References of "NeuroImage"
     in
Bookmark and Share    
Full Text
Peer Reviewed
See detailSelf-referential reflective activity and its relationship with rest : a PET study
D'Argembeau, Arnaud ULg; Collette, Fabienne ULg; Van der Linden, Martial ULg et al

in NeuroImage (2005), 25(2), 616-624

This study used positron emission tomography (PET) to identify the brain substrate of self-referential reflective activity and to investigate its relationship with brain areas that are active during the ... [more ▼]

This study used positron emission tomography (PET) to identify the brain substrate of self-referential reflective activity and to investigate its relationship with brain areas that are active during the resting state. Thirteen healthy volunteers performed reflective tasks pertaining to three different matters (the self, another person, and social issues) while they were scanned. Rest scans were also acquired, in which subjects were asked to simply relax and not think in a systematic way. The mental activity experienced during each scan was assessed with rating scales. The results showed that, although self-referential thoughts were most frequent during the self-referential task, some self-referential reflective activity also occurred during rest. Compared to rest, performing the reflective tasks was associated with increased blood flow in the dorsomedial prefrontal cortex, the left anterior middle temporal gyros, the temporal pole bilaterally, and the right cerebellum; there was a decrease of blood flow in right prefrontal regions,and in medial and right lateral parietal regions. In addition, the ventromedial prefrontal cortex (VMPFC) (1) was more active during the self-referential reflective task than during the other two reflective tasks, (2) showed common activation during rest and the self-referential task, and (3) showed a correlation between cerebral metabolism and the amount of self-referential processing. It is suggested that the VMPFC is crucial for representing knowledge pertaining to the self and that this is an important function of the resting state. [less ▲]

Detailed reference viewed: 41 (6 ULg)
Full Text
Peer Reviewed
See detailCASL fMRI of subcortico-cortical perfusion changes during memory-guided finger sequences
Garraux, Gaëtan ULg; Hallett, Mark; Talagala, S Lalith

in Neuroimage (2005), 25(1), 122-132

Arterial spin labeling (ASL) perfusion functional magnetic resonance imaging (fMRI) is an attractive alternative to BOLD fMRI. Nevertheless, current ASL fMRI techniques are limited by several factors that ... [more ▼]

Arterial spin labeling (ASL) perfusion functional magnetic resonance imaging (fMRI) is an attractive alternative to BOLD fMRI. Nevertheless, current ASL fMRI techniques are limited by several factors that hamper more routine applications in humans. One of these factors is restricted brain coverage so that whole-brain ASL fMRI studies have never been reported. The present study tested the ability of a multislice continuous ASL (CASL) fMRI approach using a small surface coil placed on the subject's neck to map changes in regional cerebral blood flow (rCBF) throughout the brain while healthy individuals (N = 15) performed memory-guided sequential finger movements at a mean rate of similar to0.5 Hz. As predicted by results from a large number of studies, reliable task-related increases in flow were detected across subjects not only in primary and associative cortical areas but also in subcortical brain regions. When normalized to baseline, rCBF increased 31% in the hand representation area (HRA) of left primary motor cortex (M1) 13% in the left supplementary motor area proper (SMA), 10% in the left dorsolateral prefrontal cortex (DLPFC), 10-18% in the bilateral intraparietal sulci, 6% in the HRA of left putamen, 10% in the left thalamus, and 17% in the right anterior cerebellum. In addition to these increases, 6% and 4% decreases in rCBF were detected in the HRA of the right M1 and the bilateral posterior cingulate sulci, respectively. These results demonstrate that perfusion-based fMRI using CASL with a separate labeling coil can now be used to characterize task-related flow changes in most of the brain volume with adequate accuracy and sensitivity. [less ▲]

Detailed reference viewed: 15 (2 ULg)
Full Text
Peer Reviewed
See detailAn empirical Bayesian solution to the source reconstruction problem in EEG
Phillips, Christophe ULg; Mattout Jeremie; Rugg, Michael D et al

in Neuroimage (2005), 24(4), 997-1011

Distributed linear solutions of the EEG source localisation problem are used routinely. In contrast to discrete dipole equivalent models, distributed linear solutions do not assume a fixed number of ... [more ▼]

Distributed linear solutions of the EEG source localisation problem are used routinely. In contrast to discrete dipole equivalent models, distributed linear solutions do not assume a fixed number of active sources and rest on a discretised fully 3D representation of the electrical activity of the brain. The ensuing inverse problem is underdetermined and constraints or priors are required to ensure the uniqueness of the solution. In a Bayesian framework, the conditional expectation of the source distribution, given the data, is attained by carefully balancing the minimisation of the residuals induced by noise and the improbability of the estimates as determined by their priors. This balance is specified by hyperparameters that control the relative importance of fitting and conforming to various constraints. Here we formulate the conventional "Weighted Minimum Norm" (WMN) solution in terms of hierarchical linear models. An "Expectation-Maximisation" (EM) algorithm is used to obtain a "Restricted Maximum Likelihood" (ReML) estimate of the hyperparameters, before estimating the "Maximum a Posteriori" solution itself. This procedure can be considered a generalisation of previous work that encompasses multiple constraints. Our approach was compared with the "classic" WMN and Maximum Smoothness solutions, using a simplified 2D source model with synthetic noisy data. The ReML solution was assessed with four types of source location priors: no priors, accurate priors, inaccurate priors, and both accurate and inaccurate priors. The ReML approach proved useful as: (1) The regularisation (or influence of the a priori source covariance) increased as the noise level increased. (2) The localisation error (LE) was negligible when accurate location priors were used. (3) When accurate and inaccurate location priors were used simultaneously, the solution was not influenced by the inaccurate priors. The ReML solution was then applied to real somatosensory-evoked responses to illustrate the application in an empirical setting. (C) 2004 Elsevier Inc. All rights reserved. [less ▲]

Detailed reference viewed: 54 (1 ULg)
Full Text
Peer Reviewed
See detailCerebral correlates of delta waves during non-REM sleep revisited.
Dang Vu, Thien Thanh ULg; Desseilles, Martin ULg; Laureys, Steven ULg et al

in NeuroImage (2005), 28(1), 14-21

We aimed at characterizing the neural correlates of delta activity during Non Rapid Eye Movement (NREM) sleep in non-sleep-deprived normal young adults, based on the statistical analysis of a positron ... [more ▼]

We aimed at characterizing the neural correlates of delta activity during Non Rapid Eye Movement (NREM) sleep in non-sleep-deprived normal young adults, based on the statistical analysis of a positron emission tomography (PET) sleep data set. One hundred fifteen PET scans were obtained using H(2)(15)O under continuous polygraphic monitoring during stages 2-4 of NREM sleep. Correlations between regional cerebral blood flow (rCBF) and delta power (1.5-4 Hz) spectral density were analyzed using statistical parametric mapping (SPM2). Delta power values obtained at central scalp locations negatively correlated during NREM sleep with rCBF in the ventromedial prefrontal cortex, the basal forebrain, the striatum, the anterior insula, and the precuneus. These regions embrace the set of brain areas in which rCBF decreases during slow wave sleep (SWS) as compared to Rapid Eye Movement (REM) sleep and wakefulness (Maquet, P., Degueldre, C., Delfiore, G., Aerts, J., Peters, J.M., Luxen, A., Franck, G., 1997. Functional neuroanatomy of human slow wave sleep. J. Neurosci. 17, 2807-S2812), supporting the notion that delta activity is a valuable prominent feature of NREM sleep. A strong association was observed between rCBF in the ventromedial prefrontal regions and delta power, in agreement with electrophysiological studies. In contrast to the results of a previous PET study investigating the brain correlates of delta activity (Hofle, N., Paus, T., Reutens, D., Fiset, P., Gotman, J., Evans, A.C., Jones, B.E., 1997. Regional cerebral blood flow changes as a function of delta and spindle activity during slow wave sleep in humans. J. Neurosci. 17, 4800-4808), in which waking scans were mixed with NREM sleep scans, no correlation was found with thalamus activity. This latter result stresses the importance of an extra-thalamic delta rhythm among the synchronous NREM sleep oscillations. Consequently, this rCBF distribution might preferentially reflect a particular modulation of the cellular processes involved in the generation of cortical delta waves during NREM sleep. [less ▲]

Detailed reference viewed: 80 (15 ULg)
Full Text
Peer Reviewed
See detailA Prominent Role for Amygdaloïd Complexes in the Variability of Heart Rate during Rapid Eye Movement (REM) Sleep
Desseilles, Martin ULg; Dang Vu, Thanh; Laureys, Steven ULg et al

in NeuroImage (2005), 26(Suppl. 1),

Detailed reference viewed: 10 (0 ULg)
Full Text
Peer Reviewed
See detailTime-of-day modulations of rCBF response in functional brain imaging studies: a meta-analysis
Schmidt, Christina; Dang Vu, Thanh; Orban, Pierre et al

in NeuroImage (2005), 26(Suppl. 1),

Detailed reference viewed: 12 (2 ULg)
Full Text
Peer Reviewed
See detailNeural correlates of fast and slow ocular sequence learning
Albouy, Geneviève ULg; Ruby, Perrine; Balteau, Evelyne ULg et al

in NeuroImage (2005), 26(Suppl. 1),

Detailed reference viewed: 6 (1 ULg)
Full Text
Peer Reviewed
See detailDifferential effects of testosterone on neuronal populations and their connections in a sensorimotor brain nucleus controlling song production in songbirds: a manganese enhanced-magnetic resonance imaging study
Van Meir, V.; Verhoye, M.; Absil, Philippe ULg et al

in Neuroimage (2004), 21(3), 914-923

Nucleus HVC (formerly called high vocal center) of songbirds contains two types of projecting neurons connecting HVC respectively to the nucleus robustus archistriatalis, RA, or to area X. These two ... [more ▼]

Nucleus HVC (formerly called high vocal center) of songbirds contains two types of projecting neurons connecting HVC respectively to the nucleus robustus archistriatalis, RA, or to area X. These two neuron classes exhibit multiple neurochemical differences and are differentially replaced by new neurons during adult life: high rates of neuronal replacement are observed in RA-projecting neurons only. The activity of these two types of neurons may also be modulated differentially by steroids. We analyzed by magnetic resonance imaging the effect of testosterone on the volume of RA and area X and on the dynamics of Mn2+ accumulation in RA and area X of female starlings that had been injected with MnCl2 through a permanent cannula implanted in HVC. Repeated visualization 6 weeks apart (before and after testosterone treatment) identified a volume increase of both nuclei in testosterone-treated birds associated with a concomitant decrease in controls. Following testosterone treatment, the total amount of Mn2+ transported to RA and area X increased but the dynamics of accumulation, reflecting in part the activity of HVC neurons, was specifically altered in area X but not in RA. These data indicate that testosterone differentially affects the RA- and area X-projecting neurons in HVC. Manganese-enhanced magnetic resonance imaging (ME-MRI) thus provides repeated measures of connected brain areas and demonstrates testosterone-dependent regionally specific changes in brain activity and functional connectivity. The slow time scales investigated by this technique (compared to functional MRI) appear ideally suited for characterizing slow processes such as those involved in brain plasticity and learning. (C) 2004 Elsevier Inc. All rights reserved. [less ▲]

Detailed reference viewed: 63 (9 ULg)
Full Text
Peer Reviewed
See detailLearned material content and acquisition level modulate cerebral reactivation during posttraining rapid-eye-movements sleep
Peigneux, Philippe ULg; Laureys, Steven ULg; Fuchs, Sonia et al

in NeuroImage (2003), 20(1), 125-134

We have previously shown that several brain areas are activated both during sequence learning at wake and during subsequent rapid-eye-movements (REM) sleep (Nat. Neurosci. 3 (2000) 831-836), suggesting ... [more ▼]

We have previously shown that several brain areas are activated both during sequence learning at wake and during subsequent rapid-eye-movements (REM) sleep (Nat. Neurosci. 3 (2000) 831-836), suggesting that REM sleep participates in the reprocessing of recent memory traces in humans. However, the nature of the reprocessed information remains open. Here, we show that regional cerebral reactivation during posttraining REM sleep is not merely related to the acquisition of basic visuomotor skills during prior practice of the serial reaction time task, but rather to the implicit acquisition of the probabilistic rules that defined stimulus sequences. Moreover, functional connections between the reactivated cuneus and the striatum-the latter being critical for implicit sequence learning-are reinforced during REM sleep after practice on a probabilistic rather than on a random sequence of stimuli. Our results therefore support the hypothesis that REM sleep is deeply involved in the reprocessing and optimization of the high-order information contained in the material to be learned. In addition, we show that the level of acquisition of probabilistic rules attained prior to sleep is correlated to the increase in regional cerebral blood flow during subsequent REM sleep. This suggests that posttraining cerebral reactivation is modulated by the strength of the memory traces developed during the learning episode. Our data provide the first experimental evidence for a link between behavioral performance and cerebral reactivation during REM sleep. (C) 2003 Elsevier Inc. All rights reserved. [less ▲]

Detailed reference viewed: 35 (4 ULg)
Full Text
Peer Reviewed
See detailPredominant ventromedial frontopolar metabolic impairment in frontotemporal dementia
Salmon, Eric ULg; Garraux, Gaëtan ULg; Delbeuck, Xavier et al

in NeuroImage (2003), 20(1), 435-440

In a multicenter study, FDG-PET images in a population of 29 patients with frontotemporal dementia (FTD) were compared to controls with similar age from each center. A conjunction analysis led to ... [more ▼]

In a multicenter study, FDG-PET images in a population of 29 patients with frontotemporal dementia (FTD) were compared to controls with similar age from each center. A conjunction analysis led to identification of the ventromedial frontopolar cortex as the single region affected in each and every FTD patients. This precise regional metabolic impairment should be integrated with recent neuropsychological researches, such as those showing that the ventromedial frontal cortex is critically involved in decision-making processes based on personal experience, feelings of rightness or social knowledge, processes that are characteristically impaired in FTD. (C) 2003 Elsevier Inc. All rights reserved. [less ▲]

Detailed reference viewed: 45 (5 ULg)
Full Text
Peer Reviewed
See detailNeural correlates of "hot" and "cold" emotional processing : a multilevel approach to the functional anatomy of emotion
Schaefer, Alexandre; Collette, Fabienne ULg; Philippot, Pierre et al

in Neuroimage (2003), 18(4), 938-949

The neural correlates of two hypothesized emotional processing modes, i.e., schematic and propositional modes, were investigated with positron emission tomography. Nineteen subjects performed an emotional ... [more ▼]

The neural correlates of two hypothesized emotional processing modes, i.e., schematic and propositional modes, were investigated with positron emission tomography. Nineteen subjects performed an emotional mental imagery task while mentally repeating sentences linked to the meaning of the imagery script. In the schematic conditions, participants repeated metaphoric sentences, whereas in the propositional conditions, the sentences were explicit questions about specific emotional appraisals of the imagery scenario. Five types of emotional scripts were proposed to the subjects (happiness, anger, affection, sadness, and a neutral scenario). The results supported the hypothesized distinction between schematic and propositional emotional processing modes. Specifically, schematic mode was associated with increased activity in the ventromedial prefrontal cortex whereas propositional mode was associated with activation of the anterolateral prefrontal cortex. In addition, interaction analyses showed that schematic versus propositional processing of happiness (compared with the neutral scenario) was associated with increased activity in the ventral striatum whereas "schematic anger" was tentatively associated with activation of the ventral pallidum. (C) 2003 Elsevier Science (USA). All rights reserved. [less ▲]

Detailed reference viewed: 10 (0 ULg)
Full Text
Peer Reviewed
See detailDevelopment and evaluation of an automated atlas-based image analysis method for microPET studies of the rat brain.
Rubins, Daniel J.; Melega, William P.; Lacan, Goran et al

in NeuroImage (2003), 20(4), 2100-18

An automated method for placement of 3D rat brain atlas-derived volumes of interest (VOIs) onto PET studies has been designed and evaluated. VOIs representing major structures of the rat brain were ... [more ▼]

An automated method for placement of 3D rat brain atlas-derived volumes of interest (VOIs) onto PET studies has been designed and evaluated. VOIs representing major structures of the rat brain were defined on a set of digitized cryosectioned images of the rat brain. For VOI placement, each PET study was registered with a synthetic PET target constructed from the VOI template. Registration was accomplished with an automated algorithm that maximized the mutual information content of the image volumes. The accuracy and precision of this method for VOI placement was determined using datasets from PET studies of the striatal dopamine and hippocampal serotonin systems. Each evaluated PET study could be registered to at least one synthetic PET target without obvious failure. Registration was critically dependent upon the initial position of the PET study relative to the synthetic PET target, but not dependent on the amount of synthetic PET target smoothing. An evaluation algorithm showed that resultant radioactivity concentration measurements of selected brain structures had errors=2% due to misalignment with the corresponding VOI. Further, radioligand binding values calculated from these measurements were found to be more precise than those calculated from measurements obtained with manually drawn regions of interest (ROIs). Overall, evaluation results demonstrated that this atlas-derived VOI method can be used to obtain unbiased measurements of radioactivity concentration from PET studies. Its automated features, and applicability to different radioligands and brain regions, will facilitate quantitative rat brain PET assessment procedures. [less ▲]

Detailed reference viewed: 36 (5 ULg)
Full Text
Peer Reviewed
See detailCortical processing of noxious somatosensory stimuli in the persistent vegetative state
Laureys, Steven ULg; Faymonville, Marie-Elisabeth ULg; Peigneux, Philippe ULg et al

in NeuroImage (2002), 17(2), 732-741

The persistent vegetative state (PVS) is a devastating medical condition characterized by preserved wakefulness contrasting with absent voluntary interaction with the environment. We used positron ... [more ▼]

The persistent vegetative state (PVS) is a devastating medical condition characterized by preserved wakefulness contrasting with absent voluntary interaction with the environment. We used positron emission tomography to assess the central processing of noxious somatosensory stimuli in the PVS. Changes in regional cerebral blood flow were measured during high-intensity electrical stimulation of the median nerve compared with rest in 15 nonsedated patients and in 15 healthy controls. Evoked potentials were recorded simultaneously. The stimuli were experienced as highly unpleasant to painful in controls. Brain glucose metabolism was also studied with [F-18]fluorodeoxyglucose in resting conditions. In PVS patients, overall cerebral metabolism was 40% of normal values. Nevertheless, noxious somatosensory stimulation-activated midbrain, contralateral thalamus, and primary somatosensory cortex in each and every PVS patient, even in the absence of detectable cortical evoked potentials. Secondary somatosensory, bilateral insular, posterior parietal, and anterior cingulate cortices did not show activation in any patient. Moreover, in PVS patients, the activated primary somatosensory cortex was functionally disconnected from secondary somatosensory, bilateral posterior parietal, premotor, polysensory superior temporal, and prefrontal cortices. In conclusion, somatosensory stimulation of PVS patients, at intensities that elicited pain in controls, resulted in increased neuronal activity in primary somatosensory cortex, even if resting brain metabolism was severely impaired. However, this activation of primary cortex seems to be isolated and dissociated from higher-order associative cortices. (C) 2002 Elsevier Science (USA). [less ▲]

Detailed reference viewed: 71 (18 ULg)
Full Text
Peer Reviewed
See detailDiscrimination between Alzheimer dementia and controls by automated analysis of multicenter FDG PET
Herholz, K.; Salmon, Eric ULg; Perani, D. et al

in Neuroimage (2002), 17(1), 302-316

A new diagnostic indicator of FDG PET scan abnormality, based on age-adjusted t statistics and an automated voxel-based procedure, is presented and validated in a large data set comprising 110 normal ... [more ▼]

A new diagnostic indicator of FDG PET scan abnormality, based on age-adjusted t statistics and an automated voxel-based procedure, is presented and validated in a large data set comprising 110 normal controls and 395 patients with probable Alzheimer's disease (AD) that were studied in eight participating centers. The effect of differences in spatial resolution of PET scanners was minimized effectively by filtering and masking. In controls FDG uptake declined significantly with age in anterior cingulate and frontolateral perisylvian cortex. In patients with probable AD decline of FDG uptake in posterior cingulate, temporoparietal, and prefrontal association cortex was related to dementia severity. These effects were clearly distinct from age effects in controls, suggesting that the disease process of AD is not related to normal aging. Women with probable AD had significantly more frontal metabolic impairment than men. The new indicator of metabolic abnormality in AD-related regions provided 93% sensitivity and specificity for distinction of mild to moderate probable AD from normals, and 84% sensitivity at 93% specificity for detection of very mild probable AD (defined by Mini Mental Score 24 or better). All regions related to AD severity were already affected in very mild AD, suggesting that all vulnerable areas are affected to a similar degree already at disease onset. Ventromedial frontal cortex was also abnormal. In conclusion, automated analysis of multicenter FDG PET is feasible, provides insights into AD pathophysiology, and can be used potentially as a sensitive biomarker for early AD diagnosis. (C) 2002 Elsevier Science (USA). [less ▲]

Detailed reference viewed: 29 (2 ULg)
Full Text
Peer Reviewed
See detailSystematic regularization of linear inverse solutions of the EEG source localization problem
Phillips, Christophe ULg; Rugg, Michael D; Friston, Karl J

in NeuroImage (2002), 17(1), 287-301

Distributed linear solutions of the EEG source localization problem are used routinely. Here we describe an approach based on the weighted minimum norm method that imposes constraints using anatomical and ... [more ▼]

Distributed linear solutions of the EEG source localization problem are used routinely. Here we describe an approach based on the weighted minimum norm method that imposes constraints using anatomical and physiological information derived from other imaging modalities to regularize the solution. In this approach the hyperparameters controlling the degree of regularization are estimated using restricted maximum likelihood (ReML). EEG data are always contaminated by noise, e.g., exogenous noise and background brain activity. The conditional expectation of the source distribution, given the data, is attained by carefully balancing the minimization of the residuals induced by noise and the improbability of the estimates as determined by their priors. This balance is specified by hyperparameters that control the relative importance of fitting and conforming to prior constraints. Here we introduce a systematic approach to this regularization problem, in the context of a linear observation model we have described previously. In this model, basis functions are extracted to reduce the solution space a priori in the spatial and temporal domains. The basis sets are motivated by knowledge of the evoked EEG response and information theory. In this paper we focus on an iterative "expectation-maximization" procedure to jointly estimate the conditional expectation of the source distribution and the ReML hyperparameters on which this solution rests. We used simulated data mixed with real EEG noise to explore the behavior of the approach with various source locations, priors, and noise levels. The results enabled us to conclude: M Solutions in the space of informed basis functions have a high face and construct validity, in relation to conventional analyses. (ii) The hyperparameters controlling the degree of regularization vary largely with source geometry and noise. The second conclusion speaks to the usefulness of using adaptative ReML hyperparameter estimates. (C) 2002 Elsevier Science (USA). [less ▲]

Detailed reference viewed: 19 (5 ULg)
Full Text
Peer Reviewed
See detailAnatomically informed basis functions for EEG source localization : Combining functional and anatomical constraints
Phillips, Christophe ULg; Rugg, Michael D; Friston, Karl J

in Neuroimage (2002), 16(3), 678-6951

Distributed linear solutions have frequently been used to solve the source localization problem in EEG. Here we introduce an approach based on the weighted minimum norm (WMN) method that imposes ... [more ▼]

Distributed linear solutions have frequently been used to solve the source localization problem in EEG. Here we introduce an approach based on the weighted minimum norm (WMN) method that imposes constraints using anatomical and physiological information derived from other imaging modalities. The anatomical constraints are used to reduce the solution space a priori by modeling the spatial source distribution with a set of basis functions. These spatial basis functions are chosen in a principled way using information theory. The reduced problem is then solved with a classical WMN method. Further (functional) constraints can be introduced in the weighting of the solution using fMRI brain responses to augment spatial priors. We used simulated data to explore the behavior of the approach over a range of the model's hyperparameters. To assess the construct validity of our method we compared it with two established approaches to the source localization problem, a simple weighted minimum norm and a maximum smoothness (Loreta-like) solution. This involved simulations, using single and multiple sources that were analyzed under different levels of confidence in the priors. (C) 2002 Elsevier Science (USA). [less ▲]

Detailed reference viewed: 19 (0 ULg)
Full Text
Peer Reviewed
See detailClassical and Bayesian inference in neuroimaging: Applications
Friston, Karl J; Glaser, Daniel E; Henson, Richard NA et al

in Neuroimage (2002), 16(2), 484-512

In Friston et al. ((2002) Neuroimage 16: 465-483) we introduced empirical Bayes as a potentially useful way to estimate and make inferences about effects in hierarchical models. In this paper we present a ... [more ▼]

In Friston et al. ((2002) Neuroimage 16: 465-483) we introduced empirical Bayes as a potentially useful way to estimate and make inferences about effects in hierarchical models. In this paper we present a series of models that exemplify the diversity of problems that can be addressed within this framework. In hierarchical linear observation models, both classical and empirical Bayesian approaches can be framed in terms of covariance component estimation (e.g., variance partitioning). To illustrate the use of the expectation-maximization (EM) algorithm in covariance component estimation we focus first on two important problems in fMRI: nonsphericity induced by (i) serial or temporal correlations among errors and (ii) variance components caused by the hierarchical nature of multisubject studies. In hierarchical observation models, variance components at higher levels can be used as constraints on the parameter estimates of lower levels. This enables the use of parametric empirical Bayesian (PEB) estimators, as distinct from classical maximum likelihood (ML) estimates. We develop this distinction to address: (i) The difference between response estimates based on ML and the conditional means from a Bayesian approach and the implications for estimates of intersubject variability. (ii) The relationship between fixed- and random-effect analyses. (iii) The specificity and sensitivity of Bayesian inference and, finally, (iv) the relative importance of the number of scans and subjects. The forgoing is concerned with within- and between-subject variability in multisubject hierarchical fMRI studies. In the second half of this paper we turn to Bayesian inference at the first (within-voxel) level, using PET data to show how priors can be derived from the (between-voxel) distribution of activations over the brain. This application uses exactly the same ideas and formalism but, in this instance, the second level is provided by observations over voxels as opposed to subjects. The ensuing posterior probability maps (PPMs) have enhanced anatomical precision and greater face validity, in relation to underlying anatomy. Furthermore, in comparison to conventional SPMs they are not confounded by the multiple comparison problem that, in a classical context, dictates high thresholds and low sensitivity. We conclude with some general comments on Bayesian approaches to image analysis and on some unresolved issues. [less ▲]

Detailed reference viewed: 7 (1 ULg)
Full Text
Peer Reviewed
See detailClassical and Bayesian inference in neuroimaging : Theory
Friston, Karl J; Penny, W.; Phillips, Christophe ULg et al

in Neuroimage (2002), 16(2), 465-483

This paper reviews hierarchical observation models, used in functional neuroimaging, in a Bayesian light. It emphasizes the common ground shared by classical and Bayesian methods to show that conventional ... [more ▼]

This paper reviews hierarchical observation models, used in functional neuroimaging, in a Bayesian light. It emphasizes the common ground shared by classical and Bayesian methods to show that conventional analyses of neuroimaging data can be usefully extended within an empirical Bayesian framework. In particular we formulate the procedures used in conventional data analysis in terms of hierarchical linear models and establish a connection between classical inference and parametric empirical Bayes (PEB) through covariance component estimation. This estimation is based on an expectation maximization or EM algorithm. The key point is that hierarchical models not only provide for appropriate inference at the highest level but that one can revisit lower levels suitably equipped to make Bayesian inferences. Bayesian inferences eschew many of the difficulties encountered with classical inference and characterize brain responses in a way that is more directly predicated on what one is interested in. The motivation for Bayesian approaches is reviewed and the theoretical background is presented in a way that relates to conventional methods, in particular restricted maximum likelihood (ReML). This paper is a technical and theoretical prelude to subsequent papers that deal with applications of the theory to a range of important issues in neuroimaging. These issues include; (i) Estimating nonsphericity or variance components in fMRI time-series that can arise from serial correlations within subject, or are induced by multisubject (i.e., hierarchical) studies. (ii) Spatiotemporal Bayesian models for imaging data, in which voxels-specific effects are constrained by responses in other voxels. (iii) Bayesian estimation of nonlinear models of hemodynamic responses and (iv) principled ways of mixing structural and functional priors in EEG source reconstruction. Although diverse, all these estimation problems are accommodated by the PEB framework described in this paper. [less ▲]

Detailed reference viewed: 10 (0 ULg)
Full Text
Peer Reviewed
See detailGeneration of rapid eye movements during paradoxical sleep in humans
Peigneux, Philippe ULg; Laureys, Steven ULg; Fuchs, Sonia et al

in NeuroImage (2001), 14(3), 701-708

Although rapid eye movements (REMs) are a prominent feature of paradoxical sleep (PS), their origin and functional significance remain poorly understood in humans. In animals, including nonhuman primates ... [more ▼]

Although rapid eye movements (REMs) are a prominent feature of paradoxical sleep (PS), their origin and functional significance remain poorly understood in humans. In animals, including nonhuman primates, REMs during PS are closely related to the occurrence of the so-called PGO waves, i.e., prominent phasic activities recorded throughout the brain but predominantly and most easily in the pons (P), the lateral geniculate bodies (G), and the occipital cortex (O). Therefore, and because the evolution of species is parsimonious, a plausible hypothesis would be that during PS in humans, REMs are generated by mechanisms similar to PGO waves. Using positron emission tomography and iterative cerebral blood flow measurements by H(2)(15)O infusions, we predicted that the brain regions where the PGO waves are the most easily recorded in animals would be differentially more active in PS than in wakefulness, in relation with the density of the REM production [i.e., we looked for the condition (PS versus wakefulness) by performance (REM density) interaction]. Accordingly, we found a significant interaction effect in the right geniculate body and in the primary occipital cortex. The result supports the hypothesis of the existence of processes similar to PGO waves in humans, responsible for REM generation. The interest in the presence of PGO waves in humans is outstanding because the cellular processes involved in, or triggered by, PGO waves might favor brain plasticity during PS. [less ▲]

Detailed reference viewed: 9 (1 ULg)