Population-level inferences for distributed MEG source localization under multiple constraints: Application to face-evoked fields

Ant Colony System Optimization for Spatiotemporal Modelling of Combined EEG and MEG Data

Electroencephalography/Magnetoencephalography (EEG/MEG) source localization involves the estimation of neural activity inside the brain volume that underlies the EEG/MEG measures observed at the sensor array. In this paper, we consider a Bayesian finite spatial mixture model for source reconstruction and implement Ant Colony System (ACS) optimization coupled with Iterated Conditional Modes (ICM) for computing estimates of the neural source activity. Our approach is evaluated using simulation studies and a real data application in which we implement a nonparametric bootstrap for interval estimation. We demonstrate improved performance of the ACS-ICM algorithm as compared to existing methodology for the same spatiotemporal model.

Low-Density EEG for Neural Activity Reconstruction Using Multivariate Empirical Mode Decomposition

Several approaches can be used to estimate neural activity. The main differences between them concern the a priori information used and its sensitivity to high noise levels. Empirical mode decomposition (EMD) has been recently applied to electroencephalography EEG-based neural activity reconstruction to provide a priori time-frequency information to improve the estimation of neural activity. EMD has the specific ability to identify independent oscillatory modes in non-stationary signals with multiple oscillatory components. However, attempts to use EMD in EEG analysis have not yet provided optimal reconstructions, due to the intrinsic mode-mixing problem of EMD. Several studies have used single-channel analysis, whereas others have used multiple-channel analysis for other applications. Here, we present the results of multiple-channel analysis using multivariate empirical mode decomposition (MEMD) to reduce the mode-mixing problem and provide useful a priori time-frequency information for the reconstruction of neuronal activity using several low-density EEG electrode montages. The methods were evaluated using real and synthetic EEG data, in which the reconstructions were performed using the multiple sparse priors (MSP) algorithm with EEG electrode montages of 32, 16, and 8 electrodes. The quality of the source reconstruction was assessed using the Wasserstein metric. A comparison of the solutions without pre-processing and those after applying MEMD showed the source reconstructions to be improved using MEMD as a priori information for the low-density montages of 8 and 16 electrodes. The mean source reconstruction error on a real EEG dataset was reduced by 59.42 and 66.04% for the 8 and 16 electrode montages, respectively, and that on a simulated EEG with three active sources, by 87.31 and 31.45% for the same electrode montages.

Accounting for Taste: A Multi-Attribute Neurocomputational Model Explains the Neural Dynamics of Choices for Self and Others

How do we make choices for others with different preferences from our own? Although neuroimaging studies implicate similar circuits in representing preferences for oneself and others, some models propose that additional corrective mechanisms come online when choices for others diverge from one's own preferences. Here we used event-related potentials (ERPs) in humans, in combination with computational modeling, to examine how social information is integrated in the time leading up to choices for oneself and others. Hungry male and female participants with unrestricted diets selected foods for themselves, a similar unrestricted eater, and a dissimilar, self-identified healthy eater. Across choices for both oneself and others, ERP value signals emerged within the same time window but differentially reflected taste and health attributes based on the recipient's preferences. Choices for the dissimilar recipient were associated with earlier activity localized to brain regions implicated in social cognition, including temporoparietal junction. Finally, response-locked analysis revealed a late ERP component specific to choices for the similar recipient, localized to the parietal lobe, that appeared to reflect differences in the response threshold based on uncertainty. A multi-attribute computational model supported the link between specific ERP components and distinct model parameters, and was not significantly improved by adding time-dependent dual processes. Model simulations suggested that longer response times previously associated with effortful correction may alternatively arise from higher choice uncertainty. Together, these results provide a parsimonious neurocomputational mechanism for social decision-making, additionally explaining divergent patterns of choice and response time data in decisions for oneself and others.SIGNIFICANCE STATEMENT How do we choose for others, particularly when they have different preferences? Whereas some studies suggest that similar neural circuits underlie decision-making for oneself and others, others argue for additional, slower perspective-taking mechanisms. Combining event-related potentials with computational modeling, we found that integration of others' preferences occurs over the same timescale as for oneself while differentially tracking recipient-relevant attributes. Although choosing for others took longer and produced differences in late-emerging neural responses, computational modeling attributed these patterns to greater response caution rather than egocentric bias correction. Computational simulations also correctly predicted when and why choosing differently for others takes longer, suggesting that a model incorporating value integration and evidence accumulation can parsimoniously account for complex patterns in social decision-making.

Working Memory Replay Prioritizes Weakly Attended Events

One view of working memory posits that maintaining a series of events requires their sequential and equal mnemonic replay. Another view is that the content of working memory maintenance is prioritized by attention. We decoded the dynamics for retaining a sequence of items using magnetoencephalography, wherein participants encoded sequences of three stimuli depicting a face, a manufactured object, or a natural item and maintained them in working memory for 5000 ms. Memory for sequence position and stimulus details were probed at the end of the maintenance period. Decoding of brain activity revealed that one of the three stimuli dominated maintenance independent of its sequence position or category; and memory was enhanced for the selectively replayed stimulus. Analysis of event-related responses during the encoding of the sequence showed that the selectively replayed stimuli were determined by the degree of attention at encoding. The selectively replayed stimuli had the weakest initial encoding indexed by weaker visual attention signals at encoding. These findings do not rule out sequential mnemonic replay but reveal that attention influences the content of working memory maintenance by prioritizing replay of weakly encoded events. We propose that the prioritization of weakly encoded stimuli protects them from interference during the maintenance period, whereas the more strongly encoded stimuli can be retrieved from long-term memory at the end of the delay period.

Temporal Dynamics of Sensorimotor Networks in Effort-Based Cost-Benefit Valuation: Early Emergence and Late Net Value Integration

Although physical effort can impose significant costs on decision-making, when and how effort cost information is incorporated into choice remains contested, reflecting a larger debate over the role of sensorimotor networks in specifying behavior. Serial information processing models, in which motor circuits simply implement the output of cognitive systems, hypothesize that effort cost factors into decisions relatively late, via integration with stimulus values into net (combined) value signals in dorsomedial frontal cortex (dmFC). In contrast, ethology-inspired approaches suggest a more active role for the dorsal sensorimotor stream, with effort cost signals emerging rapidly after stimulus onset. Here we investigated the time course of effort cost integration using event-related potentials in hungry human subjects while they made decisions about expending physical effort for appetitive foods. Consistent with the ethological perspective, we found that effort cost was represented from as early as 100-250 ms after stimulus onset, localized to dorsal sensorimotor regions including middle cingulate, somatosensory, and motor/premotor cortices. However, examining the same data time-locked to motor output revealed net value signals combining stimulus value and effort cost approximately -400 ms before response, originating from sensorimotor areas including dmFC, precuneus, and posterior parietal cortex. Granger causal connectivity analysis of the motor effector signal in the time leading to response showed interactions between these sensorimotor regions and ventrolateral prefrontal cortex, a structure associated with adjusting behavior-response mappings. These results suggest that rapid activation of sensorimotor regions interacts with cognitive valuation systems, producing a net value signal reflecting both physical effort and reward contingencies.

SIGNIFICANCE STATEMENT

Although physical effort imposes a cost on choice, when and how effort cost influences neural correlates of decision-making remains contested. This dispute reflects a larger disagreement between cognitive neuroscience and ethology over the role of sensorimotor systems in behavior: are sensorimotor circuits merely implementing the late-stage output of cognitive systems, or engaged rapidly and interactively from early in decision-making? We find that, although early representation of effort cost is associated with sensorimotor regions, these signals are also integrated with cognitive stimulus value representations in the time leading up to motor response. These data suggest that sensorimotor networks interact dynamically with cognitive systems to guide decision-making, providing a first step toward reconciling differing perspectives on sensorimotor roles in valuation and choice.

The impact of MEG source reconstruction method on source-space connectivity estimation: A comparison between minimum-norm solution and beamforming

Despite numerous important contributions, the investigation of brain connectivity with magnetoencephalography (MEG) still faces multiple challenges. One critical aspect of source-level connectivity, largely overlooked in the literature, is the putative effect of the choice of the inverse method on the subsequent cortico-cortical coupling analysis. We set out to investigate the impact of three inverse methods on source coherence detection using simulated MEG data. To this end, thousands of randomly located pairs of sources were created. Several parameters were manipulated, including inter- and intra-source correlation strength, source size and spatial configuration. The simulated pairs of sources were then used to generate sensor-level MEG measurements at varying signal-to-noise ratios (SNR). Next, the source level power and coherence maps were calculated using three methods (a) L2-Minimum-Norm Estimate (MNE), (b) Linearly Constrained Minimum Variance (LCMV) beamforming, and (c) Dynamic Imaging of Coherent Sources (DICS) beamforming. The performances of the methods were evaluated using Receiver Operating Characteristic (ROC) curves. The results indicate that beamformers perform better than MNE for coherence reconstructions if the interacting cortical sources consist of point-like sources. On the other hand, MNE provides better connectivity estimation than beamformers, if the interacting sources are simulated as extended cortical patches, where each patch consists of dipoles with identical time series (high intra-patch coherence). However, the performance of the beamformers for interacting patches improves substantially if each patch of active cortex is simulated with only partly coherent time series (partial intra-patch coherence). These results demonstrate that the choice of the inverse method impacts the results of MEG source-space coherence analysis, and that the optimal choice of the inverse solution depends on the spatial and synchronization profile of the interacting cortical sources. The insights revealed here can guide method selection and help improve data interpretation regarding MEG connectivity estimation.

Improved Localization of Seizure Onset Zones Using Spatiotemporal Constraints and Time-Varying Source Connectivity

Presurgical evaluation of brain neural activity is commonly carried out in refractory epilepsy patients to delineate as accurately as possible the seizure onset zone (SOZ) before epilepsy surgery. In practice, any subjective interpretation of electroencephalographic (EEG) recordings is hindered mainly because of the highly stochastic behavior of the epileptic activity. We propose a new method for dynamic source connectivity analysis that aims to accurately localize the seizure onset zones by explicitly including temporal, spectral, and spatial information of the brain neural activity extracted from EEG recordings. In particular, we encode the source nonstationarities in three critical stages of processing: Inverse problem solution, estimation of the time courses extracted from the regions of interest, and connectivity assessment. With the aim to correctly encode all temporal dynamics of the seizure-related neural network, a directed functional connectivity measure is employed to quantify the information flow variations over the time window of interest. Obtained results on simulated and real EEG data confirm that the proposed approach improves the accuracy of SOZ localization.

Reconstruction of Neural Activity from EEG Data Using Dynamic Spatiotemporal Constraints

We present a novel iterative regularized algorithm (IRA) for neural activity reconstruction that explicitly includes spatiotemporal constraints, performing a trade-off between space and time resolutions. For improving the spatial accuracy provided by electroencephalography (EEG) signals, we explore a basis set that describes the smooth, localized areas of potentially active brain regions. In turn, we enhance the time resolution by adding the Markovian assumption for brain activity estimation at each time period. Moreover, to deal with applications that have either distributed or localized neural activity, the spatiotemporal constraints are expressed through [Formula: see text] and [Formula: see text] norms, respectively. For the purpose of validation, we estimate the neural reconstruction performance in time and space separately. Experimental testing is carried out on artificial data, simulating stationary and non-stationary EEG signals. Also, validation is accomplished on two real-world databases, one holding Evoked Potentials and another with EEG data of focal epilepsy. Moreover, responses of functional magnetic resonance imaging for the former EEG data have been measured in advance, allowing to contrast our findings. Obtained results show that the [Formula: see text]-based IRA produces a spatial resolution that is comparable to the one achieved by some widely used sparse-based estimators of brain activity. At the same time, the [Formula: see text]-based IRA outperforms other similar smooth solutions, providing a spatial resolution that is lower than the sparse [Formula: see text]-based solution. As a result, the proposed IRA is a promising method for improving the accuracy of brain activity reconstruction.

MEG Connectivity and Power Detections with Minimum Norm Estimates Require Different Regularization Parameters

Minimum Norm Estimation (MNE) is an inverse solution method widely used to reconstruct the source time series that underlie magnetoencephalography (MEG) data. MNE addresses the ill-posed nature of MEG source estimation through regularization (e.g., Tikhonov regularization). Selecting the best regularization parameter is a critical step. Generally, once set, it is common practice to keep the same coefficient throughout a study. However, it is yet to be known whether the optimal lambda for spectral power analysis of MEG source data coincides with the optimal regularization for source-level oscillatory coupling analysis. We addressed this question via extensive Monte-Carlo simulations of MEG data, where we generated 21,600 configurations of pairs of coupled sources with varying sizes, signal-to-noise ratio (SNR), and coupling strengths. Then, we searched for the Tikhonov regularization coefficients (lambda) that maximize detection performance for (a) power and (b) coherence. For coherence, the optimal lambda was two orders of magnitude smaller than the best lambda for power. Moreover, we found that the spatial extent of the interacting sources and SNR, but not the extent of coupling, were the main parameters affecting the best choice for lambda. Our findings suggest using less regularization when measuring oscillatory coupling compared to power estimation.

Sparse EEG Source Localization Using Bernoulli Laplacian Priors

Source localization in electroencephalography has received an increasing amount of interest in the last decade. Solving the underlying ill-posed inverse problem usually requires choosing an appropriate regularization. The usual l2 norm has been considered and provides solutions with low computational complexity. However, in several situations, realistic brain activity is believed to be focused in a few focal areas. In these cases, the l2 norm is known to overestimate the activated spatial areas. One solution to this problem is to promote sparse solutions for instance based on the l1 norm that are easy to handle with optimization techniques. In this paper, we consider the use of an l0 + l1 norm to enforce sparse source activity (by ensuring the solution has few nonzero elements) while regularizing the nonzero amplitudes of the solution. More precisely, the l0 pseudonorm handles the position of the nonzero elements while the l1 norm constrains the values of their amplitudes. We use a Bernoulli-Laplace prior to introduce this combined l0 + l1 norm in a Bayesian framework. The proposed Bayesian model is shown to favor sparsity while jointly estimating the model hyperparameters using a Markov chain Monte Carlo sampling technique. We apply the model to both simulated and real EEG data, showing that the proposed method provides better results than the l2 and l1  norms regularizations in the presence of pointwise sources. A comparison with a recent method based on multiple sparse priors is also conducted.

Face processing in the brains of pre-school aged children measured with MEG

There are two competing theories concerning the development of face perception: a late maturation account and an early maturation account. Magnetoencephalography (MEG) neuroimaging holds promise for adjudicating between the two opposing accounts by providing objective neurophysiological measures of face processing, with sufficient temporal resolution to isolate face-specific brain responses from those associated with other sensory, cognitive and motor processes. The current study used a customized child MEG system to measure M100 and M170 brain responses in 15 children aged three to six years while they viewed faces, cars and their phase-scrambled counterparts. Compared to adults tested using the same stimuli in a conventional MEG system, children showed significantly larger and later M100 responses. Children's M170 responses, derived by subtracting the responses to phase-scrambled images from the corresponding images (faces or cars) were delayed in latency but otherwise resembled the adult M170. This component has not been obtained in previous studies of young children tested using conventional adult MEG systems. However children did show a markedly reduced M170 response to cars in comparison to adults. This may reflect children's lack of expertise with cars relative to faces. Taken together, these data are in accord with recent behavioural and neuroimaging data that support early maturation of the basic face processing functions.

Multiple neural mechanisms for coloring words in synesthesia

Grapheme-color synesthesia is a phenomenon in which achromatic letters/digits automatically induce particular colors. When multiple letters are integrated into a word, some synesthetes perceive that all those letters are changed into the same color, reporting lexical color to that word. Previous psychological studies found several "rules" that determine those lexical colors. The colors to most words are determined by the first letters of the words, while some words in ordinal sequences have their specific colors. Recent studies further reported the third case where lexical colors might be influenced by semantic information of words. Although neural mechanisms determining those lexical colors remained unknown, here we identified three separate neural systems in the synesthete's brain underlying three rules for illusory coloring of words. In addition to the occipito-temporal and parietal regions previously found to be associated with the grapheme-color synesthesia, neural systems for lexical coloring extended to linguistic areas in the left inferior frontal and anterior temporal regions that were engaged in semantic analyses of words. Those results indicate an involvement of wider and higher neural networks than previously assumed in a production of synesthetic colors to visual stimuli and further showed a multiplicity of synesthetic mechanisms represented in the single brain.

Localizing evoked and induced responses to faces using magnetoencephalography

A rich pattern of responses in frequency, time and space are known to be generated in the visual cortex in response to faces. Recently, a number of studies have used magnetoencephalography (MEG) to try to record these responses non-invasively – in many cases using source analysis techniques based on the beamforming method. Here we sought both to characterize best practice for measuring face-specific responses using MEG beamforming, and to determine whether the results produced by the beamformer match evidence from other modalities. We measured activity to visual presentation of face stimuli and phase-scrambled control stimuli, and performed source analyses of both induced and evoked responses using Synthetic Aperture Magnetometry. We localized the gamma-band response to bilateral lateral occipital cortex, and both the gamma-band response and the M170-evoked response to the right fusiform gyrus. Differences in the gamma-band response between faces and scrambled stimuli were confined to the frequency range 50–90 Hz; gamma-band activity at higher frequencies did not differ between the two stimulus categories. We additionally identified a component of the M220-evoked response – localized to the parieto-occipital sulcus – which was enhanced for scrambled vs. unscrambled faces. These findings help to establish that MEG beamforming can localize face-specific responses in time, frequency and space with good accuracy (when validated against established findings from functional magnetic resonance imaging and intracranial recordings), as well as contributing to the establishment of best methodological practice for the use of the beamformer method to measure face-specific responses.

Temporally dissociable mechanisms of self-control: early attentional filtering versus late value modulation

Optimal decision-making often requires exercising self-control. A growing fMRI literature has implicated the dorsolateral prefrontal cortex (dlPFC) in successful self-control, but due to the limitations inherent in BOLD measures of brain activity, the neurocomputational role of this region has not been resolved. Here we exploit the high temporal resolution and whole-brain coverage of event-related potentials (ERPs) to test the hypothesis that dlPFC affects dietary self-control through two different mechanisms: attentional filtering and value modulation. Whereas attentional filtering of sensory input should occur early in the decision process, value modulation should occur later on, after the computation of stimulus values begins. Hungry human subjects were asked to make food choices while we measured neural activity using ERP in a natural condition, in which they responded freely and did not exhibit a tendency to regulate their diet, and in a self-control condition, in which they were given a financial incentive to lose weight. We then measured various neural markers associated with the attentional filtering and value modulation mechanisms across the decision period to test for changes in neural activity during the exercise of self-control. Consistent with the hypothesis, we found evidence for top-down attentional filtering early on in the decision period (150-200 ms poststimulus onset) as well as evidence for value modulation later in the process (450-650 ms poststimulus onset). We also found evidence that dlPFC plays a role in the deployment of both mechanisms.

Temporal dynamics of neural activity underlying unconscious processing of manipulable objects

The primate visual system is assumed to comprise two main pathways: a ventral pathway for shape and color perception and a dorsal pathway for spatial processing and visuomotor control. Previous studies consistently reported strong activation in the dorsal pathway(especially in the inferior parietal region) induced by manipulable object images such as tools. However, it is controversial whether the dorsal pathway retains this preferential activity to tool images under unconscious perception. In the present study, we used magnetoencephalography (MEG) and investigated spatio-temporal dynamics of neural responses to visible and invisible tool images. A presentation of visible tool images elicited a strong neural response over the parietal regions in the left hemisphere peaking at 400 msec. This response unique to the processing of tool information in the left parietal regions was still observed when conscious perception of tool images was inhibited by interocular suppression. Furthermore, analyses of neural oscillation signals revealed a suppression of m rhythm (8-13 Hz), a neural index of movement execution or imagery,induced by both visible and invisible tools. Those results indicated that the neural circuit to process the tool information was preserved under unconscious perception, highlighting an implicit aspect of the dorsal pathway.

Different cortical dynamics in face and body perception: an MEG study

Evidence from functional neuroimaging indicates that visual perception of human faces and bodies is carried out by distributed networks of face and body-sensitive areas in the occipito-temporal cortex. However, the dynamics of activity in these areas, needed to understand their respective functional roles, are still largely unknown. We monitored brain activity with millisecond time resolution by recording magnetoencephalographic (MEG) responses while participants viewed photographs of faces, bodies, and control stimuli. The cortical activity underlying the evoked responses was estimated with anatomically-constrained noise-normalised minimum-norm estimate and statistically analysed with spatiotemporal cluster analysis.

Our findings point to distinct spatiotemporal organization of the neural systems for face and body perception. Face-selective cortical currents were found at early latencies (120–200 ms) in a widespread occipito-temporal network including the ventral temporal cortex (VTC). In contrast, early body-related responses were confined to the lateral occipito-temporal cortex (LOTC). These were followed by strong sustained body-selective responses in the orbitofrontal cortex from 200–700 ms, and in the lateral temporal cortex and VTC after 500 ms latency. Our data suggest that the VTC region has a key role in the early processing of faces, but not of bodies. Instead, the LOTC, which includes the extra-striate body area (EBA), appears the dominant area for early body perception, whereas the VTC contributes to late and post-perceptual processing.

Environmental reverberation affects processing of sound intensity in right temporal cortex

Although sound reverberation is considered a nuisance variable in most studies investigating auditory processing, it can serve as a cue for loudness constancy, a phenomenon describing constant loudness perception in spite of changing sound source distance. In this study, we manipulated room reverberation characteristics to test their effect on psychophysical loudness constancy and we tested with magnetoencephalography on human subjects for neural responses reflecting loudness constancy. Psychophysically, we found that loudness constancy was present in strong, but not weak, reverberation conditions. In contrast, the dependence of sound distance judgment on actual distance was similar across conditions. We observed brain activity reflecting behavioral loudness constancy, i.e. inverse scaling of the evoked magnetic fields with distance for weak reverberation but constant responses across distance for strong reverberation from ~210 to 270 ms after stimulus onset. Distributed magnetoencephalography source reconstruction revealed underlying neural generators within the right middle temporal and right inferior anterior temporal lobe. Our data suggest a dissociation of loudness constancy and distance perception, implying a direct usage of reverberation cues for constructing constant loudness across distance. Furthermore, our magnetoencephalography data suggest involvement of auditory association areas in the right middle and right inferior anterior temporal cortex in this process.

Temporal dynamics of neural activity in an integration of visual and contextual information in an esthetic preference task

While viewing works of art in galleries, we evaluate them by integrating at least two types of information: their visual properties (e.g., colors, symmetry, and proportion) and contextual information accompanying them (e.g., titles and names of artists). How rapidly the brain integrates visual and contextual information of artworks remains to be investigated. Using electroencephalography (EEG), we investigated neural activity when subjects with no professional experience in art viewed images of sculptures (masterpieces from the Classical and Renaissance periods, characterized by a canonical proportion of the golden ratio) and performed a five-scale rating of how appealing they were. At the beginning of each trial, we manipulated the expectations of the subjects for an upcoming sculpture by presenting information about its authenticity (either "genuine" or "fake"), although all images were actually taken from genuine artworks. The image of the sculpture was then presented, either in its original proportion or after being deformed by a photo-editing software. This 2 × 2 factorial design enabled us to identify whether each component of the EEG response was sensitive to contextual information (genuine or fake), visual information (original or deformed), or both. Results revealed that amplitudes of a positive EEG component emerging at 200-300ms after the presentation of the artworks (mainly distributed over the parietal cortex) were significantly modulated by both visual and contextual factors, indicating a rapid integration of these two types of information in the brain.

MEG source localization of spatially extended generators of epileptic activity: comparing entropic and hierarchical bayesian approaches

Localizing the generators of epileptic activity in the brain using Electro-EncephaloGraphy (EEG) or Magneto-EncephaloGraphy (MEG) signals is of particular interest during the pre-surgical investigation of epilepsy. Epileptic discharges can be detectable from background brain activity, provided they are associated with spatially extended generators. Using realistic simulations of epileptic activity, this study evaluates the ability of distributed source localization methods to accurately estimate the location of the generators and their sensitivity to the spatial extent of such generators when using MEG data. Source localization methods based on two types of realistic models have been investigated: (i) brain activity may be modeled using cortical parcels and (ii) brain activity is assumed to be locally smooth within each parcel. A Data Driven Parcellization (DDP) method was used to segment the cortical surface into non-overlapping parcels and diffusion-based spatial priors were used to model local spatial smoothness within parcels. These models were implemented within the Maximum Entropy on the Mean (MEM) and the Hierarchical Bayesian (HB) source localization frameworks. We proposed new methods in this context and compared them with other standard ones using Monte Carlo simulations of realistic MEG data involving sources of several spatial extents and depths. Detection accuracy of each method was quantified using Receiver Operating Characteristic (ROC) analysis and localization error metrics. Our results showed that methods implemented within the MEM framework were sensitive to all spatial extents of the sources ranging from 3 cm² to 30 cm², whatever were the number and size of the parcels defining the model. To reach a similar level of accuracy within the HB framework, a model using parcels larger than the size of the sources should be considered.

Cortical neural activities associated with binocular rivalry: an EEG-fMRI integrative study

Binocular rivalry occurs over time when each eye is simultaneously presented with different visual stimuli. We have addressed which brain regions are and the mechanisms involved in binocular rivalry using EEG and fMRI measurements of cortical activities during observations of competitive (rivalry) and cooperative (fusion) drifting vertical gratings. By applying an EEG-fMRI integrative method, we analyzed the time courses of multiple cortical sources of event-related potentials obtained under rivalry or fusion conditions. We detected significant cortical activities at bilateral MT+/V5 and the right prefrontal eye field in the rivalry condition; however, this may not reflect intrinsic alternation in binocular rivalry. Our findings suggest that novel integrative methods are necessary to investigate the distributed cortical network associated with binocular rivalry, through analysis of multiple cortical sources of event-related desynchronization and/or synchronization in certain frequency bands.

Effect of local anesthesia on trigeminal somatosensory-evoked magnetic fields

For objective neurophysiological evaluation of the function of the trigeminal system, magnetoencephalography- based TSEF (trigeminal somatosensory-evoked field) assessment would be valuable in providing spatial and temporal profiles of cortical responses. However, this necessitates knowledge of how TSEF varies with trigeminal nerve dysfunctions. We introduced a conduction block of the trigeminal nerve using local anesthesia (lidocaine) to temporally mimic nerve dysfunctions, and monitored TSEF changes. Following an electrical stimulation of the lower lip, a magnetic response with peak latency of approximately 20 ms was identified in all participants. Dipole for the peak was estimated on the post-central gyrus in the participant's own magnetic resonance image. After normalization to Montreal Neurological Institute (MNI) space and inter-participant data integration, the summary equivalent current dipole localization among participants remained in the post-central gyrus, suggesting validity of the use of MNI space. Partial anesthesia of the lower lip led to a loss of the waveform characteristics of TSEF for electrical stimulation to the trigeminal nerve. We verified that the 20-ms latency cortical response of TSEF components localized at the primary sensory cortex can serve as a robust neurofunctional marker of experimental trigeminal nerve dysfunction.

Temporal dynamics of neural activity at the moment of emergence of conscious percept

From which regions of the brain do conscious representations of visual stimuli emerge? This is an important but controversial issue in neuroscience because some studies have reported a major role of the higher visual regions of the ventral pathway in conscious perception, whereas others have found neural correlates of consciousness as early as in the primary visual areas and in the thalamus. One reason for this controversy has been the difficulty in focusing on neural activity at the moment when conscious percepts are generated in the brain, excluding any bottom-up responses (not directly related to consciousness) that are induced by stimuli. In this study, we address this issue with a new approach that can induce a rapid change in conscious perception with little influence from bottom-up responses. Our results reveal that the first consciousness-related activity emerges from the higher visual region of the ventral pathway. However, this activity is rapidly diffused to the entire brain, including the early visual cortex. These results thus integrate previous "higher" and "lower" views on the emergence of neural correlates of consciousness, providing a new perspective for the temporal dynamics of consciousness.

A magnetoencephalographic study of face processing: M170, gamma-band oscillations and source localization

EEG studies suggested that the N170 ERP and Gamma-band responses to faces reflect early and later stages of a multiple-level face-perception mechanism, respectively. However, these conclusions should be considered cautiously because EEG-recorded Gamma may be contaminated by noncephalic activity such as microsaccades. Moreover, EEG studies of Gamma cannot easily reveal its intracranial sources. Here we recorded MEG rather than EEG, assessed the sources of the M170 and Gamma oscillations using beamformer, and explored the sensitivity of these neural manifestations to global, featural and configural information in faces. The M170 was larger in response to faces and face components than in response to watches. Scrambling the configuration of the inner components of the face even if presented without the face contour reduced and delayed the M170. The amplitude of MEG Gamma oscillations (30-70 Hz) was higher than baseline during an epoch between 230-570 ms from stimulus onset and was particularly sensitive to the configuration of the stimuli, regardless of their category. However, in the lower part of this frequency range (30-40 Hz) only physiognomic stimuli elevated the MEG above baseline. Both the M170 and Gamma were generated in a posterior-ventral network including the fusiform, inferior-occipital and lingual gyri, all in the right hemisphere. The generation of Gamma involved additional sources in the visual system, bilaterally. We suggest that the evoked M170 manifests a face-perception mechanism based on the global characteristics of face, whereas the induced Gamma oscillations are associated with the integration of visual input into a pre-existent coherent perceptual representation.

The problem of low variance voxels in statistical parametric mapping; a new hat avoids a 'haircut'

Statistical parametric mapping (SPM) locates significant clusters based on a ratio of signal to noise (a ‘contrast’ of the parameters divided by its standard error) meaning that very low noise regions, for example outside the brain, can attain artefactually high statistical values. Similarly, the commonly applied preprocessing step of Gaussian spatial smoothing can shift the peak statistical significance away from the peak of the contrast and towards regions of lower variance. These problems have previously been identified in positron emission tomography (PET) (Reimold et al., 2006) and voxel-based morphometry (VBM) (Acosta-Cabronero et al., 2008), but can also appear in functional magnetic resonance imaging (fMRI) studies. Additionally, for source-reconstructed magneto- and electro-encephalography (M/EEG), the problems are particularly severe because sparsity-favouring priors constrain meaningfully large signal and variance to a small set of compactly supported regions within the brain. (Acosta-Cabronero et al., 2008) suggested adding noise to background voxels (the ‘haircut’), effectively increasing their noise variance, but at the cost of contaminating neighbouring regions with the added noise once smoothed. Following theory and simulations, we propose to modify – directly and solely – the noise variance estimate, and investigate this solution on real imaging data from a range of modalities.

Different neural pathways to negative affect in youth with pediatric bipolar disorder and severe mood dysregulation

Questions persist regarding the presentation of bipolar disorder (BD) in youth and the nosological significance of irritability. Of particular interest is whether severe mood dysregulation (SMD), characterized by severe non-episodic irritability, hyper-arousal, and hyper-reactivity to negative emotional stimuli, is a developmental presentation of pediatric BD and, therefore, whether the two conditions are pathophysiologically similar. We administered the affective Posner paradigm, an attentional task with a condition involving blocked goal attainment via rigged feedback. The sample included 60 youth (20 BD, 20 SMD, and 20 controls) ages 8-17. Magnetoencephalography (MEG) examined neuronal activity (4-50 Hz) following negative versus positive feedback. We also examined reaction time (RT), response accuracy, and self-reported affect. Both BD and SMD youth reported being less happy than controls during the rigged condition. Also, SMD youth reported greater arousal following negative feedback than both BD and controls, and they responded to negative feedback with significantly greater activation of the anterior cingulate cortex (ACC) and medial frontal gyrus (MFG) than controls. Compared to SMD and controls, BD youth displayed greater superior frontal gyrus (SFG) activation and decreased insula activation following negative feedback. Data suggest a greater negative affective response to blocked goal attainment in SMD versus BD and control youth. This occurs in tandem with hyperactivation of medial frontal regions in SMD youth, while BD youth show dysfunction in the SFG and insula. Data add to a growing empirical base that differentiates pediatric BD and SMD and begin to elucidate potential neural mechanisms of irritability.

A Parametric Empirical Bayesian Framework for the EEG/MEG Inverse Problem: Generative Models for Multi-Subject and Multi-Modal Integration

We review recent methodological developments within a parametric empirical Bayesian (PEB) framework for reconstructing intracranial sources of extracranial electroencephalographic (EEG) and magnetoencephalographic (MEG) data under linear Gaussian assumptions. The PEB framework offers a natural way to integrate multiple constraints (spatial priors) on this inverse problem, such as those derived from different modalities (e.g., from functional magnetic resonance imaging, fMRI) or from multiple replications (e.g., subjects). Using variations of the same basic generative model, we illustrate the application of PEB to three cases: (1) symmetric integration (fusion) of MEG and EEG; (2) asymmetric integration of MEG or EEG with fMRI, and (3) group-optimization of spatial priors across subjects. We evaluate these applications on multi-modal data acquired from 18 subjects, focusing on energy induced by face perception within a time–frequency window of 100–220 ms, 8–18 Hz. We show the benefits of multi-modal, multi-subject integration in terms of the model evidence and the reproducibility (over subjects) of cortical responses to faces.

A parametric empirical Bayesian framework for fMRI-constrained MEG/EEG source reconstruction

We describe an asymmetric approach to fMRI and MEG/EEG fusion in which fMRI data are treated as empirical priors on electromagnetic sources, such that their influence depends on the MEG/EEG data, by virtue of maximizing the model evidence. This is important if the causes of the MEG/EEG signals differ from those of the fMRI signal. Furthermore, each suprathreshold fMRI cluster is treated as a separate prior, which is important if fMRI data reflect neural activity arising at different times within the EEG/MEG data. We present methodological considerations when mapping from a 3D fMRI Statistical Parametric Map to a 2D cortical surface and thence to the covariance components used within our Parametric Empirical Bayesian framework. Our previous introduction of a canonical (inverse‐normalized) cortical mesh also allows deployment of fMRI priors that live in a template space; for example, from a group analysis of different individuals. We evaluate the ensuing scheme with MEG and EEG data recorded simultaneously from 12 participants, using the same face‐processing paradigm under which independent fMRI data were obtained. Because the fMRI priors become part of the generative model, we use the model evidence to compare (i) multiple versus single, (ii) valid versus invalid, (iii) binary versus continuous, and (iv) variance versus covariance fMRI priors. For these data, multiple, valid, binary, and variance fMRI priors proved best for a standard Minimum Norm inversion. Interestingly, however, inversion using Multiple Sparse Priors benefited little from additional fMRI priors, suggesting that they already provide a sufficiently flexible generative model. Hum Brain Mapp, 2010. © 2010 Wiley‐Liss, Inc.

Bayesian symmetrical EEG/fMRI fusion with spatially adaptive priors

In this paper, we propose a novel symmetrical EEG/fMRI fusion method which combines EEG and fMRI by means of a common generative model. We use a total variation (TV) prior to model the spatial distribution of the cortical current responses and hemodynamic response functions, and utilize spatially adaptive temporal priors to model their temporal shapes. The spatial adaptivity of the prior model allows for adaptation to the local characteristics of the estimated responses and leads to high estimation performance for the cortical current distribution and the hemodynamic response functions. We utilize a Bayesian formulation with a variational Bayesian framework and obtain a fully automatic fusion algorithm. Simulations with synthetic data and experiments with real data from a multimodal study on face perception demonstrate the performance of the proposed method.

Canonical source reconstruction for MEG

We describe a simple and efficient solution to the problem of reconstructing electromagnetic sources into a canonical or standard anatomical space. Its simplicity rests upon incorporating subject-specific anatomy into the forward model in a way that eschews the need for cortical surface extraction. The forward model starts with a canonical cortical mesh, defined in a standard stereotactic space. The mesh is warped, in a nonlinear fashion, to match the subject's anatomy. This warping is the inverse of the transformation derived from spatial normalization of the subject's structural MRI image, using fully automated procedures that have been established for other imaging modalities. Electromagnetic lead fields are computed using the warped mesh, in conjunction with a spherical head model (which does not rely on individual anatomy). The ensuing forward model is inverted using an empirical Bayesian scheme that we have described previously in several publications. Critically, because anatomical information enters the forward model, there is no need to spatially normalize the reconstructed source activity. In other words, each source, comprising the mesh, has a predetermined and unique anatomical attribution within standard stereotactic space. This enables the pooling of data from multiple subjects and the reporting of results in stereotactic coordinates. Furthermore, it allows the graceful fusion of fMRI and MEG data within the same anatomical framework.

ERP generator anomalies in presymptomatic carriers of the Alzheimer's disease E280A PS-1 mutation

Although subtle anatomical anomalies long precede the onset of clinical symptoms in Alzheimer's disease, their impact on the reorganization of brain networks underlying cognitive functions has not been fully explored. A unique window into this reorganization is provided by presymptomatic cases of familial Alzheimer's disease (FAD). Here we studied neural circuitry related to semantic processing in presymptomatic FAD cases by estimating the intracranial sources of the N400 event-related potential (ERP). ERPs were obtained during a semantic-matching task from 24 presymptomatic carriers and 25 symptomatic carriers of the E280A presenilin-1 (PS-1) mutation, as well as 27 noncarriers (from the same families). As expected, the symptomatic-carrier group performed worse in the matching task and had lower N400 amplitudes than both asymptomatic groups, which did not differ from each other on these variables. However, N400 topography differed in mutation carrier groups with respect to the noncarriers. Intracranial source analysis evinced that the presymptomatic-carriers presented a decrease of N400 generator strength in right inferior-temporal and medial cingulate areas and increased generator strength in the left hippocampus and parahippocampus compared to the controls. This represents alterations in neural function without translation into behavioral impairments. Compared to controls, the symptomatic-carriers presented a similar anatomical shift in the distribution of N400 generators to that found in presymptomatic-carriers, albeit with a larger reduction in generator strength. The redistribution of N400 generators in presymptomatic-carriers indicates that early focal degeneration associated with the mutation induces neural reorganization, possibly contributing to a functional compensation that enables normal performance in the semantic task.

A preliminary study of the neural mechanisms of frustration in pediatric bipolar disorder using magnetoencephalography

BACKGROUND

Irritability is prevalent and impairing in pediatric bipolar disorder (BD) but has been minimally studied using neuroimaging techniques. We used magnetoencephalography (MEG) to study theta band oscillations in the anterior cingulate cortex (ACC) during frustration in BD youth. ACC theta power is associated with attention to emotional stimuli, and the ACC may mediate responses to frustrating stimuli.

METHODS

We used the affective Posner task, an attention paradigm that uses rigged feedback to induce frustration, to compare 20 medicated BD youth (14.9+/-2.0 years; 45% male) and 20 healthy controls (14.7+/-1.7 years; 45% male). MEG measured neuronal activity after negative and positive feedback; we also compared groups on reaction time, response accuracy, and self-reported affect. Patients met strict DSM-IV BD criteria and were euthymic. Controls had no psychiatric history.

RESULTS

BD youth reported more negative affective responses than controls. After negative feedback, BD subjects, relative to controls, displayed greater theta power in the right ACC and bilateral parietal lobe. After positive feedback, BD subjects displayed lower theta power in the left ACC than did controls. Correlations between MEG, behavior, and affect were nonsignificant.

CONCLUSION

In this first MEG study of BD youth, BD youth displayed patterns of theta oscillations in the ACC and parietal lobe in response to frustration-inducing negative feedback that differed from healthy controls. These data suggest that BD youth may display heightened processing of negative feedback and exaggerated self-monitoring after frustrating emotional stimuli. Future studies are needed with unmedicated bipolar youth, and comparison ADHD and anxiety groups.

Modulation of perception and brain activity by predictable trajectories of facial expressions

People track facial expression dynamics with ease to accurately perceive distinct emotions. Although the superior temporal sulcus (STS) appears to possess mechanisms for perceiving changeable facial attributes such as expressions, the nature of the underlying neural computations is not known. Motivated by novel theoretical accounts, we hypothesized that visual and motor areas represent expressions as anticipated motion trajectories. Using magnetoencephalography, we show predictable transitions between fearful and neutral expressions (compared with scrambled and static presentations) heighten activity in visual cortex as quickly as 165 ms poststimulus onset and later (237 ms) engage fusiform gyrus, STS and premotor areas. Consistent with proposed models of biological motion representation, we suggest that visual areas predictively represent coherent facial trajectories. We show that such representations bias emotion perception of subsequent static faces, suggesting that facial movements elicit predictions that bias perception. Our findings reveal critical processes evoked in the perception of dynamic stimuli such as facial expressions, which can endow perception with temporal continuity.

Early Visually Evoked Electrophysiological Responses Over the Human Brain (P1, N170) Show Stable Patterns of Face-Sensitivity from 4 years to Adulthood

Whether the development of face recognition abilities truly reflects changes in how faces, specifically, are perceived, or rather can be attributed to more general perceptual or cognitive development, is debated. Event-related potential (ERP) recordings on the scalp offer promise for this issue because they allow brain responses to complex visual stimuli to be relatively well isolated from other sensory, cognitive and motor processes. ERP studies in 5- to 16-year-old children report large age-related changes in amplitude, latency (decreases) and topographical distribution of the early visual components, the P1 and the occipito-temporal N170. To test the face specificity of these effects, we recorded high-density ERPs to pictures of faces, cars, and their phase-scrambled versions from 72 children between the ages of 4 and 17, and a group of adults. We found that none of the previously reported age-dependent changes in amplitude, latency or topography of the P1 or N170 were specific to faces. Most importantly, when we controlled for age-related variations of the P1, the N170 appeared remarkably similar in amplitude and topography across development, with much smaller age-related decreases in latencies than previously reported. At all ages the N170 showed equivalent face-sensitivity: it had the same topography and right hemisphere dominance, it was absent for meaningless (scrambled) stimuli, and larger and earlier for faces than cars. The data also illustrate the large amount of inter-individual and inter-trial variance in young children's data, which causes the N170 to merge with a later component, the N250, in grand-averaged data. Based on our observations, we suggest that the previously reported “bi-fid” N170 of young children is in fact the N250. Overall, our data indicate that the electrophysiological markers of face-sensitive perceptual processes are present from 4 years of age and do not appear to change throughout development.

Selecting forward models for MEG source-reconstruction using model-evidence

We investigated four key aspects of forward models for distributed solutions to the MEG inverse problem: 1) the nature of the cortical mesh constraining sources (derived from an individual's MRI, or inverse-normalised from a template mesh); 2) the use of single-sphere, overlapping spheres, or Boundary Element Model (BEM) head-models; 3) the density of the cortical mesh (3000 vs. 7000 vertices); and 4) whether source orientations were constrained to be normal to that mesh. These were compared within the context of two types of spatial prior on the sources: a single prior corresponding to a standard L2-minimum-norm (MNM) inversion, or multiple sparse priors (MSP). The resulting generative models were compared using a free-energy approximation to the Bayesian model-evidence after fitting multiple epochs of responses to faces or scrambled faces. Statistical tests of the free-energy, across nine participants, showed clear superiority of MSP over MNM models; with the former reconstructing deeper sources. Furthermore, there was 1) no evidence that an individually-defined cortical mesh was superior to an inverse-normalised canonical mesh, but 2) clear evidence that a BEM was superior to spherical head-models, provided individually-defined inner skull and scalp meshes were used. Finally, for MSP models, there was evidence that the combination of 3) higher density cortical meshes and 4) dipoles constrained to be normal to the mesh was superior to lower-density or freely-oriented sources (in contrast to the MNM models, in which free-orientation was optimal). These results have practical implications for MEG source reconstruction, particularly in the context of group studies.

MEG and EEG data fusion: simultaneous localisation of face-evoked responses

We present an empirical Bayesian scheme for distributed multimodal inversion of electromagnetic forward models of EEG and MEG signals. We used a generative model with common source activity and separate error components for each modality. Under this scheme, the weightings of error for each modality, relative to source components, are estimated automatically from the data, by optimising the model-evidence. This obviates the need for arbitrary user-defined weightings. To evaluate the scheme, we acquired three types of data simultaneously from twelve participants: total magnetic flux (as recorded by 102 magnetometers), orthogonal in-plane gradients of the magnetic field (as recorded by 204 planar gradiometers) and voltage differences in the electrical field (recorded by 70 electrodes). We assessed the relative precision of each sensor-type in terms of signal-to-noise ratio (SNR); using empirical sample variances and optimised estimators from the generative model. We then compared the localisation of face-evoked responses, using each modality separately, with that obtained by their "fusion" under the common generative model. Finally, we quantified the conditional precisions of the source estimates using their posterior covariance, confirming that EEG can improve MEG-based source reconstructions.

Bayesian model selection for group studies

Bayesian model selection (BMS) is a powerful method for determining the most likely among a set of competing hypotheses about the mechanisms that generated observed data. BMS has recently found widespread application in neuroimaging, particularly in the context of dynamic causal modelling (DCM). However, so far, combining BMS results from several subjects has relied on simple (fixed effects) metrics, e.g. the group Bayes factor (GBF), that do not account for group heterogeneity or outliers. In this paper, we compare the GBF with two random effects methods for BMS at the between-subject or group level. These methods provide inference on model-space using a classical and Bayesian perspective respectively. First, a classical (frequentist) approach uses the log model evidence as a subject-specific summary statistic. This enables one to use analysis of variance to test for differences in log-evidences over models, relative to inter-subject differences. We then consider the same problem in Bayesian terms and describe a novel hierarchical model, which is optimised to furnish a probability density on the models themselves. This new variational Bayes method rests on treating the model as a random variable and estimating the parameters of a Dirichlet distribution which describes the probabilities for all models considered. These probabilities then define a multinomial distribution over model space, allowing one to compute how likely it is that a specific model generated the data of a randomly chosen subject as well as the exceedance probability of one model being more likely than any other model. Using empirical and synthetic data, we show that optimising a conditional density of the model probabilities, given the log-evidences for each model over subjects, is more informative and appropriate than both the GBF and frequentist tests of the log-evidences. In particular, we found that the hierarchical Bayesian approach is considerably more robust than either of the other approaches in the presence of outliers. We expect that this new random effects method will prove useful for a wide range of group studies, not only in the context of DCM, but also for other modelling endeavours, e.g. comparing different source reconstruction methods for EEG/MEG or selecting among competing computational models of learning and decision-making.

Forward and backward connections in the brain: a DCM study of functional asymmetries

In this paper, we provide evidence for functional asymmetries in forward and backward connections that define hierarchical architectures in the brain. We exploit the fact that modulatory or nonlinear influences of one neuronal system on another (i.e., effective connectivity) entail coupling between different frequencies. Functional asymmetry in forward and backward connections was addressed by comparing dynamic causal models of MEG responses induced by visual processing of normal and scrambled faces. We compared models with and without nonlinear (between-frequency) coupling in both forward and backward connections. Bayesian model comparison indicated that the best model had nonlinear forward and backward connections. Using the best model we then quantified frequency-specific causal influences mediating observed spectral responses. We found a striking asymmetry between forward and backward connections; in which high (gamma) frequencies in higher cortical areas suppressed low (alpha) frequencies in lower areas. This suppression was significantly greater than the homologous coupling in the forward connections. Furthermore, exactly the asymmetry was observed when we examined face-selective coupling (i.e., coupling under faces minus scrambled faces). These results highlight the importance of nonlinear coupling among brain regions and point to a functional asymmetry between forward and backward connections in the human brain that is consistent with anatomical and physiological evidence from animal studies. This asymmetry is also consistent with functional architectures implied by theories of perceptual inference in the brain, based on hierarchical generative models.

Does physical interstimulus variance account for early electrophysiological face sensitive responses in the human brain? Ten lessons on the N170

A recent event-related potential (ERP) study (Thierry G., Martin, C.D., Downing, P., Pegna, A.J. 2007. Controlling for interstimulus perceptual variance abolishes N170 face selectivity. Nature Neuroscience, 10, 505-11) claimed that the larger occipito-temporal N170 response to pictures of faces than other categories -- the N170 effect -- is due to a methodological artifact in stimulus selection, specifically, a greater interstimulus physical variance between pictures of objects than faces in previous ERP studies which, when controlled, eliminates this N170 effect. This statement casts doubts on the validity of the conclusions reached by a whole tradition of electrophysiological experiments published over the past 15 years and questions the very interest of using the N170 to probe the time course of face processes in the human brain. Here we claim that this physical variance factor is ill-defined by Thierry et al. and cannot account for previous observations of a smaller N170 amplitude to nonface objects than faces without latency increase and component "smearing". Most importantly, this factor was controlled in previous studies that reported robust N170 effects. We demonstrate that the absence of N170 effect in the study of Thierry et al. is due to methodological flaws in the reported experiments, most notably measuring the N170 at the wrong electrode sites. Moreover, the authors attributed a modulation of N170 amplitude in their study to a differential interstimulus physical variance while it probably reflects a biased comparison of different quality sets of individual images. Here, by taking Thierry et al.'s study as an exemplar case of what should not be done in ERP research of visual categorization processes, we provide clarifications on a number of methodological and theoretical issues about the N170 and its largest amplitude to faces. More generally, we discuss the potential role of differential visual homogeneity of object categories as well as low-level visual properties versus high-level visual processes in accounting for early face-preferential responses and the question of the speed at which visual stimuli are categorized as faces. This survey of the literature points to the N170 as a critical event in the time course of face processes in the human brain.