Showing papers by "Bertrand Thirion published in 2009"

Recruitment of an Area Involved in Eye Movements During Mental Arithmetic

Abstract: Throughout the history of mathematics, concepts of number and space have been tightly intertwined. We tested the hypothesis that cortical circuits for spatial attention contribute to mental arithmetic in humans. We trained a multivariate classifier algorithm to infer the direction of an eye movement, left or right, from the brain activation measured in the posterior parietal cortex. Without further training, the classifier then generalized to an arithmetic task. Its left versus right classification could be used to sort out subtraction versus addition trials, whether performed with symbols or with sets of dots. These findings are consistent with the suggestion that mental arithmetic co-opts parietal circuitry associated with spatial coding.

Article Deciphering Cortical Number Coding from Human Brain Activity Patterns

Abstract: Summary Background: Neuropsychology and human functional neuroimaging have implicated human parietal cortex in numerical processing, and macaque electrophysiology has shown that intraparietal areas house neurons tuned to numerosity. Yet although the areas responding overall during numerical tasks have been well defined by neuroimaging, a direct demonstration of individual number coding by spatial patterns has thus far been elusive. Results: We used multivariate pattern recognition on highresolution functional imaging data to decode the information content of fine-scale signals evoked by different individual numbers. Parietal activation patterns for individual numerosities could be accurately discriminated and generalized across changes in low-level stimulus parameters. Distinct patterns were evoked by symbolic and nonsymbolic number formats, and individual digits were less accurately decoded (albeit still with significant accuracy) than numbers of dots. Interestingly, the numerosity of dot sets could be predicted above chance from the brain activation patterns evoked by digits, but not vice versa. Finally, number-evoked patterns changed in a gradual fashion as a function of numerical distance for the nonsymbolic notation, compatible with some degree of orderly layout of individual number representations. Conclusions: Our findings demonstrate partial format invariance of individual number codes that is compatible with more numerous but more broadly tuned populations for nonsymbolic than for symbolic numbers, as postulated by recent computational models. In more general terms, our results illustrate the potential of functional magnetic resonance imaging pattern recognition to understand the detailed format of representations within a single semantic category, and beyond sensory cortical areas for which columnar architectures are well established.

A disconnection account of Gerstmann syndrome: functional neuroanatomy evidence.

Abstract: Objective To examine the functional neuroanatomy that could account for pure Gerstmann syndrome, which is the selective association of acalculia, finger agnosia, left-right disorientation, and agraphia. Methods We used structural and functional neuroimaging at high spatial resolution in healthy subjects to seek a shared cortical substrate of the Grundstorung posited by Gerstmann, ie, a common functional denominator accounting for this clinical tetrad. We construed a functional activation paradigm that mirrors each of the four clinical deficits in Gerstmann syndrome and determined cortical activation patterns. We then applied fiber tracking to diffusion tensor images and used cortical activation foci in the four functional domains as seed regions. Results None of the subjects showed parietal overlap of cortical activation patterns from the four cognitive domains. In every subject, however, the parietal activation patterns across all four domains consistently connected to a small region of subcortical parietal white matter at a location that is congruent with the lesion in a well-documented case of pure Gerstmann syndrome. Interpretation Our functional neuroimaging findings are not in agreement with Gerstmann's postulate of damage to a common cognitive function underpinning clinical semiology. Our evidence from intact functional neuroanatomy suggests that pure forms of Gerstmann's tetrad do not arise from lesion to a shared cortical substrate but from intraparietal disconnection after damage to a focal region of subcortical white matter. Ann Neurol 2009;66:654–662

Discriminative Network Models of Schizophrenia

Abstract: 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 [4] 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 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.

CanICA: Model-based extraction of reproducible group-level ICA patterns from fMRI time series

Abstract: Spatial Independent Component Analysis (ICA) is an increasingly used data-driven method to analyze functional Magnetic Resonance Imaging (fMRI) data. To date, it has been used to extract meaningful patterns without prior information. However, ICA is not robust to mild data variation and remains a parameter-sensitive algorithm. The validity of the extracted patterns is hard to establish, as well as the significance of differences between patterns extracted from different groups of subjects. We start from a generative model of the fMRI group data to introduce a probabilistic ICA pattern-extraction algorithm, called CanICA (Canonical ICA). Thanks to an explicit noise model and canonical correlation analysis, our method is auto-calibrated and identifies the group-reproducible data subspace before performing ICA. We compare our method to state-of-the-art multi-subject fMRI ICA methods and show that the features extracted are more reproducible.

Adaptive multi-class Bayesian sparse regression - An application to brain activity classification

Abstract: In this article we describe a novel method for regularized regression and apply it to the prediction of a behavioural variable from brain activation images. In the context of neuroimaging, regression or classification techniques are often plagued with the curse of dimensionality, due to the extremely high number of voxels and the limited number of activation maps. A commonly-used solution is the regularization of the weights used in the parametric prediction function. It entails the difficult issue of introducing an adapted amount of regularization in the model; this question can be addressed in a Bayesian framework, but model specification needs a careful design to balance adaptiveness and sparsity. Thus, we introduce an adaptive multi-class regularization to deal with this cluster-based structure of the data. Based on a hierarchical model and estimated in a Variational Bayes framework, our algorithm is robust to overfit and more adaptive than other regularization methods. Results on simulated data and preliminary results on real data show the accuracy of the method in the context of brain activation images.

Structural and probabilistic methods for group analysis in functional neuroimaging

Abstract: Neuroimaging is currently the main modality for the non-invasive exploration of brain structure and function. In this presentation, we will focus on Magnetic Resonance Imaging (MRI) that is the only modality that provides a spatially-resolved and full coverage of the brain volume. The use of such data to better understand brain function or to diagnose various brain diseases has to face two main issues: i) dealing with heterogeneous informations such as tissue segmentation obtained through anatomical MRI, functional region characterization obtained in functional MRI and connectivity measures obtained in diffusion MRI; ii) making meaningful between-subject comparison, in spite of the impressive variability of brain shapes across individuals. We will present how to build models of brain organization while accounting for between-individual variability with various statistical tools. Finally, we will show how we address the modelling of between subjects variability by comparing neuroimaging measures to behavioural and genetic information that also characterize inter-individual variability.

CanICA: Model-based extraction of reproducible group-level ICA patterns from fMRI time series

Abstract: Spatial Independent Component Analysis (ICA) is an increasingly used data-driven method to analyze functional Magnetic Resonance Imaging (fMRI) data. To date, it has been used to extract meaningful patterns without prior information. However, ICA is not robust to mild data variation and remains a parameter-sensitive algorithm. The validity of the extracted patterns is hard to establish, as well as the significance of differences between patterns extracted from different groups of subjects. We start from a generative model of the fMRI group data to introduce a probabilistic ICA pattern-extraction algorithm, called CanICA (Canonical ICA). Thanks to an explicit noise model and canonical correlation analysis, our method is auto-calibrated and identifies the group-reproducible data subspace before performing ICA. We compare our method to state-of-the-art multi-subject fMRI ICA methods and show that the features extracted are more reproducible.

Surface-based versus volume-based fMRI group analysis: a case study

Abstract: Being able to detect reliably functional activity in a popula- tion of subjects is crucial in human brain mapping, both for the under- standing of cognitive functions in normal subjects and for the analysis of patient data The usual approach proceeds by normalizing brain volumes to a common 3D 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 3D Nevertheless, few as- sessments 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 e ects (RFX) and mixed-e ects analyses (MFX) We nd that, using a voxel-level thresholding, and to some extent, cluster-level thresholding, the surface-based approach generally detects more, but smaller active regions than the corresponding volume-based approach for both RFX and MFX procedures, and that surface-based supra-threshold regions are more reproducible by bootstrap