Paillard
  • Angel David Reyero Lobo
  • Vitaliy Kolodyazhniy
  • Bertrand Thirion
  • Denis-Alexander Engemann

Causal machine learning (ML) promises to provide powerful tools for estimating individual treatment effects. While causal methods have placed some emphasis on heterogeneity in treatment response, it is of paramount importance to clarify the nature of this heterogeneity, by highlighting which variables drive it. We propose PermuCATE, an algorithm based on the Conditional Permutation Importance (CPI) method, for statistically rigorous global variable importance assessment in the estimation of the Conditional Average Treatment Effect (CATE). Theoretical analysis of the finite sample regime and empirical studies show that PermuCATE has lower variance than the Leave-One-Covariate-Out (LOCO) method and provides a reliable measure of variable importance. This property increases statistical power, which is crucial for causal inference applications with finite sample sizes. We empirically demonstrate the benefits of PermuCATE in simulated and real datasets, including complex settings with high-dimensional, correlated variables.

Details

NeurIPS Conference 2025 Conference Paper

Riemannian Flow Matching for Brain Connectivity Matrices via Pullback Geometry

  • Antoine Collas
  • Ce Ju
  • Nicolas Salvy
  • Bertrand Thirion

Generating realistic brain connectivity matrices is key to analyzing population heterogeneity in brain organization, understanding disease, and augmenting data in challenging classification problems. Functional connectivity matrices lie in constrained spaces—such as the set of symmetric positive definite or correlation matrices—that can be modeled as Riemannian manifolds. However, using Riemannian tools typically requires redefining core operations (geodesics, norms, integration), making generative modeling computationally inefficient. In this work, we propose DiffeoCFM, an approach that enables conditional flow matching (CFM) on matrix manifolds by exploiting pullback metrics induced by global diffeomorphisms on Euclidean spaces. We show that Riemannian CFM with such metrics is equivalent to applying standard CFM after data transformation. This equivalence allows efficient vector field learning, and fast sampling with standard ODE solvers. We instantiate DiffeoCFM with two different settings: the matrix logarithm for covariance matrices and the normalized Cholesky decomposition for correlation matrices. We evaluate DiffeoCFM on three large-scale fMRI datasets with more than $4600$ scans from $2800$ subjects (ADNI, ABIDE, OASIS‑3) and two EEG motor imagery datasets with over $30000$ trials from $26$ subjects (BNCI2014‑002 and BNCI2015‑001). It enables fast training and achieves state-of-the-art performance, all while preserving manifold constraints. Code: https: //github. com/antoinecollas/DiffeoCFM

PDF Details

TMLR Journal 2025 Journal Article

Sample-efficient decoding of visual stimuli from fMRI through inter-individual functional alignment

  • Alexis Thual
  • Yohann Benchetrit
  • Felix Geilert
  • Jérémy Rapin
  • Iurii Makarov
  • Stanislas Dehaene
  • Bertrand Thirion
  • Hubert Banville

Deep learning is leading to major advances in the realm of brain decoding from functional Magnetic Resonance Imaging (fMRI). However, the large inter-individual variability in brain characteristics has constrained most studies to train models on one participant at a time. This limitation hampers the training of deep learning models, which typically requires very large datasets. Here, we propose to boost brain decoding of videos and static images across participants by aligning brain responses of training and left-out participants. Evaluated on a retrieval task, compared to the anatomically-aligned baseline, our method halves the median rank in out-of-subject setups. It also outperforms classical within-subject approaches when fewer than 100 minutes of data is available for the tested participant. Furthermore, we show that our alignment framework handles multiple subjects, which improves accuracy upon classical single-subject approaches. Finally, we show that this method aligns neural representations in accordance with brain anatomy. Overall, this study lays the foundations for leveraging extensive neuroimaging datasets and enhancing the decoding of individual brains when a limited amount of brain-imaging data is available.

PDF Details

NeurIPS Conference 2025 Conference Paper

Unsupervised Learning for Optimal Transport plan prediction between unbalanced graphs

  • Sonia Mazelet
  • Rémi Flamary
  • Bertrand Thirion

Optimal transport between graphs, based on Gromov-Wasserstein and other extensions, is a powerful tool for comparing and aligning graph structures. However, solving the associated non-convex optimization problems is computationally expensive, which limits the scalability of these methods to large graphs. In this work, we present Unbalanced Learning of Optimal Transport (ULOT), a deep learning method that predicts optimal transport plans between two graphs. Our method is trained by minimizing the fused unbalanced Gromov-Wasserstein (FUGW) loss. We propose a novel neural architecture with cross-attention that is conditioned on the FUGW tradeoff hyperparameters. We evaluate ULOT on synthetic stochastic block model (SBM) graphs and on real cortical surface data obtained from fMRI. ULOT predicts transport plans with competitive loss up to two orders of magnitude faster than classical solvers. Furthermore, the predicted plan can be used as a warm start for classical solvers to accelerate their convergence. Finally, the predicted transport plan is fully differentiable with respect to the graph inputs and FUGW hyperparameters, enabling the optimization of functionals of the ULOT plan.

PDF Details

AAAI Conference 2024 Conference Paper

Variable Importance in High-Dimensional Settings Requires Grouping

  • Ahmad Chamma
  • Bertrand Thirion
  • Denis Engemann

Explaining the decision process of machine learning algorithms is nowadays crucial for both model’s performance enhancement and human comprehension. This can be achieved by assessing the variable importance of single variables, even for high-capacity non-linear methods, e.g. Deep Neural Networks (DNNs). While only removal-based approaches, such as Permutation Importance (PI), can bring statistical validity, they return misleading results when variables are correlated. Conditional Permutation Importance (CPI) bypasses PI’s limitations in such cases. However, in high-dimensional settings, where high correlations between the variables cancel their conditional importance, the use of CPI as well as other methods leads to unreliable results, besides prohibitive computation costs. Grouping variables statistically via clustering or some prior knowledge gains some power back and leads to better interpretations. In this work, we introduce BCPI (Block-Based Conditional Permutation Importance), a new generic framework for variable importance computation with statistical guarantees handling both single and group cases. Furthermore, as handling groups with high cardinality (such as a set of observations of a given modality) are both time-consuming and resource-intensive, we also introduce a new stacking approach extending the DNN architecture with sub-linear layers adapted to the group structure. We show that the ensuing approach extended with stacking controls the type-I error even with highly-correlated groups and shows top accuracy across benchmarks. Furthermore, we perform a real-world data analysis in a large-scale medical dataset where we aim to show the consistency between our results and the literature for a biomarker prediction.

PDF Details DOI

NeurIPS Conference 2023 Conference Paper

False Discovery Proportion control for aggregated Knockoffs

  • Alexandre Blain
  • Bertrand Thirion
  • Olivier Grisel
  • Pierre Neuvial

Controlled variable selection is an important analytical step in various scientific fields, such as brain imaging or genomics. In these high-dimensional data settings, considering too many variables leads to poor models and high costs, hence the need for statistical guarantees on false positives. Knockoffs are a popular statistical tool for conditional variable selection in high dimension. However, they control for the expected proportion of false discoveries (FDR) and not the actual proportion of false discoveries (FDP). We present a new method, KOPI, that controls the proportion of false discoveries for Knockoff-based inference. The proposed method also relies on a new type of aggregation to address the undesirable randomness associated with classical Knockoff inference. We demonstrate FDP control and substantial power gains over existing Knockoff-based methods in various simulation settings and achieve good sensitivity/specificity tradeoffs on brain imaging data.

PDF Details

NeurIPS Conference 2023 Conference Paper

Statistically Valid Variable Importance Assessment through Conditional Permutations

  • Ahmad Chamma
  • Denis A. Engemann
  • Bertrand Thirion

Variable importance assessment has become a crucial step in machine-learning applications when using complex learners, such as deep neural networks, on large-scale data. Removal-based importance assessment is currently the reference approach, particularly when statistical guarantees are sought to justify variable inclusion. It is often implemented with variable permutation schemes. On the flip side, these approaches risk misidentifying unimportant variables as important in the presence of correlations among covariates. Here we develop a systematic approach for studying Conditional Permutation Importance (CPI) that is model agnostic and computationally lean, as well as reusable benchmarks of state-of-the-art variable importance estimators. We show theoretically and empirically that \textit{CPI} overcomes the limitations of standard permutation importance by providing accurate type-I error control. When used with a deep neural network, \textit{CPI} consistently showed top accuracy across benchmarks. An experiment on real-world data analysis in a large-scale medical dataset showed that \textit{CPI} provides a more parsimonious selection of statistically significant variables. Our results suggest that \textit{CPI} can be readily used as drop-in replacement for permutation-based methods.

PDF Details

NeurIPS Conference 2022 Conference Paper

A Conditional Randomization Test for Sparse Logistic Regression in High-Dimension

  • Binh T. Nguyen
  • Bertrand Thirion
  • Sylvain Arlot

Identifying the relevant variables for a classification model with correct confidence levels is a central but difficult task in high-dimension. Despite the core role of sparse logistic regression in statistics and machine learning, it still lacks a good solution for accurate inference in the regime where the number of features $p$ is as large as or larger than the number of samples $n$. Here we tackle this problem by improving the Conditional Randomization Test (CRT). The original CRT algorithm shows promise as a way to output p-values while making few assumptions on the distribution of the test statistics. As it comes with a prohibitive computational cost even in mildly high-dimensional problems, faster solutions based on distillation have been proposed. Yet, they rely on unrealistic hypotheses and result in low-power solutions. To improve this, we propose \emph{CRT-logit}, an algorithm that combines a variable-distillation step and a decorrelation step that takes into account the geometry of $\ell_1$-penalized logistic regression problem. We provide a theoretical analysis of this procedure, and demonstrate its effectiveness on simulations, along with experiments on large-scale brain-imaging and genomics datasets.

PDF Details

NeurIPS Conference 2022 Conference Paper

Aligning individual brains with fused unbalanced Gromov Wasserstein

  • Alexis Thual
  • Quang Huy Tran
  • Tatiana Zemskova
  • Nicolas Courty
  • Rémi Flamary
  • Stanislas Dehaene
  • Bertrand Thirion

Individual brains vary in both anatomy and functional organization, even within a given species. Inter-individual variability is a major impediment when trying to draw generalizable conclusions from neuroimaging data collected on groups of subjects. Current co-registration procedures rely on limited data, and thus lead to very coarse inter-subject alignments. In this work, we present a novel method for inter-subject alignment based on Optimal Transport, denoted as Fused Unbalanced Gromov Wasserstein (FUGW). The method aligns two cortical surfaces based on the similarity of their functional signatures in response to a variety of stimuli, while penalizing large deformations of individual topographic organization. We demonstrate that FUGW is suited for whole-brain landmark-free alignment. The unbalanced feature allows to deal with the fact that functional areas vary in size across subjects. Results show that FUGW alignment significantly increases between-subject correlation of activity during new independent fMRI tasks and runs, and leads to more precise maps of fMRI results at the group level.

PDF Details

ICML Conference 2022 Conference Paper

Neural Language Models are not Born Equal to Fit Brain Data, but Training Helps

  • Alexandre Pasquiou
  • Yair Lakretz
  • John T. Hale
  • Bertrand Thirion
  • Christophe Pallier

Neural Language Models (NLMs) have made tremendous advances during the last years, achieving impressive performance on various linguistic tasks. Capitalizing on this, studies in neuroscience have started to use NLMs to study neural activity in the human brain during language processing. However, many questions remain unanswered regarding which factors determine the ability of a neural language model to capture brain activity (aka its ’brain score’). Here, we make first steps in this direction and examine the impact of test loss, training corpus and model architecture (comparing GloVe, LSTM, GPT-2 and BERT), on the prediction of functional Magnetic Resonance Imaging time-courses of participants listening to an audiobook. We find that (1) untrained versions of each model already explain significant amount of signal in the brain by capturing similarity in brain responses across identical words, with the untrained LSTM outperforming the transformer-based models, being less impacted by the effect of context; (2) that training NLP models improves brain scores in the same brain regions irrespective of the model’s architecture; (3) that Perplexity (test loss) is not a good predictor of brain score; (4) that training data have a strong influence on the outcome and, notably, that off-the-shelf models may lack statistical power to detect brain activations. Overall, we outline the impact of model-training choices, and suggest good practices for future studies aiming at explaining the human language system using neural language models.

Details

YNIMG Journal 2022 Journal Article

Notip: Non-parametric true discovery proportion control for brain imaging

  • Alexandre Blain
  • Bertrand Thirion
  • Pierre Neuvial

Cluster-level inference procedures are widely used for brain mapping. These methods compare the size of clusters obtained by thresholding brain maps to an upper bound under the global null hypothesis, computed using Random Field Theory or permutations. However, the guarantees obtained by this type of inference - i.e. at least one voxel is truly activated in the cluster - are not informative with regards to the strength of the signal therein. There is thus a need for methods to assess the amount of signal within clusters; yet such methods have to take into account that clusters are defined based on the data, which creates circularity in the inference scheme. This has motivated the use of post hoc estimates that allow statistically valid estimation of the proportion of activated voxels in clusters. In the context of fMRI data, the All-Resolutions Inference framework introduced in Rosenblatt et al. (2018) provides post hoc estimates of the proportion of activated voxels. However, this method relies on parametric threshold families, which results in conservative inference. In this paper, we leverage randomization methods to adapt to data characteristics and obtain tighter false discovery control. We obtain Notip, for Non-parametric True Discovery Proportion control: a powerful, non-parametric method that yields statistically valid guarantees on the proportion of activated voxels in data-derived clusters. Numerical experiments demonstrate substantial gains in number of detections compared with state-of-the-art methods on 36 fMRI datasets. The conditions under which the proposed method brings benefits are also discussed.

Details DOI

YNIMG Journal 2021 Journal Article

An empirical evaluation of functional alignment using inter-subject decoding

  • Thomas Bazeille
  • Elizabeth DuPre
  • Hugo Richard
  • Jean-Baptiste Poline
  • Bertrand Thirion

Inter-individual variability in the functional organization of the brain presents a major obstacle to identifying generalizable neural coding principles. Functional alignment-a class of methods that matches subjects' neural signals based on their functional similarity-is a promising strategy for addressing this variability. To date, however, a range of functional alignment methods have been proposed and their relative performance is still unclear. In this work, we benchmark five functional alignment methods for inter-subject decoding on four publicly available datasets. Specifically, we consider three existing methods: piecewise Procrustes, searchlight Procrustes, and piecewise Optimal Transport. We also introduce and benchmark two new extensions of functional alignment methods: piecewise Shared Response Modelling (SRM), and intra-subject alignment. We find that functional alignment generally improves inter-subject decoding accuracy though the best performing method depends on the research context. Specifically, SRM and Optimal Transport perform well at both the region-of-interest level of analysis as well as at the whole-brain scale when aggregated through a piecewise scheme. We also benchmark the computational efficiency of each of the surveyed methods, providing insight into their usability and scalability. Taking inter-subject decoding accuracy as a quantification of inter-subject similarity, our results support the use of functional alignment to improve inter-subject comparisons in the face of variable structure-function organization. We provide open implementations of all methods used.

Details DOI

YNIMG Journal 2021 Journal Article

Brain topography beyond parcellations: Local gradients of functional maps

  • Elvis Dohmatob
  • Hugo Richard
  • Ana Luísa Pinho
  • Bertrand Thirion

Functional neuroimaging provides the unique opportunity to characterize brain regions based on their response to tasks or ongoing activity. As such, it holds the premise to capture brain spatial organization. Yet, the conceptual framework to describe this organization has remained elusive: on the one hand, parcellations build implicitly on a piecewise constant organization, i.e. flat regions separated by sharp boundaries; on the other hand, the recently popularized concept of functional gradient hints instead at a smooth structure. Noting that both views converge to a topographic scheme that pieces together local variations of functional features, we perform a quantitative assessment of local gradient-based models. Using as a driving case the prediction of functional Magnetic Resonance Imaging (fMRI) data -concretely, the prediction of task-fMRI from rest-fMRI maps across subjects- we develop a parcel-wise linear regression model based on a dictionary of reference topographies. Our method uses multiple random parcellations -as opposed to a single fixed parcellation- and aggregates estimates across these parcellations to predict functional features in left-out subjects. Our experiments demonstrate the existence of an optimal cardinality of the parcellation to capture local gradients of functional maps.

Details DOI

YNIMG Journal 2021 Journal Article

Decoding with confidence: Statistical control on decoder maps

  • Jérôme-Alexis Chevalier
  • Tuan-Binh Nguyen
  • Joseph Salmon
  • Gaël Varoquaux
  • Bertrand Thirion

In brain imaging, decoding is widely used to infer relationships between brain and cognition, or to craft brain-imaging biomarkers of pathologies. Yet, standard decoding procedures do not come with statistical guarantees, and thus do not give confidence bounds to interpret the pattern maps that they produce. Indeed, in whole-brain decoding settings, the number of explanatory variables is much greater than the number of samples, hence classical statistical inference methodology cannot be applied. Specifically, the standard practice that consists in thresholding decoding maps is not a correct inference procedure. We contribute a new statistical-testing framework for this type of inference. To overcome the statistical inefficiency of voxel-level control, we generalize the Family Wise Error Rate (FWER) to account for a spatial tolerance δ, introducing the δ-Family Wise Error Rate (δ-FWER). Then, we present a decoding procedure that can control the δ-FWER: the Ensemble of Clustered Desparsified Lasso (EnCluDL), a procedure for multivariate statistical inference on high-dimensional structured data. We evaluate the statistical properties of EnCluDL with a thorough empirical study, along with three alternative procedures including decoder map thresholding. We show that EnCluDL exhibits the best recovery properties while ensuring the expected statistical control.

Details DOI

YNIMG Journal 2021 Journal Article

Functional annotation of human cognitive states using deep graph convolution

  • Yu Zhang
  • Loïc Tetrel
  • Bertrand Thirion
  • Pierre Bellec

A key goal in neuroscience is to understand brain mechanisms of cognitive functions. An emerging approach is "brain decoding", which consists of inferring a set of experimental conditions performed by a participant, using pattern classification of brain activity. Few works so far have attempted to train a brain decoding model that would generalize across many different cognitive tasks drawn from multiple cognitive domains. To tackle this problem, we proposed a multidomain brain decoder that automatically learns the spatiotemporal dynamics of brain response within a short time window using a deep learning approach. We evaluated the decoding model on a large population of 1200 participants, under 21 different experimental conditions spanning six different cognitive domains, acquired from the Human Connectome Project task-fMRI database. Using a 10s window of fMRI response, the 21 cognitive states were identified with a test accuracy of 90% (chance level 4.8%). Performance remained good when using a 6s window (82%). It was even feasible to decode cognitive states from a single fMRI volume (720ms), with the performance following the shape of the hemodynamic response. Moreover, a saliency map analysis demonstrated that the high decoding performance was driven by the response of biologically meaningful brain regions. Together, we provide an automated tool to annotate human brain activity with fine temporal resolution and fine cognitive granularity. Our model shows potential applications as a reference model for domain adaptation, possibly making contributions in a variety of domains, including neurological and psychiatric disorders.

Details DOI

NeurIPS Conference 2021 Conference Paper

Shared Independent Component Analysis for Multi-Subject Neuroimaging

  • Hugo Richard
  • Pierre Ablin
  • Bertrand Thirion
  • Alexandre Gramfort
  • Aapo Hyvarinen

We consider shared response modeling, a multi-view learning problem where one wants to identify common components from multiple datasets or views. We introduce Shared Independent Component Analysis (ShICA) that models eachview as a linear transform of shared independent components contaminated by additive Gaussian noise. We show that this model is identifiable if the components are either non-Gaussian or have enough diversity in noise variances. We then show that in some cases multi-set canonical correlation analysis can recover the correct unmixing matrices, but that even a small amount of sampling noise makes Multiset CCA fail. To solve this problem, we propose to use joint diagonalization after Multiset CCA, leading to a new approach called ShICA-J. We show via simulations that ShICA-J leads to improved results while being very fast to fit. While ShICA-J is based on second-order statistics, we further propose to leverage non-Gaussianity of the components using a maximum-likelihood method, ShICA-ML, that is both more accurate and more costly. Further, ShICA comes with a principled method for shared components estimation. Finally, we provide empirical evidence on fMRI and MEG datasets that ShICA yields more accurate estimation of the componentsthan alternatives.

PDF Details

ICML Conference 2020 Conference Paper

Aggregation of Multiple Knockoffs

  • Tuan-Binh Nguyen
  • Jérôme-Alexis Chevalier
  • Bertrand Thirion
  • Sylvain Arlot

We develop an extension of the knockoff inference procedure, introduced by Barber & Candes (2015). This new method, called Aggregation of Multiple Knockoffs (AKO), addresses the instability inherent to the random nature of knockoff-based inference. Specifically, AKO improves both the stability and power compared with the original knockoff algorithm while still maintaining guarantees for false discovery rate control. We provide a new inference procedure, prove its core properties, and demonstrate its benefits in a set of experiments on synthetic and real datasets.

Details

YNIMG Journal 2020 Journal Article

Fine-grain atlases of functional modes for fMRI analysis

  • Kamalaker Dadi
  • Gaël Varoquaux
  • Antonia Machlouzarides-Shalit
  • Krzysztof J. Gorgolewski
  • Demian Wassermann
  • Bertrand Thirion
  • Arthur Mensch

Population imaging markedly increased the size of functional-imaging datasets, shedding new light on the neural basis of inter-individual differences. Analyzing these large data entails new scalability challenges, computational and statistical. For this reason, brain images are typically summarized in a few signals, for instance reducing voxel-level measures with brain atlases or functional modes. A good choice of the corresponding brain networks is important, as most data analyses start from these reduced signals. We contribute finely-resolved atlases of functional modes, comprising from 64 to 1024 networks. These dictionaries of functional modes (DiFuMo) are trained on millions of fMRI functional brain volumes of total size 2. 4 ​TB, spanned over 27 studies and many research groups. We demonstrate the benefits of extracting reduced signals on our fine-grain atlases for many classic functional data analysis pipelines: stimuli decoding from 12, 334 brain responses, standard GLM analysis of fMRI across sessions and individuals, extraction of resting-state functional-connectomes biomarkers for 2500 individuals, data compression and meta-analysis over more than 15, 000 statistical maps. In each of these analysis scenarii, we compare the performance of our functional atlases with that of other popular references, and to a simple voxel-level analysis. Results highlight the importance of using high-dimensional “soft” functional atlases, to represent and analyze brain activity while capturing its functional gradients. Analyses on high-dimensional modes achieve similar statistical performance as at the voxel level, but with much reduced computational cost and higher interpretability. In addition to making them available, we provide meaningful names for these modes, based on their anatomical location. It will facilitate reporting of results.

Details DOI

NeurIPS Conference 2020 Conference Paper

Modeling Shared responses in Neuroimaging Studies through MultiView ICA

  • Hugo Richard
  • Luigi Gresele
  • Aapo Hyvarinen
  • Bertrand Thirion
  • Alexandre Gramfort
  • Pierre Ablin

Group studies involving large cohorts of subjects are important to draw general conclusions about brain functional organization. However, the aggregation of data coming from multiple subjects is challenging, since it requires accounting for large variability in anatomy, functional topography and stimulus response across individuals. Data modeling is especially hard for ecologically relevant conditions such as movie watching, where the experimental setup does not imply well-defined cognitive operations. We propose a novel MultiView Independent Component Analysis (ICA) model for group studies, where data from each subject are modeled as a linear combination of shared independent sources plus noise. Contrary to most group-ICA procedures, the likelihood of the model is available in closed form. We develop an alternate quasi-Newton method for maximizing the likelihood, which is robust and converges quickly. We demonstrate the usefulness of our approach first on fMRI data, where our model demonstrates improved sensitivity in identifying common sources among subjects. Moreover, the sources recovered by our model exhibit lower between-sessions variability than other methods. On magnetoencephalography (MEG) data, our method yields more accurate source localization on phantom data. Applied on 200 subjects from the Cam-CAN dataset, it reveals a clear sequence of evoked activity in sensor and source space.

PDF Details

YNIMG Journal 2020 Journal Article

Multi-subject MEG/EEG source imaging with sparse multi-task regression

  • Hicham Janati
  • Thomas Bazeille
  • Bertrand Thirion
  • Marco Cuturi
  • Alexandre Gramfort

Magnetoencephalography and electroencephalography (M/EEG) are non-invasive modalities that measure the weak electromagnetic fields generated by neural activity. Estimating the location and magnitude of the current sources that generated these electromagnetic fields is an inverse problem. Although it can be cast as a linear regression, this problem is severely ill-posed as the number of observations, which equals the number of sensors, is small. When considering a group study, a common approach consists in carrying out the regression tasks independently for each subject using techniques such as MNE or sLORETA. An alternative is to jointly localize sources for all subjects taken together, while enforcing some similarity between them. By pooling S subjects in a single joint regression, the number of observations is S times larger, potentially making the problem better posed and offering the ability to identify more sources with greater precision. Here we show how the coupling of the different regression problems can be done through a multi-task regularization that promotes focal source estimates. To take into account intersubject variabilities, we propose the Minimum Wasserstein Estimates (MWE). Thanks to a new joint regression method based on optimal transport (OT) metrics, MWE does not enforce perfect overlap of activation foci for all subjects but rather promotes spatial proximity on the cortical mantle. Besides, by estimating the noise level of each subject, MWE copes with the subject-specific signal-to-noise ratios with only one regularization parameter. On realistic simulations, MWE decreases the localization error by up to 4 ​mm per source compared to individual solutions. Experiments on the Cam-CAN dataset show improvements in spatial specificity in population imaging compared to individual models such as dSPM as well as a state-of-the-art Bayesian group level model. Our analysis of a multimodal dataset shows how multi-subject source localization reduces the gap between MEG and fMRI for brain mapping.

Details DOI

NeurIPS Conference 2020 Conference Paper

Statistical control for spatio-temporal MEG/EEG source imaging with desparsified mutli-task Lasso

  • Jerome-Alexis Chevalier
  • Joseph Salmon
  • Alexandre Gramfort
  • Bertrand Thirion

Detecting where and when brain regions activate in a cognitive task or in a given clinical condition is the promise of non-invasive techniques like magnetoencephalography (MEG) or electroencephalography (EEG). This problem, referred to as source localization, or source imaging, poses however a high-dimensional statistical inference challenge. While sparsity promoting regularizations have been proposed to address the regression problem, it remains unclear how to ensure statistical control of false detections in this setting. Moreover, MEG/EEG source imaging requires to work with spatio-temporal data and autocorrelated noise. To deal with this, we adapt the desparsified Lasso estimator ---an estimator tailored for high dimensional linear model that asymptotically follows a Gaussian distribution under sparsity and moderate feature correlation assumptions--- to temporal data corrupted with autocorrelated noise. We call it the desparsified multi-task Lasso (d-MTLasso). We combine d-MTLasso with spatially constrained clustering to reduce data dimension and with ensembling to mitigate the arbitrary choice of clustering; the resulting estimator is called ensemble of clustered desparsified multi-task Lasso (ecd-MTLasso). With respect to the current procedures, the two advantages of ecd-MTLasso are that i)it offers statistical guarantees and ii)it allows to trade spatial specificity for sensitivity, leading to a powerful adaptive method. Extensive simulations on realistic head geometries, as well as empirical results on various MEG datasets, demonstrate the high recovery performance of ecd-MTLasso and its primary practical benefit: offer a statistically principled way to threshold MEG/EEG source maps.

PDF Details

YNIMG Journal 2019 Journal Article

Benchmarking functional connectome-based predictive models for resting-state fMRI

  • Kamalaker Dadi
  • Mehdi Rahim
  • Alexandre Abraham
  • Darya Chyzhyk
  • Michael Milham
  • Bertrand Thirion
  • Gaël Varoquaux

Functional connectomes reveal biomarkers of individual psychological or clinical traits. However, there is great variability in the analytic pipelines typically used to derive them from rest-fMRI cohorts. Here, we consider a specific type of studies, using predictive models on the edge weights of functional connectomes, for which we highlight the best modeling choices. We systematically study the prediction performances of models in 6 different cohorts and a total of 2000 individuals, encompassing neuro-degenerative (Alzheimer's, Post-traumatic stress disorder), neuro-psychiatric (Schizophrenia, Autism), drug impact (Cannabis use) clinical settings and psychological trait (fluid intelligence). The typical prediction procedure from rest-fMRI consists of three main steps: defining brain regions, representing the interactions, and supervised learning. For each step we benchmark typical choices: 8 different ways of defining regions –either pre-defined or generated from the rest-fMRI data– 3 measures to build functional connectomes from the extracted time-series, and 10 classification models to compare functional interactions across subjects. Our benchmarks summarize more than 240 different pipelines and outline modeling choices that show consistent prediction performances in spite of variations in the populations and sites. We find that regions defined from functional data work best; that it is beneficial to capture between-region interactions with tangent-based parametrization of covariances, a midway between correlations and partial correlation; and that simple linear predictors such as a logistic regression give the best predictions. Our work is a step forward to establishing reproducible imaging-based biomarkers for clinical settings.

Details DOI

ICML Conference 2019 Conference Paper

Feature Grouping as a Stochastic Regularizer for High-Dimensional Structured Data

  • Sergül Aydöre
  • Bertrand Thirion
  • Gaël Varoquaux

In many applications where collecting data is expensive, for example neuroscience or medical imaging, the sample size is typically small compared to the feature dimension. These datasets call for intelligent regularization that exploits known structure, such as correlations between the features arising from the measurement device. However, existing structured regularizers need specially crafted solvers, which are difficult to apply to complex models. We propose a new regularizer specifically designed to leverage structure in the data in a way that can be applied efficiently to complex models. Our approach relies on feature grouping, using a fast clustering algorithm inside a stochastic gradient descent loop: given a family of feature groupings that capture feature covariations, we randomly select these groups at each iteration. Experiments on two real-world datasets demonstrate that the proposed approach produces models that generalize better than those trained with conventional regularizers, and also improves convergence speed, and has a linear computational cost.

Details

YNIMG Journal 2019 Journal Article

The functional database of the ARCHI project: Potential and perspectives

  • Philippe Pinel
  • Baudouin Forgeot d’Arc
  • Stanislas Dehaene
  • Thomas Bourgeron
  • Bertrand Thirion
  • Denis Le Bihan
  • Cyril Poupon

More than two decades of functional magnetic resonance imaging (fMRI) of the human brain have succeeded to identify, with a growing level of precision, the neural basis of multiple cognitive skills within various domains (perception, sensorimotor processes, language, emotion and social cognition …). Progress has been made in the comprehension of the functional organization of localized brain areas. However, the long time required for fMRI acquisition limits the number of experimental conditions performed in a single individual. As a consequence, distinct brain localizations have mostly been studied in separate groups of participants, and their functional relationships at the individual level remain poorly understood. To address this issue, we report here preliminary results on a database of fMRI data acquired on 78 individuals who each performed a total of 29 experimental conditions, grouped in 4 cross-domains functional localizers. This protocol has been designed to efficiently isolate, in a single session, the brain activity associated with language, numerical representation, social perception and reasoning, premotor and visuomotor representations. Analyses are reported at the group and at the individual level, to establish the ability of our protocol to selectively capture distinct regions of interest in a very short time. Test-retest reliability was assessed in a subset of participants. The activity evoked by the different contrasts of the protocol is located in distinct brain networks that, individually, largely replicate previous findings and, taken together, cover a large proportion of the cortical surface. We provide detailed analyses of a subset of regions of relevance: the left frontal, left temporal and middle frontal cortices. These preliminary analyses highlight how combining such a large set of functional contrasts may contribute to establish a finer-grained brain atlas of cognitive functions, especially in regions of high functional overlap. Detailed structural images (structural connectivity, micro-structures, axonal diameter) acquired in the same individuals in the context of the ARCHI database provide a promising situation to explore functional/structural interdependence. Additionally, this protocol might also be used as a way to establish individual neurofunctional signatures in large cohorts.

Details DOI

YNIMG Journal 2018 Journal Article

Decoding fMRI activity in the time domain improves classification performance

  • João Loula
  • Gaël Varoquaux
  • Bertrand Thirion

Most current functional Magnetic Resonance Imaging (fMRI) decoding analyses rely on statistical summaries of the data resulting from a deconvolution approach: each stimulation event is associated with a brain response. This standard approach leads to simple learning procedures, yet it is ill-suited for decoding events with short inter-stimulus intervals. In order to overcome this issue, we propose a novel framework that separates the spatial and temporal components of the prediction by decoding the fMRI time-series continuously, i. e. scan-by-scan. The stimulation events can then be identified through a deconvolution of the reconstructed time series. We show that this model performs as well as or better than standard approaches across several datasets, most notably in regimes with small inter-stimuli intervals (3–5s), while also offering predictions that are highly interpretable in the time domain. This opens the way toward analyzing datasets not normally thought of as suitable for decoding and makes it possible to run decoding on studies with reduced scan time.

Details DOI

YNIMG Journal 2018 Journal Article

FReM – Scalable and stable decoding with fast regularized ensemble of models

  • Andrés Hoyos-Idrobo
  • Gaël Varoquaux
  • Yannick Schwartz
  • Bertrand Thirion

Brain decoding relates behavior to brain activity through predictive models. These are also used to identify brain regions involved in the cognitive operations related to the observed behavior. Training such multivariate models is a high-dimensional statistical problem that calls for suitable priors. State of the art priors –eg small total-variation– enforce spatial structure on the maps to stabilize them and improve prediction. However, they come with a hefty computational cost. We build upon very fast dimension reduction with spatial structure and model ensembling to achieve decoders that are fast on large datasets and increase the stability of the predictions and the maps. Our approach, fast regularized ensemble of models (FReM), includes an implicit spatial regularization by using a voxel grouping with a fast clustering algorithm. In addition, it aggregates different estimators obtained across splits of a cross-validation loop, each time keeping the best possible model. Experiments on a large number of brain imaging datasets show that our combination of voxel clustering and model ensembling improves decoding maps stability and reduces the variance of prediction accuracy. Importantly, our method requires less samples than state-of-the-art methods to achieve a given level of prediction accuracy. Finally, FreM is much faster than other spatially-regularized methods and, in addition, it can better exploit parallel computing resources.

Details DOI

YNIMG Journal 2017 Journal Article

Assessing and tuning brain decoders: Cross-validation, caveats, and guidelines

  • Gaël Varoquaux
  • Pradeep Reddy Raamana
  • Denis A. Engemann
  • Andrés Hoyos-Idrobo
  • Yannick Schwartz
  • Bertrand Thirion

Decoding, i. e. prediction from brain images or signals, calls for empirical evaluation of its predictive power. Such evaluation is achieved via cross-validation, a method also used to tune decoders' hyper-parameters. This paper is a review on cross-validation procedures for decoding in neuroimaging. It includes a didactic overview of the relevant theoretical considerations. Practical aspects are highlighted with an extensive empirical study of the common decoders in within- and across-subject predictions, on multiple datasets –anatomical and functional MRI and MEG– and simulations. Theory and experiments outline that the popular “leave-one-out” strategy leads to unstable and biased estimates, and a repeated random splits method should be preferred. Experiments outline the large error bars of cross-validation in neuroimaging settings: typical confidence intervals of 10%. Nested cross-validation can tune decoders' parameters while avoiding circularity bias. However we find that it can be favorable to use sane defaults, in particular for non-sparse decoders.

Details DOI

YNIMG Journal 2017 Journal Article

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

  • Alexandre Abraham
  • Michael P. Milham
  • Adriana Di Martino
  • R. Cameron Craddock
  • Dimitris Samaras
  • Bertrand Thirion
  • Gael Varoquaux

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.

Details DOI

YNICL Journal 2017 Journal Article

Distinct alterations in Parkinson's medication-state and disease-state connectivity

  • Bernard Ng
  • Gael Varoquaux
  • Jean Baptiste Poline
  • Bertrand Thirion
  • Michael D. Greicius
  • Kathleen L. Poston

Altered brain connectivity has been described in people with Parkinson's disease and in response to dopaminergic medications. However, it is unclear whether dopaminergic medications primarily 'normalize' disease related connectivity changes or if they induce unique alterations in brain connectivity. Further, it is unclear how these disease- and medication-associated changes in brain connectivity relate differently to specific motor manifestations of disease, such as bradykinesia/rigidity and tremor. In this study, we applied a novel covariance projection approach in combination with a bootstrapped permutation test to resting state functional MRI data from 57 Parkinson's disease and 20 healthy control participants to determine the Parkinson's medication-state and disease-state connectivity changes associated with different motor manifestations of disease. First, we identified brain connections that best classified Parkinson's disease ON versus OFF dopamine and Parkinson's disease versus healthy controls, achieving 96.9 ± 5.9% and 72.7 ± 12.4% classification accuracy, respectively. Second, we investigated the connections that significantly contribute to the classifications. We found that the connections greater in Parkinson's disease OFF compared to ON dopamine are primarily between motor (cerebellum and putamen) and posterior cortical regions, such as the posterior cingulate cortex. By contrast, connections that are greater in ON compared to OFF dopamine are between the right and left medial prefrontal cortex. We also identified the connections that are greater in healthy control compared to Parkinson's disease and found the most significant connections are associated with primary motor regions, such as the striatum and the supplementary motor area. Notably, these are different connections than those identified in Parkinson's disease OFF compared to ON. Third, we determined which of the Parkinson's medication-state and disease-state connections are associated with the severity of different motor symptoms. We found two connections correlate with both bradykinesia/rigidity severity and tremor severity, whereas four connections correlate with only bradykinesia/rigidity severity, and five connections correlate with only tremor severity. Connections that correlate with only tremor severity are anchored by the cerebellum and the supplemental motor area, but only those connections that include the supplemental motor area predict dopaminergic improvement in tremor. Our results suggest that dopaminergic medications do not simply 'normalize' abnormal brain connectivity associated with Parkinson's disease, but rather dopamine drives distinct connectivity changes, only some of which are associated with improved motor symptoms. In addition, the dissociation between of connections related to severity of bradykinesia/rigidity versus tremor highlights the distinct abnormalities in brain circuitry underlying these specific motor symptoms.

Details DOI

YNIMG Journal 2017 Journal Article

Joint prediction of multiple scores captures better individual traits from brain images

  • Mehdi Rahim
  • Bertrand Thirion
  • Danilo Bzdok
  • Irène Buvat
  • Gaël Varoquaux

To probe individual variations in brain organization, population imaging relates features of brain images to rich descriptions of the subjects such as genetic information or behavioral and clinical assessments. Capturing common trends across these measurements is important: they jointly characterize the disease status of patient groups. In particular, mapping imaging features to behavioral scores with predictive models opens the way toward more precise diagnosis. Here we propose to jointly predict all the dimensions (behavioral scores) that make up the individual profiles, using so-called multi-output models. This approach often boosts prediction accuracy by capturing latent shared information across scores. We demonstrate the efficiency of multi-output models on two independent resting-state fMRI datasets targeting different brain disorders (Alzheimer's Disease and schizophrenia). Furthermore, the model with joint prediction generalizes much better to a new cohort: a model learned on one study is more accurately transferred to an independent one. Finally, we show how multi-output models can easily be extended to multi-modal settings, combining heterogeneous data sources for a better overall accuracy.

Details DOI

NeurIPS Conference 2017 Conference Paper

Learning Neural Representations of Human Cognition across Many fMRI Studies

  • Arthur Mensch
  • Julien Mairal
  • Danilo Bzdok
  • Bertrand Thirion
  • Gael Varoquaux

Cognitive neuroscience is enjoying rapid increase in extensive public brain-imaging datasets. It opens the door to large-scale statistical models. Finding a unified perspective for all available data calls for scalable and automated solutions to an old challenge: how to aggregate heterogeneous information on brain function into a universal cognitive system that relates mental operations/cognitive processes/psychological tasks to brain networks? We cast this challenge in a machine-learning approach to predict conditions from statistical brain maps across different studies. For this, we leverage multi-task learning and multi-scale dimension reduction to learn low-dimensional representations of brain images that carry cognitive information and can be robustly associated with psychological stimuli. Our multi-dataset classification model achieves the best prediction performance on several large reference datasets, compared to models without cognitive-aware low-dimension representations; it brings a substantial performance boost to the analysis of small datasets, and can be introspected to identify universal template cognitive concepts.

PDF Details

YNIMG Journal 2017 Journal Article

Seeing it all: Convolutional network layers map the function of the human visual system

  • Michael Eickenberg
  • Alexandre Gramfort
  • Gaël Varoquaux
  • Bertrand Thirion

Convolutional networks used for computer vision represent candidate models for the computations performed in mammalian visual systems. We use them as a detailed model of human brain activity during the viewing of natural images by constructing predictive models based on their different layers and BOLD fMRI activations. Analyzing the predictive performance across layers yields characteristic fingerprints for each visual brain region: early visual areas are better described by lower level convolutional net layers and later visual areas by higher level net layers, exhibiting a progression across ventral and dorsal streams. Our predictive model generalizes beyond brain responses to natural images. We illustrate this on two experiments, namely retinotopy and face-place oppositions, by synthesizing brain activity and performing classical brain mapping upon it. The synthesis recovers the activations observed in the corresponding fMRI studies, showing that this deep encoding model captures representations of brain function that are universal across experimental paradigms.

Details DOI

ICML Conference 2016 Conference Paper

Dictionary Learning for Massive Matrix Factorization

  • Arthur Mensch
  • Julien Mairal
  • Bertrand Thirion
  • Gaël Varoquaux

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.

Details

NeurIPS Conference 2016 Conference Paper

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

  • Elvis Dohmatob
  • Arthur Mensch
  • Gael Varoquaux
  • Bertrand Thirion

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.

PDF Details

YNIMG Journal 2015 Journal Article

Data-driven HRF estimation for encoding and decoding models

  • Fabian Pedregosa
  • Michael Eickenberg
  • Philippe Ciuciu
  • Bertrand Thirion
  • Alexandre Gramfort

Despite the common usage of a canonical, data-independent, hemodynamic response function (HRF), it is known that the shape of the HRF varies across brain regions and subjects. This suggests that a data-driven estimation of this function could lead to more statistical power when modeling BOLD fMRI data. However, unconstrained estimation of the HRF can yield highly unstable results when the number of free parameters is large. We develop a method for the joint estimation of activation and HRF by means of a rank constraint, forcing the estimated HRF to be equal across events or experimental conditions, yet permitting it to differ across voxels. Model estimation leads to an optimization problem that we propose to solve with an efficient quasi-Newton method, exploiting fast gradient computations. This model, called GLM with Rank-1 constraint (R1-GLM), can be extended to the setting of GLM with separate designs which has been shown to improve decoding accuracy in brain activity decoding experiments. We compare 10 different HRF modeling methods in terms of encoding and decoding scores on two different datasets. Our results show that the R1-GLM model outperforms competing methods in both encoding and decoding settings, positioning it as an attractive method both from the points of view of accuracy and computational efficiency.

Details DOI

YNIMG Journal 2015 Journal Article

Robust regression for large-scale neuroimaging studies

  • Virgile Fritsch
  • Benoit Da Mota
  • Eva Loth
  • Gaël Varoquaux
  • Tobias Banaschewski
  • Gareth J. Barker
  • Arun L.W. Bokde
  • Rüdiger Brühl

Multi-subject datasets used in neuroimaging group studies have a complex structure, as they exhibit non-stationary statistical properties across regions and display various artifacts. While studies with small sample sizes can rarely be shown to deviate from standard hypotheses (such as the normality of the residuals) due to the poor sensitivity of normality tests with low degrees of freedom, large-scale studies (e. g. >100 subjects) exhibit more obvious deviations from these hypotheses and call for more refined models for statistical inference. Here, we demonstrate the benefits of robust regression as a tool for analyzing large neuroimaging cohorts. First, we use an analytic test based on robust parameter estimates; based on simulations, this procedure is shown to provide an accurate statistical control without resorting to permutations. Second, we show that robust regression yields more detections than standard algorithms using as an example an imaging genetics study with 392 subjects. Third, we show that robust regression can avoid false positives in a large-scale analysis of brain–behavior relationships with over 1500 subjects. Finally we embed robust regression in the Randomized Parcellation Based Inference (RPBI) method and demonstrate that this combination further improves the sensitivity of tests carried out across the whole brain. Altogether, our results show that robust procedures provide important advantages in large-scale neuroimaging group studies.

Details DOI

NeurIPS Conference 2015 Conference Paper

Semi-Supervised Factored Logistic Regression for High-Dimensional Neuroimaging Data

  • Danilo Bzdok
  • Michael Eickenberg
  • Olivier Grisel
  • Bertrand Thirion
  • Gael Varoquaux

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.

PDF Details

YNIMG Journal 2014 Journal Article

Randomized parcellation based inference

  • Benoit Da Mota
  • Virgile Fritsch
  • Gaël Varoquaux
  • Tobias Banaschewski
  • Gareth J. Barker
  • Arun L.W. Bokde
  • Uli Bromberg
  • Patricia Conrod

Neuroimaging group analyses are used to relate inter-subject signal differences observed in brain imaging with behavioral or genetic variables and to assess risks factors of brain diseases. The lack of stability and of sensitivity of current voxel-based analysis schemes may however lead to non-reproducible results. We introduce a new approach to overcome the limitations of standard methods, in which active voxels are detected according to a consensus on several random parcellations of the brain images, while a permutation test controls the false positive risk. Both on synthetic and real data, this approach shows higher sensitivity, better accuracy and higher reproducibility than state-of-the-art methods. In a neuroimaging–genetic application, we find that it succeeds in detecting a significant association between a genetic variant next to the COMT gene and the BOLD signal in the left thalamus for a functional Magnetic Resonance Imaging contrast associated with incorrect responses of the subjects from a Stop Signal Task protocol.

Details DOI

NeurIPS Conference 2013 Conference Paper

Mapping paradigm ontologies to and from the brain

  • Yannick Schwartz
  • Bertrand Thirion
  • Gael Varoquaux

Imaging neuroscience links brain activation maps to behavior and cognition via correlational studies. Due to the nature of the individual experiments, based on eliciting neural response from a small number of stimuli, this link is incomplete, and unidirectional from the causal point of view. To come to conclusions on the function implied by the activation of brain regions, it is necessary to combine a wide exploration of the various brain functions and some inversion of the statistical inference. Here we introduce a methodology for accumulating knowledge towards a bidirectional link between observed brain activity and the corresponding function. We rely on a large corpus of imaging studies and a predictive engine. Technically, the challenges are to find commonality between the studies without denaturing the richness of the corpus. The key elements that we contribute are labeling the tasks performed with a cognitive ontology, and modeling the long tail of rare paradigms in the corpus. To our knowledge, our approach is the first demonstration of predicting the cognitive content of completely new brain images. To that end, we propose a method that predicts the experimental paradigms across different studies.

PDF Details

YNIMG Journal 2012 Journal Article

An empirical comparison of surface-based and volume-based group studies in neuroimaging

  • Alan Tucholka
  • Virgile Fritsch
  • Jean-Baptiste Poline
  • Bertrand Thirion

Being able to detect reliably functional activity in a population of subjects is crucial in human brain mapping, both for the understanding of cognitive functions in normal subjects and for the analysis of patient data. The usual approach proceeds by normalizing brain volumes to a common three-dimensional template. However, a large part of the data acquired in fMRI aims at localizing cortical activity, and methods working on the cortical surface may provide better inter-subject registration than the standard procedures that process the data in the volume. Nevertheless, few assessments of the performance of surface-based (2D) versus volume-based (3D) procedures have been shown so far, mostly because inter-subject cortical surface maps are not easily obtained. In this paper we present a systematic comparison of 2D versus 3D group-level inference procedures, by using cluster-level and voxel-level statistics assessed by permutation, in random effects (RFX) and mixed-effects analyses (MFX). We consider different schemes to perform meaningful comparisons between thresholded statistical maps in the volume and on the cortical surface. We find that surface-based multi-subject statistical analyses are generally more sensitive than their volume-based counterpart, in the sense that they detect slightly denser networks of regions when performing peak-level detection; this effect is less clear for cluster-level inference and is reduced by smoothing. Surface-based inference also increases the reliability of the activation maps.

Details DOI

YNIMG Journal 2012 Journal Article

Significant correlation between a set of genetic polymorphisms and a functional brain network revealed by feature selection and sparse Partial Least Squares

  • Édith Le Floch
  • Vincent Guillemot
  • Vincent Frouin
  • Philippe Pinel
  • Christophe Lalanne
  • Laura Trinchera
  • Arthur Tenenhaus
  • Antonio Moreno

Brain imaging is increasingly recognised as an intermediate phenotype to understand the complex path between genetics and behavioural or clinical phenotypes. In this context, a first goal is to propose methods to identify the part of genetic variability that explains some neuroimaging variability. Classical univariate approaches often ignore the potential joint effects that may exist between genes or the potential covariations between brain regions. In this paper, we propose instead to investigate an exploratory multivariate method in order to identify a set of Single Nucleotide Polymorphisms (SNPs) covarying with a set of neuroimaging phenotypes derived from functional Magnetic Resonance Imaging (fMRI). Recently, Partial Least Squares (PLS) regression or Canonical Correlation Analysis (CCA) have been proposed to analyse DNA and transcriptomics. Here, we propose to transpose this idea to the DNA vs. imaging context. However, in very high-dimensional settings like in imaging genetics studies, such multivariate methods may encounter overfitting issues. Thus we investigate the use of different strategies of regularisation and dimension reduction techniques combined with PLS or CCA to face the very high dimensionality of imaging genetics studies. We propose a comparison study of the different strategies on a simulated dataset first and then on a real dataset composed of 94 subjects, around 600, 000 SNPs and 34 functional MRI lateralisation indexes computed from reading and speech comprehension contrast maps. We estimate the generalisability of the multivariate association with a cross-validation scheme and demonstrate the significance of this link, using a permutation procedure. Univariate selection appears to be necessary to reduce the dimensionality. However, the significant association uncovered by this two-step approach combining univariate filtering and L1-regularised PLS suggests that discovering meaningful genetic associations calls for a multivariate approach.

Details DOI

YNIMG Journal 2012 Journal Article

Very large fMRI study using the IMAGEN database: Sensitivity–specificity and population effect modeling in relation to the underlying anatomy

  • Benjamin Thyreau
  • Yannick Schwartz
  • Bertrand Thirion
  • Vincent Frouin
  • Eva Loth
  • Sabine Vollstädt-Klein
  • Tomas Paus
  • Eric Artiges

In this paper we investigate the use of classical fMRI Random Effect (RFX) group statistics when analyzing a very large cohort and the possible improvement brought from anatomical information. Using 1326 subjects from the IMAGEN study, we first give a global picture of the evolution of the group effect t-value from a simple face-watching contrast with increasing cohort size. We obtain a wide activated pattern, far from being limited to the reasonably expected brain areas, illustrating the difference between statistical significance and practical significance. This motivates us to inject tissue-probability information into the group estimation, we model the BOLD contrast using a matter-weighted mixture of Gaussians and compare it to the common, single-Gaussian model. In both cases, the model parameters are estimated per-voxel for one subgroup, and the likelihood of both models is computed on a second, separate subgroup to reflect model generalization capacity. Various group sizes are tested, and significance is asserted using a 10-fold cross-validation scheme. We conclude that adding matter information consistently improves the quantitative analysis of BOLD responses in some areas of the brain, particularly those where accurate inter-subject registration remains challenging.

Details DOI

YNIMG Journal 2011 Journal Article

Phase delays within visual cortex shape the response to steady-state visual stimulation

  • Benoit Cottereau
  • Jean Lorenceau
  • Alexandre Gramfort
  • Maureen Clerc
  • Bertrand Thirion
  • Sylvain Baillet

Although the spatial organization of visual areas can be revealed by functional Magnetic Resonance Imaging (fMRI), the synoptic, non-invasive access to the temporal characteristics of the information flow amongst distributed visual processes remains a technical and methodological challenge. Using frequency-encoded steady-state visual stimulation together with a combination of time-resolved functional magnetic source imaging from magnetoencephalography (MEG) and anatomical magnetic resonance imaging (MRI), this study evidences maps of visuotopic sustained oscillatory neural responses distributed across the visual cortex. Our results further reveal relative phase delays across responding striate and extra-striate visual areas, which thereby shape the chronometry of neural processes amongst these regions. The methodology developed in this study points at further developments in time-resolved analyses of distributed visual processes in the millisecond range, and to new ways of exploring the dynamics of functional processes within the human visual cortex non-invasively.

Details DOI

JMLR Journal 2011 Journal Article

Scikit-learn: Machine Learning in Python

  • Fabian Pedregosa
  • Gaël Varoquaux
  • Alexandre Gramfort
  • Vincent Michel
  • Bertrand Thirion
  • Olivier Grisel
  • Mathieu Blondel
  • Peter Prettenhofer

Scikit-learn is a Python module integrating a wide range of state-of-the-art machine learning algorithms for medium-scale supervised and unsupervised problems. This package focuses on bringing machine learning to non-specialists using a general-purpose high-level language. Emphasis is put on ease of use, performance, documentation, and API consistency. It has minimal dependencies and is distributed under the simplified BSD license, encouraging its use in both academic and commercial settings. Source code, binaries, and documentation can be downloaded from http://scikit-learn.sourceforge.net. [abs] [ pdf ][ bib ] [ code ] &copy JMLR 2011. ( edit, beta )

PDF Details

NeurIPS Conference 2010 Conference Paper

Brain covariance selection: better individual functional connectivity models using population prior

  • Gael Varoquaux
  • Alexandre Gramfort
  • Jean-Baptiste Poline
  • Bertrand Thirion

Spontaneous brain activity, as observed in functional neuroimaging, has been shown to display reproducible structure that expresses brain architecture and carries markers of brain pathologies. An important view of modern neuroscience is that such large-scale structure of coherent activity reflects modularity properties of brain connectivity graphs. However, to date, there has been no demonstration that the limited and noisy data available in spontaneous activity observations could be used to learn full-brain probabilistic models that generalize to new data. Learning such models entails two main challenges: i) modeling full brain connectivity is a difficult estimation problem that faces the curse of dimensionality and ii) variability between subjects, coupled with the variability of functional signals between experimental runs, makes the use of multiple datasets challenging. We describe subject-level brain functional connectivity structure as a multivariate Gaussian process and introduce a new strategy to estimate it from group data, by imposing a common structure on the graphical model in the population. We show that individual models learned from functional Magnetic Resonance Imaging (fMRI) data using this population prior generalize better to unseen data than models based on alternative regularization schemes. To our knowledge, this is the first report of a cross-validated model of spontaneous brain activity. Finally, we use the estimated graphical model to explore the large-scale characteristics of functional architecture and show for the first time that known cognitive networks appear as the integrated communities of functional connectivity graph.

PDF Details

NeurIPS Conference 2009 Conference Paper

Discriminative Network Models of Schizophrenia

  • Irina Rish
  • Benjamin Thyreau
  • Bertrand Thirion
  • Marion Plaze
  • Marie-laure Paillere-martinot
  • Catherine Martelli
  • Jean-Luc Martinot
  • Jean-Baptiste Poline

Schizophrenia is a complex psychiatric disorder that has eluded a characterization in terms of local abnormalities of brain activity, and is hypothesized to affect the collective, ``emergent working of the brain. We propose a novel data-driven approach to capture emergent features using functional brain networks [Eguiluzet al] extracted from fMRI data, and demonstrate its advantage over traditional region-of-interest (ROI) and local, task-specific linear activation analyzes. Our results suggest that schizophrenia is indeed associated with disruption of global, emergent brain properties related to its functioning as a network, which cannot be explained by alteration of local activation patterns. Moreover, further exploitation of interactions by sparse Markov Random Field classifiers shows clear gain over linear methods, such as Gaussian Naive Bayes and SVM, allowing to reach 86% accuracy (over 50% baseline - random guess), which is quite remarkable given that it is based on a single fMRI experiment using a simple auditory task.

PDF Details