Robust regression for large-scale neuroimaging studies.

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.

Randomized parcellation based inference.

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.

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

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.

Oxytocin receptor genotype modulates ventral striatal activity to social cues and response to stressful life events.

BACKGROUND: Common variants in the oxytocin receptor gene (OXTR) have been shown to influence social and affective behavior and to moderate the effect of adverse experiences on risk for social-affective problems. However, the intermediate neurobiological mechanisms are not fully understood. Although human functional neuroimaging studies have reported that oxytocin effects on social behavior and emotional states are mediated by amygdala function, animal models indicate that oxytocin receptors in the ventral striatum (VS) modulate sensitivity to social reinforcers. This study aimed to comprehensively investigate OXTR-dependent brain mechanisms associated with social-affective problems. METHODS: In a sample of 1445 adolescents we tested the effect of 23-tagging single nucleotide polymorphisms across the OXTR region and stressful life events (SLEs) on functional magnetic resonance imaging blood oxygen level-dependent activity in the VS and amygdala to animated angry faces. Single nucleotide polymorphisms for which gene-wide significant effects on brain function were found were then carried forward to examine associations with social-affective problems. RESULTS: A gene-wide significant effect of rs237915 showed that adolescents with minor CC-genotype had significantly lower VS activity than CT/TT-carriers. Significant or nominally significant gene × environment effects on emotional problems (in girls) and peer problems (in boys) revealed a strong increase in clinical symptoms as a function of SLEs in CT/TT-carriers but not CC-homozygotes. However, in low-SLE environments, CC-homozygotes had more emotional problems (girls) and peer problems (boys). Moreover, among CC-homozygotes, reduced VS activity was related to more peer problems. CONCLUSIONS: These findings suggest that a common OXTR-variant affects brain responsiveness to negative social cues and that in "risk-carriers" reduced sensitivity is simultaneously associated with more social-affective problems in "favorable environments" and greater resilience against stressful experiences.

Enhancing the reproducibility of group analysis with randomized brain parcellations.

Neuroimaging group analyses are used to compare the intersubject variability observed in brain organization with behavioural 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. A new approach is introduced 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 syntetic and real data, this approach shows higher sensitivity, better recovery and higher reproducibility than standard methods and succeeds in detecting a significant association in an imaging-genetic study between a genetic variant next to the COMT gene and a region in the left thalamus on a functional magnetic resonance imaging contrast.

Detecting outliers in high-dimensional neuroimaging datasets with robust covariance estimators.

Medical imaging datasets often contain deviant observations, the so-called outliers, due to acquisition or preprocessing artifacts or resulting from large intrinsic inter-subject variability. These can undermine the statistical procedures used in group studies as the latter assume that the cohorts are composed of homogeneous samples with anatomical or functional features clustered around a central mode. The effects of outlying subjects can be mitigated by detecting and removing them with explicit statistical control. With the emergence of large medical imaging databases, exhaustive data screening is no longer possible, and automated outlier detection methods are currently gaining interest. The datasets used in medical imaging are often high-dimensional and strongly correlated. The outlier detection procedure should therefore rely on high-dimensional statistical multivariate models. However, state-of-the-art procedures, based on the Minimum Covariance Determinant (MCD) estimator, are not well-suited for such high-dimensional settings. In this work, we introduce regularization in the MCD framework and investigate different regularization schemes. We carry out extensive simulations to provide backing for practical choices in absence of ground truth knowledge. We demonstrate on functional neuroimaging datasets that outlier detection can be performed with small sample sizes and improves group studies.

Detecting outlying subjects in high-dimensional neuroimaging datasets with regularized minimum covariance determinant.

Medical imaging datasets used in clinical studies or basic research often comprise highly variable multi-subject data. Statistically-controlled inclusion of a subject in a group study, i.e. deciding whether its images should be considered as samples from a given population or whether they should be rejected as outlier data, is a challenging issue. While the informal approaches often used do not provide any statistical assessment that a given dataset is indeed an outlier, traditional statistical procedures are not well-suited to the noisy, high-dimensional, settings encountered in medical imaging, e.g. with functional brain images. In this work, we modify the classical Minimum Covariance Determinant approach by adding a regularization term, that ensures that the estimation is well-posed in high-dimensional settings and in the presence of many outliers. We show on simulated and real data that outliers can be detected satisfactorily, even in situations where the number of dimensions of the data exceeds the number of observations.