Mapping human brain function with MEG and EEG: methods and validation

Mapping kainate activation of inner neurons in the rat retina

Kainate receptors mediate fast, excitatory synaptic transmission for a range of inner neurons in the mammalian retina. However, allocation of functional kainate receptors to known cell types and their sensitivity remains unresolved. Using the cation channel probe 1-amino-4-guanidobutane agmatine (AGB), we investigated kainate sensitivity of neurochemically identified cell populations within the structurally intact rat retina. Most inner retinal neuron populations responded to kainate in a concentration-dependent manner. OFF cone bipolar cells demonstrated the highest sensitivity of all inner neurons to kainate. Immunocytochemical localization of AGB and macromolecular markers confirmed that type 2 bipolar cells were part of this kainate-sensitive population. The majority of amacrine (ACs) and ganglion cells (GCs) showed kainate responses with different sensitivities between major neurochemical classes (γ-aminobutyric acid [GABA]/glycine ACs > glycine ACs > GABA ACs; glutamate [Glu]/weakly GABA GCs > Glu GCs). Conventional and displaced cholinergic ACs were highly responsive to kainate, whereas dopaminergic ACs do not appear to express functional kainate receptors. These findings further contribute to our understanding of neuronal networks in complex multicellular tissues.

Localized high gamma motor oscillations respond to perceived biologic motion

Power changes in the beta frequency range (17-25 Hz) in the human motor and premotor areas during action observation have been associated with the mirror neuron system and have been studied extensively. These changes mimic motor activity during actual motion execution, albeit reduced in strength. Recent noninvasive (EEG/magnetoencephalography) and invasive studies (electrocorticography) have shown that during actual motion, beta power changes are accompanied by highly localized changes in the high gamma band (70-100 Hz). In this study, we investigate, using 27-channel EEG in combination with a generic head model and a cortical mapping algorithm, whether such high gamma changes are also present during motion observation. Subjects were presented with a 2.7-second video of a moving hand, contrasted with a video of moving scenery of equal length. Our results show nonlateralized beta band decrease in power in response to the moving hand versus the response to the moving scenery. We also find significant increase in high gamma power. However, unlike the beta band response, increases in this band are lateralized, with a preference for the hemisphere of the dominant hand.

A review of EEG and MEG for brainnetome research

The majority of brain activities are performed by functionally integrating separate regions of the brain. Therefore, the synchronous operation of the brain's multiple regions or neuronal assemblies can be represented as a network with nodes that are interconnected by links. Because of the complexity of brain interactions and their varying effects at different levels of complexity, one of the corresponding authors of this paper recently proposed the brainnetome as a new -ome to explore and integrate the brain network at different scales. Because electroencephalography (EEG) and magnetoencephalography (MEG) are noninvasive and have outstanding temporal resolution and because they are the primary clinical techniques used to capture the dynamics of neuronal connections, they lend themselves to the analysis of the neural networks comprising the brainnetome. Because of EEG/MEG's applicability to brainnetome analyses, the aim of this review is to identify the procedures that can be used to form a network using EEG/MEG data in sensor or source space and to promote EEG/MEG network analysis for either neuroscience or clinical applications. To accomplish this aim, we show the relationship of the brainnetome to brain networks at the macroscale and provide a systematic review of network construction using EEG and MEG. Some potential applications of the EEG/MEG brainnetome are to use newly developed methods to associate the properties of a brainnetome with indices of cognition or disease conditions. Associations based on EEG/MEG brainnetome analysis may improve the comprehension of the functioning of the brain in neuroscience research or the recognition of abnormal patterns in neurological disease.

Dipole source localization of mouse electroencephalogram using the Fieldtrip toolbox

The mouse model is an important research tool in neurosciences to examine brain function and diseases with genetic perturbation in different brain regions. However, the limited techniques to map activated brain regions under specific experimental manipulations has been a drawback of the mouse model compared to human functional brain mapping. Here, we present a functional brain mapping method for fast and robust in vivo brain mapping of the mouse brain. The method is based on the acquisition of high density electroencephalography (EEG) with a microarray and EEG source estimation to localize the electrophysiological origins. We adapted the Fieldtrip toolbox for the source estimation, taking advantage of its software openness and flexibility in modeling the EEG volume conduction. Three source estimation techniques were compared: Distribution source modeling with minimum-norm estimation (MNE), scanning with multiple signal classification (MUSIC), and single-dipole fitting. Known sources to evaluate the performance of the localization methods were provided using optogenetic tools. The accuracy was quantified based on the receiver operating characteristic (ROC) analysis. The mean detection accuracy was high, with a false positive rate less than 1.3% and 7% at the sensitivity of 90% plotted with the MNE and MUSIC algorithms, respectively. The mean center-to-center distance was less than 1.2 mm in single dipole fitting algorithm. Mouse microarray EEG source localization using microarray allows a reliable method for functional brain mapping in awake mouse opening an access to cross-species study with human brain.

Short-time windowed covariance: a metric for identifying non-stationary, event-related covariant cortical sites

BACKGROUND

Electrocorticography (ECoG) signals can provide high spatio-temporal resolution and high signal to noise ratio recordings of local neural activity from the surface of the brain. Previous studies have shown that broad-band, spatially focal, high-frequency increases in ECoG signals are highly correlated with movement and other cognitive tasks and can be volitionally modulated. However, significant additional information may be present in inter-electrode interactions, but adding additional higher order inter-electrode interactions can be impractical from a computational aspect, if not impossible.

NEW METHOD

In this paper we present a new method of calculating high frequency interactions between electrodes called Short-Time Windowed Covariance (STWC) that builds on mathematical techniques currently used in neural signal analysis, along with an implementation that accelerates the algorithm by orders of magnitude by leveraging commodity, off-the-shelf graphics processing unit (GPU) hardware.

RESULTS

Using the hardware-accelerated implementation of STWC, we identify many types of event-related inter-electrode interactions from human ECoG recordings on global and local scales that have not been identified by previous methods. Unique temporal patterns are observed for digit flexion in both low- (10mm spacing) and high-resolution (3mm spacing) electrode arrays.

COMPARISON WITH EXISTING METHODS

Covariance is a commonly used metric for identifying correlated signals, but the standard covariance calculations do not allow for temporally varying covariance. In contrast STWC allows and identifies event-driven changes in covariance without identifying spurious noise correlations.

CONCLUSIONS

STWC can be used to identify event-related neural interactions whose high computational load is well suited to GPU capabilities.

Signal-to-noise ratio of the MEG signal after preprocessing

BACKGROUND

Magnetoencephalography (MEG) provides a direct measure of brain activity with high combined spatiotemporal resolution. Preprocessing is necessary to reduce contributions from environmental interference and biological noise.

NEW METHOD

The effect on the signal-to-noise ratio of different preprocessing techniques is evaluated. The signal-to-noise ratio (SNR) was defined as the ratio between the mean signal amplitude (evoked field) and the standard error of the mean over trials.

RESULTS

Recordings from 26 subjects obtained during and event-related visual paradigm with an Elekta MEG scanner were employed. Two methods were considered as first-step noise reduction: Signal Space Separation and temporal Signal Space Separation, which decompose the signal into components with origin inside and outside the head. Both algorithm increased the SNR by approximately 100%. Epoch-based methods, aimed at identifying and rejecting epochs containing eye blinks, muscular artifacts and sensor jumps provided an SNR improvement of 5-10%. Decomposition methods evaluated were independent component analysis (ICA) and second-order blind identification (SOBI). The increase in SNR was of about 36% with ICA and 33% with SOBI.

COMPARISON WITH EXISTING METHODS

No previous systematic evaluation of the effect of the typical preprocessing steps in the SNR of the MEG signal has been performed.

CONCLUSIONS

The application of either SSS or tSSS is mandatory in Elekta systems. No significant differences were found between the two. While epoch-based methods have been routinely applied the less often considered decomposition methods were clearly superior and therefore their use seems advisable.

An evaluation of EEG scanner's dependence on the imaging technique, forward model computation method, and array dimensionality

EEG source reconstruction involves solving an inverse problem that is highly ill-posed and dependent on a generally fixed forward propagation model. In this contribution we compare a low and high density EEG setup's dependence on correct forward modeling. Specifically, we examine how different forward models affect the source estimates obtained using four inverse solvers Minimum-Norm, LORETA, Minimum-Variance Adaptive Beamformer, and Sparse Bayesian Learning.

A new wavelet transform to sparsely represent cortical current densities for EEG/MEG inverse problems

The present study investigated the use of transform sparseness of cortical current density on human brain surface to improve electroencephalography/magnetoencephalography (EEG/MEG) inverse solutions. Transform sparseness was assessed by evaluating compressibility of cortical current densities in transform domains. To do that, a structure compression method from computer graphics was first adopted to compress cortical surface structure, either regular or irregular, into hierarchical multi-resolution meshes. Then, a new face-based wavelet method based on generated multi-resolution meshes was proposed to compress current density functions defined on cortical surfaces. Twelve cortical surface models were built by three EEG/MEG softwares and their structural compressibility was evaluated and compared by the proposed method. Monte Carlo simulations were implemented to evaluate the performance of the proposed wavelet method in compressing various cortical current density distributions as compared to other two available vertex-based wavelet methods. The present results indicate that the face-based wavelet method can achieve higher transform sparseness than vertex-based wavelet methods. Furthermore, basis functions from the face-based wavelet method have lower coherence against typical EEG and MEG measurement systems than vertex-based wavelet methods. Both high transform sparseness and low coherent measurements suggest that the proposed face-based wavelet method can improve the performance of L1-norm regularized EEG/MEG inverse solutions, which was further demonstrated in simulations and experimental setups using MEG data. Thus, this new transform on complicated cortical structure is promising to significantly advance EEG/MEG inverse source imaging technologies.

Localization of coherent sources by simultaneous MEG and EEG beamformer

Simultaneous magnetoencephalography (MEG) and electroencephalography (EEG) analysis is known generally to yield better localization performance than a single modality only. For simultaneous analysis, MEG and EEG data should be combined to maximize synergistic effects. Recently, beamformer for simultaneous MEG/EEG analysis was proposed to localize both radial and tangential components well, while single modality analyses could not detect them, or had relatively higher location bias. In practice, most interesting brain sources are likely to be activated coherently; however, conventional beamformer may not work properly for such coherent sources. To overcome this difficulty, a linearly constrained minimum variance (LCMV) beamformer may be used with a source suppression strategy. In this work, simultaneous MEG/EEG LCMV beamformer using source suppression was formulated firstly to investigate its capability over various suppression strategies. The localization performance of our proposed approach was examined mainly for coherent sources and compared thoroughly with the conventional simultaneous and single modality approaches, over various suppression strategies. For this purpose, we used numerous simulated data, as well as empirical auditory stimulation data. In addition, some strategic issues of simultaneous MEG/EEG analysis were discussed. Overall, we found that our simultaneous MEG/EEG LCMV beamformer using a source suppression strategy is greatly beneficial in localizing coherent sources.

Functional brain network organisation of children between 2 and 5 years derived from reconstructed activity of cortical sources of high-density EEG recordings

There is increasing interest in applying connectivity analysis to brain measures (Rubinov and Sporns, 2010), but most studies have relied on fMRI, which substantially limits the participant groups and numbers that can be studied. High-density EEG recordings offer a comparatively inexpensive easy-to-use alternative, but require channel-level connectivity analysis which currently lacks a common analytic framework and is very limited in spatial resolution. To address this problem, we have developed a new technique for studies of network development that overcomes the spatial constraint and obtains functional networks of cortical areas by using EEG source reconstruction with age-matched average MRI templates (He et al., 1999). In contrast to previously reported channel-level analysis, this approach provides information about the cortical areas most likely to be involved in the network as well as their functional relationship (Babiloni et al., 2005; De Vico Fallani et al., 2007). In this study, we applied source reconstruction with age-matched templates to task-free high-density EEG recordings in typically-developing children between 2 and 6 years of age (O'Reilly, 2012). Graph theory was then applied to the association strengths of 68 cortical regions of interest based on the Desikan-Killiany atlas. We found linear increases of mean node degree, mean clustering coefficient and maximum betweenness centrality between 2 years and 6 years of age. Characteristic path length was negatively correlated with age. The correlation of the network measures with age indicates network development towards more closely integrated networks similar to reports from other imaging modalities (Fair et al., 2008; Power et al., 2010). We also applied eigenvalue decomposition to obtain functional modules (Clayden et al., 2013). Connection strength within these modules did not change with age, and the modules resembled hub networks previously described for MRI (Hagmann et al., 2010; Power et al., 2010). The high temporal resolution of EEG additionally allowed us to distinguish between frequency bands potentially reflecting dynamic coupling between different neural oscillators. Generally, network parameters were similar for networks based on different frequency bands, but frequency band did emerge as a significant factor for clustering coefficient and characteristic path length. In conclusion, the current analysis shows that source reconstruction of high-density EEG recordings with appropriate head models offers a valuable tool for estimating network parameters in studies of brain development. The findings replicate the pattern of closer functional integration over development described for other imaging modalities (Fair et al., 2008; Power et al., 2010).

MEG source localization using invariance of noise space

We propose INvariance of Noise (INN) space as a novel method for source localization of magnetoencephalography (MEG) data. The method is based on the fact that modulations of source strengths across time change the energy in signal subspace but leave the noise subspace invariant. We compare INN with classical MUSIC, RAP-MUSIC, and beamformer approaches using simulated data while varying signal-to-noise ratios as well as distance and temporal correlation between two sources. We also demonstrate the utility of INN with actual auditory evoked MEG responses in eight subjects. In all cases, INN performed well, especially when the sources were closely spaced, highly correlated, or one source was considerably stronger than the other.

A signal-processing pipeline for magnetoencephalography resting-state networks

To study functional connectivity using magnetoencephalographic (MEG) data, the high-quality source-level reconstruction of brain activity constitutes a critical element. MEG resting-state networks (RSNs) have been documented by means of a dedicated processing pipeline: MEG recordings are decomposed by independent component analysis (ICA) into artifact and brain components (ICs); next, the channel maps associated with the latter ones are projected into the source space and the resulting voxel-wise weights are used to linearly combine the IC time courses. An extensive description of the proposed pipeline is provided here, along with an assessment of its performances with respect to alternative approaches. The following investigations were carried out: (1) ICA decomposition algorithm. Synthetic data are used to assess the sensitivity of the ICA results to the decomposition algorithm, by testing FastICA, INFOMAX, and SOBI. FastICA with deflation approach, a standard solution, provides the best decomposition. (2) Recombination of brain ICs versus subtraction of artifactual ICs (at the channel level). Both the recombination of the brain ICs in the sensor space and the classical procedure of subtracting the artifactual ICs from the recordings provide a suitable reconstruction, with a lower distortion using the latter approach. (3) Recombination of brain ICs after localization versus localization of artifact-corrected recordings. The brain IC recombination after source localization, as implemented in the proposed pipeline, provides a lower source-level signal distortion. (4) Detection of RSNs. The accuracy in source-level reconstruction by the proposed pipeline is confirmed by an improved specificity in the retrieval of RSNs from experimental data.

Wavelet based sparse source imaging technique

The present study proposed a novel multi-resolution wavelet to efficiently compress cortical current densities on the highly convoluted cortical surface. The basis function of the proposed wavelet is supported on triangular faces of the cortical mesh and it is thus named as the face-based wavelet to be distinguished from other vertex-based wavelets. The proposed face-based wavelet was used as a transform to gain the sparse representation of cortical sources and then was integrated into the framework of L1-norm regularizations with the purpose to improve the performance of sparse source imaging (SSI) in solving EEG/MEG inverse problems. Monte Carlo simulations were conducted with multiple extended sources (up to ten) at random locations. Experimental MEG data from an auditory induced language task was further adopted to evaluate the performance of the proposed wavelet based SSI technique. The present results indicated that the face-based wavelet can efficiently compress cortical current densities and has better performance than the vertex-based wavelet in helping inverse source reconstructions in terms of estimation accuracies in source localization and source extent. Experimental results further indicated improved detection performance of the face-based wavelet as compared with the vertex-based wavelet in the framework of SSI. It thus suggests the proposed wavelet based SSI can become a promising tool in studying brain functions and networks.

Source reconstruction accuracy of MEG and EEG Bayesian inversion approaches

Electro- and magnetoencephalography allow for non-invasive investigation of human brain activation and corresponding networks with high temporal resolution. Still, no correct network detection is possible without reliable source localization. In this paper, we examine four different source localization schemes under a common Variational Bayesian framework. A Bayesian approach to the Minimum Norm Model (MNM), an Empirical Bayesian Beamformer (EBB) and two iterative Bayesian schemes (Automatic Relevance Determination (ARD) and Greedy Search (GS)) are quantitatively compared. While EBB and MNM each use a single empirical prior, ARD and GS employ a library of anatomical priors that define possible source configurations. The localization performance was investigated as a function of (i) the number of sources (one vs. two vs. three), (ii) the signal to noise ratio (SNR; 5 levels) and (iii) the temporal correlation of source time courses (for the cases of two or three sources). We also tested whether the use of additional bilateral priors specifying source covariance for ARD and GS algorithms improved performance. Our results show that MNM proves effective only with single source configurations. EBB shows a spatial accuracy of few millimeters with high SNRs and low correlation between sources. In contrast, ARD and GS are more robust to noise and less affected by temporal correlations between sources. However, the spatial accuracy of ARD and GS is generally limited to the order of one centimeter. We found that the use of correlated covariance priors made no difference to ARD/GS performance.

Sparse imaging of cortical electrical current densities via wavelet transforms

While the cerebral cortex in the human brain is of functional importance, functions defined on this structure are difficult to analyze spatially due to its highly convoluted irregular geometry. This study developed a novel L1-norm regularization method using a newly proposed multi-resolution face-based wavelet method to estimate cortical electrical activities in electroencephalography (EEG) and magnetoencephalography (MEG) inverse problems. The proposed wavelets were developed based on multi-resolution models built from irregular cortical surface meshes, which were realized in this study too. The multi-resolution wavelet analysis was used to seek sparse representation of cortical current densities in transformed domains, which was expected due to the compressibility of wavelets, and evaluated using Monte Carlo simulations. The EEG/MEG inverse problems were solved with the use of the novel L1-norm regularization method exploring the sparseness in the wavelet domain. The inverse solutions obtained from the new method using MEG data were evaluated by Monte Carlo simulations too. The present results indicated that cortical current densities could be efficiently compressed using the proposed face-based wavelet method, which exhibited better performance than the vertex-based wavelet method. In both simulations and auditory experimental data analysis, the proposed L1-norm regularization method showed better source detection accuracy and less estimation errors than other two classic methods, i.e. weighted minimum norm (wMNE) and cortical low-resolution electromagnetic tomography (cLORETA). This study suggests that the L1-norm regularization method with the use of face-based wavelets is a promising tool for studying functional activations of the human brain.

DYNAMO: concurrent dynamic multi-model source localization method for EEG and/or MEG

This work presents a new dipolar method to estimate the neural sources from separate or combined EEG and MEG data. The novelty lies in the simultaneous estimation and integration of neural sources from different dynamic models with different parameters, leading to a dynamic multi-model solution for the EEG/MEG source localization problem. The first key aspect of this method is defining the source model as a dipolar dynamic system, which allows for the estimation of the probability distribution of the sources within the Bayesian filter estimation framework. A second important aspect is the consideration of several banks of filters that simultaneously estimate and integrate the neural sources of different models. A third relevant aspect is that the final probability estimate is a result of the probabilistic integration of the neural sources of numerous models. Such characteristics lead to a new approach that does not require a prior definition neither of the number of sources or of the underlying temporal dynamics, allowing for the specification of multiple initial prior estimates. The method was validated by three sensor modalities with simulated data designed to impose difficult estimation situations, and with real EEG data recorded in a feedback error-related potential paradigm. On the basis of these evaluations, the method was able to localize the sources with high accuracy.

Sparse current source estimation for MEG using loose orientation constraints

Spatially focal source estimates for magnetoencephalography (MEG) and electroencephalography (EEG) data can be obtained by imposing a minimum ℓ(1) -norm constraint on the distribution of the source currents. Anatomical information about the expected locations and orientations of the sources can be included in the source models. In particular, the sources can be assumed to be oriented perpendicular to the cortical surface. We introduce a minimum ℓ(1) -norm estimation source modeling approach with loose orientation constraints (ℓ(1) LOC), which integrates the estimation of the orientation, location, and strength of the source currents into a cost function to jointly model the residual error and the ℓ(1) -norm of the source estimates. Evaluation with simulated MEG data indicated that the ℓ(1) LOC method can provide low spatial dispersion, high localization accuracy, and high source detection rates. Application to somatosensory and auditory MEG data resulted in physiologically reasonable source distributions. The proposed ℓ(1) LOC method appears useful for incorporating anatomical information about the source orientations into sparse source estimation of MEG data.

Performance evaluation of the Champagne source reconstruction algorithm on simulated and real M/EEG data

In this paper, we present an extensive performance evaluation of a novel source localization algorithm, Champagne. It is derived in an empirical Bayesian framework that yields sparse solutions to the inverse problem. It is robust to correlated sources and learns the statistics of non-stimulus-evoked activity to suppress the effect of noise and interfering brain activity. We tested Champagne on both simulated and real M/EEG data. The source locations used for the simulated data were chosen to test the performance on challenging source configurations. In simulations, we found that Champagne outperforms the benchmark algorithms in terms of both the accuracy of the source localizations and the correct estimation of source time courses. We also demonstrate that Champagne is more robust to correlated brain activity present in real MEG data and is able to resolve many distinct and functionally relevant brain areas with real MEG and EEG data.

A spatiotemporal dynamic distributed solution to the MEG inverse problem

MEG/EEG are non-invasive imaging techniques that record brain activity with high temporal resolution. However, estimation of brain source currents from surface recordings requires solving an ill-conditioned inverse problem. Converging lines of evidence in neuroscience, from neuronal network models to resting-state imaging and neurophysiology, suggest that cortical activation is a distributed spatiotemporal dynamic process, supported by both local and long-distance neuroanatomic connections. Because spatiotemporal dynamics of this kind are central to brain physiology, inverse solutions could be improved by incorporating models of these dynamics. In this article, we present a model for cortical activity based on nearest-neighbor autoregression that incorporates local spatiotemporal interactions between distributed sources in a manner consistent with neurophysiology and neuroanatomy. We develop a dynamic maximum a posteriori expectation-maximization (dMAP-EM) source localization algorithm for estimation of cortical sources and model parameters based on the Kalman Filter, the Fixed Interval Smoother, and the EM algorithms. We apply the dMAP-EM algorithm to simulated experiments as well as to human experimental data. Furthermore, we derive expressions to relate our dynamic estimation formulas to those of standard static models, and show how dynamic methods optimally assimilate past and future data. Our results establish the feasibility of spatiotemporal dynamic estimation in large-scale distributed source spaces with several thousand source locations and hundreds of sensors, with resulting inverse solutions that provide substantial performance improvements over static methods.

Dynamic activation of frontal, parietal, and sensory regions underlying anticipatory visual spatial attention

Although it is well established that multiple frontal, parietal, and occipital regions in humans are involved in anticipatory deployment of visual spatial attention, less is known about the electrophysiological signals in each region across multiple subsecond periods of attentional deployment. We used MEG measures of cortical stimulus-locked, signal-averaged (event-related field) activity during a task in which a symbolic cue directed covert attention to the relevant location on each trial. Direction-specific attention effects occurred in different cortical regions for each of multiple time periods during the delay between the cue and imperative stimulus. A sequence of activation from V1/V2 to extrastriate, parietal, and frontal regions occurred within 110 ms after cue, possibly related to extraction of cue meaning. Direction-specific activations ∼300 ms after cue in frontal eye field (FEF), lateral intraparietal area (LIP), and cuneus support early covert targeting of the cued location. This was followed by coactivation of a frontal-parietal system [superior frontal gyrus (SFG), middle frontal gyrus (MFG), LIP, anterior intraparietal sulcus (IPSa)] that may coordinate the transition from targeting the cued location to sustained deployment of attention to both space and feature in the last period. The last period involved direction-specific activity in parietal regions and both dorsal and ventral sensory regions [LIP, IPSa, ventral IPS, lateral occipital region, and fusiform gyrus], which was accompanied by activation that was not direction specific in right hemisphere frontal regions (FEF, SFG, MFG). Behavioral performance corresponded with the magnitude of attention-related activity in different brain regions at each time period during deployment. The results add to the emerging electrophysiological characterization of different cortical networks that operate during anticipatory deployment of visual spatial attention.

Simultaneous EEG and MEG source reconstruction in sparse electromagnetic source imaging

Electroencephalography (EEG) and magnetoencephalography (MEG) have different sensitivities to differently configured brain activations, making them complimentary in providing independent information for better detection and inverse reconstruction of brain sources. In the present study, we developed an integrative approach, which integrates a novel sparse electromagnetic source imaging method, i.e., variation-based cortical current density (VB-SCCD), together with the combined use of EEG and MEG data in reconstructing complex brain activity. To perform simultaneous analysis of multimodal data, we proposed to normalize EEG and MEG signals according to their individual noise levels to create unit-free measures. Our Monte Carlo simulations demonstrated that this integrative approach is capable of reconstructing complex cortical brain activations (up to 10 simultaneously activated and randomly located sources). Results from experimental data showed that complex brain activations evoked in a face recognition task were successfully reconstructed using the integrative approach, which were consistent with other research findings and validated by independent data from functional magnetic resonance imaging using the same stimulus protocol. Reconstructed cortical brain activations from both simulations and experimental data provided precise source localizations as well as accurate spatial extents of localized sources. In comparison with studies using EEG or MEG alone, the performance of cortical source reconstructions using combined EEG and MEG was significantly improved. We demonstrated that this new sparse ESI methodology with integrated analysis of EEG and MEG data could accurately probe spatiotemporal processes of complex human brain activations. This is promising for noninvasively studying large-scale brain networks of high clinical and scientific significance.

Sensory handedness is not reflected in cortical responses after basic nerve stimulation: a MEG study

Motor dominance is well established, but sensory dominance is much less clear. We therefore studied the cortical evoked magnetic fields using magnetoencephalography (MEG) in a group of 20 healthy right handed subjects in order to examine whether standard electrical stimulation of the median and ulnar nerve demonstrated sensory lateralization. The global field power (GFP) curves, as an indication of cortical activation, did not depict sensory lateralization to the dominant left hemisphere. Comparison of the M20, M30, and M70 peak latencies and GFP values exhibited no statistical differences between the hemispheres, indicating no sensory hemispherical dominance at these latencies for each nerve. Field maps at these latencies presented a first and second polarity reversal for both median and ulnar stimulation. Spatial dipole position parameters did not reveal statistical left-right differences at the M20, M30 and M70 peaks for both nerves. Neither did the dipolar strengths at M20, M30 and M70 show a statistical left-right difference for both nerves. Finally, the Laterality Indices of the M20, M30 and M70 strengths did not indicate complete lateralization to one of the hemispheres. After electrical median and ulnar nerve stimulation no evidence was found for sensory hand dominance in brain responses of either hand, as measured by MEG. The results can provide a new assessment of patients with sensory dysfunctions or perceptual distortion when sensory dominance occurs way beyond the estimated norm.

Anatomically constrained minimum variance beamforming applied to EEG

Neural activity as measured non-invasively using electroencephalography (EEG) or magnetoencephalography (MEG) originates in the cortical gray matter. In the cortex, pyramidal cells are organized in columns and activated coherently, leading to current flow perpendicular to the cortical surface. In recent years, beamforming algorithms have been developed, which use this property as an anatomical constraint for the locations and directions of potential sources in MEG data analysis. Here, we extend this work to EEG recordings, which require a more sophisticated forward model due to the blurring of the electric current at tissue boundaries where the conductivity changes. Using CT scans, we create a realistic three-layer head model consisting of tessellated surfaces that represent the cerebrospinal fluid-skull, skull-scalp, and scalp-air boundaries. The cortical gray matter surface, the anatomical constraint for the source dipoles, is extracted from MRI scans. EEG beamforming is implemented on simulated sets of EEG data for three different head models: single spherical, multi-shell spherical, and multi-shell realistic. Using the same conditions for simulated EEG and MEG data, it is shown (and quantified by receiver operating characteristic analysis) that EEG beamforming detects radially oriented sources, to which MEG lacks sensitivity. By merging several techniques, such as linearly constrained minimum variance beamforming, realistic geometry forward solutions, and cortical constraints, we demonstrate it is possible to localize and estimate the dynamics of dipolar and spatially extended (distributed) sources of neural activity.

A spline framework for estimating the EEG surface laplacian using the Euclidean metric

This letter develops a framework for EEG analysis and similar applications based on polyharmonic splines. This development overcomes a basic problem with the method of splines in the Euclidean setting: that it does not work on low-degree algebraic surfaces such as spherical and ellipsoidal scalp models. The method's capability is illustrated through simulations on the three-sphere model and using empirical data.

Retinotopic mapping of the primary visual cortex - a challenge for MEG imaging of the human cortex

Magnetoencephalography (MEG) can be used to reconstruct neuronal activity with high spatial and temporal resolution. However, this reconstruction problem is ill-posed, and requires the use of prior constraints in order to produce a unique solution. At present there are a multitude of inversion algorithms, each employing different assumptions, but one major problem when comparing the accuracy of these different approaches is that often the true underlying electrical state of the brain is unknown. In this study, we explore one paradigm, retinotopic mapping in the primary visual cortex (V1), for which the ground truth is known to a reasonable degree of accuracy, enabling the comparison of MEG source reconstructions with the true electrical state of the brain. Specifically, we attempted to localize, using a beanforming method, the induced responses in the visual cortex generated by a high contrast, retinotopically varying stimulus. Although well described in primate studies, it has been an open question whether the induced gamma power in humans due to high contrast gratings derives from V1 rather than the prestriate cortex (V2). We show that the beanformer source estimate in the gamma and theta bands does vary in a manner consistent with the known retinotopy of V1. However, these peak locations, although retinotopically organized, did not accurately localize to the cortical surface. We considered possible causes for this discrepancy and suggest that improved MEG/magnetic resonance imaging co-registration and the use of more accurate source models that take into account the spatial extent and shape of the active cortex may, in future, improve the accuracy of the source reconstructions.

Gustatory imagery reveals functional connectivity from the prefrontal to insular cortices traced with magnetoencephalography

Our experience and prejudice concerning food play an important role in modulating gustatory information processing; gustatory memory stored in the central nervous system influences gustatory information arising from the peripheral nervous system. We have elucidated the mechanism of the “top-down” modulation of taste perception in humans using functional magnetic resonance imaging (fMRI) and demonstrated that gustatory imagery is mediated by the prefrontal (PFC) and insular cortices (IC). However, the temporal order of activation of these brain regions during gustatory imagery is still an open issue. To explore the source of “top-down” signals during gustatory imagery tasks, we analyzed the temporal activation patterns of activated regions in the cerebral cortex using another non-invasive brain imaging technique, magnetoencephalography (MEG). Gustatory imagery tasks were presented by words (Letter G-V) or pictures (Picture G-V) of foods/beverages, and participants were requested to recall their taste. In the Letter G-V session, 7/9 (77.8%) participants showed activation in the IC with a latency of 401.7±34.7 ms (n = 7) from the onset of word exhibition. In 5/7 (71.4%) participants who exhibited IC activation, the PFC was activated prior to the IC at a latency of 315.2±56.5 ms (n = 5), which was significantly shorter than the latency to the IC activation. In the Picture G-V session, the IC was activated in 6/9 (66.7%) participants, and only 1/9 (11.1%) participants showed activation in the PFC. There was no significant dominance between the right and left IC or PFC during gustatory imagery. These results support those from our previous fMRI study in that the Letter G-V session rather than the Picture G-V session effectively activates the PFC and IC and strengthen the hypothesis that the PFC mediates “top-down” control of retrieving gustatory information from the storage of long-term memories and in turn activates the IC.

Electrophysiological imaging of brain activity and connectivity-challenges and opportunities

Unlocking the dynamic inner workings of the brain continues to remain a grand challenge of the 21st century. To this end, functional neuroimaging modalities represent an outstanding approach to better understand the mechanisms of both normal and abnormal brain functions. The ability to image brain function with ever increasing spatial and temporal resolution has made a significant leap over the past several decades. Further delineation of functional networks could lead to improved understanding of brain function in both normal and diseased states. This paper reviews recent advancements and current challenges in dynamic functional neuroimaging techniques, including electrophysiological source imaging, multimodal neuroimaging integrating fMRI with EEG/MEG, and functional connectivity imaging.

Brainstorm: a user-friendly application for MEG/EEG analysis

Brainstorm is a collaborative open-source application dedicated to magnetoencephalography (MEG) and electroencephalography (EEG) data visualization and processing, with an emphasis on cortical source estimation techniques and their integration with anatomical magnetic resonance imaging (MRI) data. The primary objective of the software is to connect MEG/EEG neuroscience investigators with both the best-established and cutting-edge methods through a simple and intuitive graphical user interface (GUI).

Sparse cortical current density imaging in motor potentials induced by finger movement

Predominant components in electro- or magneto-encephalography (EEG/MEG) are scalp projections of synchronized neuronal electrical activity distributed over cortical structures. Reconstruction of cortical sources underlying EEG/MEG can thus be achieved with the use of the cortical current density (CCD) model. We have developed a sparse electromagnetic source imaging method based on the CCD model, named as the variation-based cortical current density (VB-SCCD) algorithm, and have shown that it has much enhanced performance in reconstructing extended cortical sources in simulations (Ding 2009 Phys. Med. Biol. 54 2683-97). The present study aims to evaluate the performance of VB-SCCD, for the first time, using experimental data obtained from six participants. The results indicate that the VB-SCCD algorithm is able to successfully reveal spatially distributed cortical sources behind motor potentials induced by visually cued repetitive finger movements, and their dynamic patterns, with millisecond resolution. These findings of motor sources and cortical systems are supported by the physiological knowledge of motor control and evidence from various neuroimaging studies with similar experiments. Furthermore, our present results indicate the improvement of cortical source resolvability of VB-SCCD, as compared with two other classical algorithms. The proposed solver embedded in VB-SCCD is able to handle large-scale computational problems, which makes the use of high-density CCD models possible and, thus, reduces model misspecifications. The present results suggest that VB-SCCD provides high resolution source reconstruction capability and is a promising tool for studying complicated dynamic systems of brain activity for basic neuroscience and clinical neuropsychiatric research.

Dynamic EEG-informed fMRI modeling of the pain matrix using 20-ms root mean square segments

Previous studies on the spatio-temporal dynamics of cortical pain processing using electroencephalography (EEG), magnetoencephalography (MEG), or intracranial recordings point towards a high degree of parallelism, e.g. parallel instead of sequential activation of primary and secondary somatosensory areas or simultaneous activation of somatosensory areas and the mid-cingulate cortex. However, because of the inverse problem, EEG and MEG provide only limited spatial resolution and certainty about the generators of cortical pain-induced electromagnetic activity, especially when multiple sources are simultaneously active. On the other hand, intracranial recordings are invasive and do not provide whole-brain coverage. In this study, we thought to investigate the spatio-temporal dynamics of cortical pain processing in 10 healthy subjects using simultaneous EEG/functional magnetic resonance imaging (fMRI). Voltages of 20 ms segments of the EEG root mean square (a global, largely reference-free measure of event-related EEG activity) in a time window 0-400 ms poststimulus were used to model trial-to-trial fluctuations in the fMRI blood oxygen level dependent (BOLD) signal. EEG-derived regressors explained additional variance in the BOLD signal from 140 ms poststimulus onward. According to this analysis, the contralateral parietal operculum was the first cortical area to become activated upon painful laser stimulation. The activation pattern in BOLD analyses informed by subsequent EEG-time windows suggests largely parallel signal processing in the bilateral operculo-insular and mid-cingulate cortices. In that regard, our data are in line with previous reports. However, the approach presented here is noninvasive and bypasses the inverse problem using only temporal information from the EEG.

The generation of tetrahedral mesh models for neuroanatomical MRI

In this article, we describe a detailed method for automatically generating tetrahedral meshes from 3D images having multiple region labels. An adaptively sized tetrahedral mesh modeling approach is described that is capable of producing meshes conforming precisely to the voxelized regions in the image. Efficient tetrahedral mesh improvement is then performed minimizing an energy function containing three terms: a smoothing term to remove the voxelization, a fidelity term to maintain continuity with the image data, and a novel elasticity term to prevent the tetrahedra from becoming flattened or inverted as the mesh deforms while allowing the voxelization to be removed entirely. The meshing algorithm is applied to structural MR image data that has been automatically segmented into 56 neuroanatomical sub-divisions as well as on two other examples. The resulting tetrahedral representation has several desirable properties such as tetrahedra with dihedral angles away from 0 and 180 degrees, smoothness, and a high resolution. Tetrahedral modeling via the approach described here has applications in modeling brain structure in normal as well as diseased brain in human and non-human data and facilitates examination of 3D object deformations resulting from neurological illness (e.g. Alzheimer's disease), development, and/or aging.

Sensitivity of beamformer source analysis to deficiencies in forward modeling

Beamforming approaches have recently been developed for the field of electroencephalography (EEG) and magnetoencephalography (MEG) source analysis and opened up new applications within various fields of neuroscience. While the number of beamformer applications thus increases fast-paced, fundamental methodological considerations, especially the dependence of beamformer performance on leadfield accuracy, is still quite unclear. In this article, we present a systematic study on the influence of improper volume conductor modeling on the source reconstruction performance of an EEG-data based synthetic aperture magnetometry (SAM) beamforming approach. A finite element model of a human head is derived from multimodal MR images and serves as a realistic volume conductor model. By means of a theoretical analysis followed by a series of computer simulations insight is gained into beamformer performance with respect to reconstruction errors in peak location, peak amplitude, and peak width resulting from geometry and anisotropy volume conductor misspecifications, sensor noise, and insufficient sensor coverage. We conclude that depending on source position, sensor coverage, and accuracy of the volume conductor model, localization errors up to several centimeters must be expected. As we could show that the beamformer tries to find the best fitting leadfield (least squares) with respect to its scanning space, this result can be generalized to other localization methods. More specific, amplitude, and width of the beamformer peaks significantly depend on the interaction between noise and accuracy of the volume conductor model. The beamformer can strongly profit from a high signal-to-noise ratio, but this requires a sufficiently realistic volume conductor model.

Occipital gamma activation during Vipassana meditation

Long-term Vipassana meditators sat in meditation vs. a control rest (mind-wandering) state for 21 min in a counterbalanced design with spontaneous EEG recorded. Meditation state dynamics were measured with spectral decomposition of the last 6 min of the eyes-closed silent meditation compared to control state. Meditation was associated with a decrease in frontal delta (1-4 Hz) power, especially pronounced in those participants not reporting drowsiness during meditation. Relative increase in frontal theta (4-8 Hz) power was observed during meditation, as well as significantly increased parieto-occipital gamma (35-45 Hz) power, but no other state effects were found for the theta (4-8 Hz), alpha (8-12 Hz), or beta (12-25 Hz) bands. Alpha power was sensitive to condition order, and more experienced meditators exhibited no tendency toward enhanced alpha during meditation relative to the control task. All participants tended to exhibit decreased alpha in association with reported drowsiness. Cross-experimental session occipital gamma power was the greatest in meditators with a daily practice of 10+ years, and the meditation-related gamma power increase was similarly the strongest in such advanced practitioners. The findings suggest that long-term Vipassana meditation contributes to increased occipital gamma power related to long-term meditational expertise and enhanced sensory awareness.

Identifying true cortical interactions in MEG using the nulling beamformer

Modeling functional brain interaction networks using non-invasive EEG and MEG data is more challenging than using intracranial recording data. This is because most interaction measures are not robust to the cross-talk (interference) between cortical regions, which may arise due to the limited spatial resolution of EEG/MEG inverse procedures. In this article, we describe a modified beamforming approach to accurately measure cortical interactions from EEG/MEG data, designed to suppress cross-talk between cortical regions. We estimate interaction measures from the output of the modified beamformer and test for statistical significance using permutation tests. Since the underlying neuronal sources and their interactions are unknown in real MEG data, we demonstrate the performance of the proposed beamforming method in a novel simulation scheme, where intracranial recordings from a macaque monkey are used as neural sources to simulate realistic MEG signals. The advantage of this approach is that local field potentials are more realistic representations of true neuronal sources than simulation models and therefore are more suitable to indicate the performance of our nulling beamforming method.

Topographic localization of brain activation in diffuse optical imaging using spherical wavelets

Diffuse optical imaging is a non-invasive technique that uses near-infrared light to measure changes in brain activity through an array of sensors placed on the surface of the head. Compared to functional MRI, optical imaging has the advantage of being portable while offering the ability to record functional changes in both oxy- and deoxy-hemoglobin within the brain at a high temporal resolution. However, the reconstruction of accurate spatial images of brain activity from optical measurements represents an ill-posed and underdetermined problem that requires regularization. These reconstructions benefit from incorporating prior information about the underlying spatial structure and function of the brain. In this work, we describe a novel image reconstruction approach which uses surface-based wavelets derived from structural MRI to incorporate high-resolution anatomical and structural prior information about the brain. This surface-based approach is used to approximate brain activation patterns through the reconstruction and presentation of topographical (two-dimensional) maps of brain activation directly onto the folded surface of the cortex. The set of wavelet coefficients is directly estimated by a truncated singular-value decomposition based pseudo-inversion of the wavelet projection of the optical forward model. We use a reconstruction metric based on Shannon entropy which quantifies the sparse loading of the wavelet coefficients and is used to determine the optimal truncation and regularization of this inverse model. In this work, examples of the performance of this model are illustrated for several cases of numerical simulation and experimental data with comparison to functional magnetic resonance imaging.

MEG's ability to localise accurately weak transient neural sources

OBJECTIVE

To investigate the accurate localisation of weak, transient, neural sources under conditions of varying difficulty.

METHODS

Multiple dipolar sources placed within a head-shaped phantom at superficial and deep locations were driven separately or simultaneously by a short-lasting current with varied amplitudes. Artificial MEG signals that were very similar to the human High Frequency Oscillations (HFO) were produced. MEG signals of HFO were also recorded from median nerve stimulation. Different inverse techniques were used to localise the phantom dipoles and the human HFO generators.

RESULTS

The human HFO were measured around 200 and 600Hz by using only 120 trials. The 200Hz HFO were localised to BA3b. The superficial phantom's source was localised with an accuracy of 2-3mm by all inverse techniques (120 trials). The 'subcortical' source was localised with an error of approximately 5mm. Localisation of deeper 'thalamic' sources required more trials.

CONCLUSION

MEG can detect and localise weak transient activations and the human HFO with an accuracy of a few mm at cortical and subcortical regions even when a small number of trials are used.

SIGNIFICANCE

Localizing HFO to specific anatomical structures has high clinical utility, for example in epilepsy, where discrete HFO appears to be generated just before focal epileptic activity.

High gamma mapping using EEG

High gamma (HG) power changes during motor activity, especially at frequencies above 70 Hz, play an important role in functional cortical mapping and as control signals for BCI (brain-computer interface) applications. Most studies of HG activity have used ECoG (electrocorticography) which provides high-quality spatially localized signals, but is an invasive method. Recent studies have shown that non-invasive modalities such as EEG and MEG can also detect task-related HG power changes. We show here that a 27 channel EEG (electroencephalography) montage provides high-quality spatially localized signals non-invasively for HG frequencies ranging from 83 to 101 Hz. We used a generic head model, a weighted minimum norm least squares (MNLS) inverse method, and a self-paced finger movement paradigm. The use of an inverse method enables us to map the EEG onto a generic cortex model. We find the HG activity during the task to be well localized in the contralateral motor area. We find HG power increases prior to finger movement, with average latencies of 462 ms and 82 ms before EMG (electromyogram) onset. We also find significant phase-locking between contra- and ipsilateral motor areas over a similar HG frequency range; here the synchronization onset precedes the EMG by 400 ms. We also compare our results to ECoG data from a similar paradigm and find EEG mapping and ECoG in good agreement. Our findings demonstrate that mapped EEG provides information on two important parameters for functional mapping and BCI which are usually only found in HG of ECoG signals: spatially localized power increases and bihemispheric phase-locking.

Robust Bayesian estimation of the location, orientation, and time course of multiple correlated neural sources using MEG

The synchronous brain activity measured via MEG (or EEG) can be interpreted as arising from a collection (possibly large) of current dipoles or sources located throughout the cortex. Estimating the number, location, and time course of these sources remains a challenging task, one that is significantly compounded by the effects of source correlations and unknown orientations and by the presence of interference from spontaneous brain activity, sensor noise, and other artifacts. This paper derives an empirical Bayesian method for addressing each of these issues in a principled fashion. The resulting algorithm guarantees descent of a cost function uniquely designed to handle unknown orientations and arbitrary correlations. Robust interference suppression is also easily incorporated. In a restricted setting, the proposed method is shown to produce theoretically zero reconstruction error estimating multiple dipoles even in the presence of strong correlations and unknown orientations, unlike a variety of existing Bayesian localization methods or common signal processing techniques such as beamforming and sLORETA. Empirical results on both simulated and real data sets verify the efficacy of this approach.

Somatosensory system in two types of motor reorganization in congenital hemiparesis: topography and function

This study investigates the (re-)organization of somatosensory functions following early brain lesions. Using functional magnetic resonance imaging (fMRI), passive hand movement was studied. Transcranial magnetic stimulation (TMS) and magnetoencephalography (MEG) were used as complementary methods. fMRI data was analyzed on the first level with regard to topographical variability; second-level group effects as well as the overall integrity of the somatosensory circuitry were also assessed. Subjects with unilateral brain lesions occurring in the third trimester of pregnancy or perinatally with different types of motor reorganization were included: patients with regular, contralateral motor organization following middle cerebral artery strokes (CONTRA(MCA), n = 6) and patients with reorganized, ipsilateral motor functions due to periventricular lesions (IPSI(PL), n = 8). Motor impairment was similar, but sensory impairment was more pronounced in the CONTRA(MCA) group. Using fMRI and MEG, both groups showed a normal pattern with a contralateral somatosensory representation, despite the transhemispherically reorganized primary motor cortex in the IPSI(PL) group, as verified by TMS. Activation topography for the paretic hands was more variable than for the nonparetic hand in both groups. The cortico-cerebellar circuitry was well-preserved in almost all subjects. We conclude that in both models of motor reorganization, no interhemispheric reorganization of somatosensory functions occurred. Also, no relevant intrahemispheric reorganization was observed apart from a higher topographical variability of fMRI activations. This preserved pattern of somatosensory organization argues in favor of a differential lesion effect on motor and somatosensory functions and demonstrates a limited compensatory potential for the latter.

Reconstructing cortical current density by exploring sparseness in the transform domain

In the present study, we have developed a novel electromagnetic source imaging approach to reconstruct extended cortical sources by means of cortical current density (CCD) modeling and a novel EEG imaging algorithm which explores sparseness in cortical source representations through the use of L1-norm in objective functions. The new sparse cortical current density (SCCD) imaging algorithm is unique since it reconstructs cortical sources by attaining sparseness in a transform domain (the variation map of cortical source distributions). While large variations are expected to occur along boundaries (sparseness) between active and inactive cortical regions, cortical sources can be reconstructed and their spatial extents can be estimated by locating these boundaries. We studied the SCCD algorithm using numerous simulations to investigate its capability in reconstructing cortical sources with different extents and in reconstructing multiple cortical sources with different extent contrasts. The SCCD algorithm was compared with two L2-norm solutions, i.e. weighted minimum norm estimate (wMNE) and cortical LORETA. Our simulation data from the comparison study show that the proposed sparse source imaging algorithm is able to accurately and efficiently recover extended cortical sources and is promising to provide high-accuracy estimation of cortical source extents.

Spatiotemporal dipole source localization of face processing ERPs in adolescents: a preliminary study

Background

Despite extensive investigation of the neural systems for face perception and emotion recognition in adults and young children in the past, the precise temporal activation of brain sources specific to the processing of emotional facial expressions in older children and adolescents is not well known. This preliminary study aims to trace the spatiotemporal dynamics of facial emotion processing during adolescence and provide a basis for future developmental studies and comparisons with patient populations that have social-emotional deficits such as autism.

Methods

We presented pictures showing happy, angry, fearful, or neutral facial expressions to healthy adolescents (aged 10–16 years) and recorded 128-channel event-related potentials (ERPs) while they performed an emotion discrimination task. ERP components were analyzed for effects of age and emotion on amplitude and latency. The underlying cortical sources of scalp ERP activity were modeled as multiple equivalent current dipoles using Brain Electrical Source Analysis (BESA).

Results

Initial global/holistic processing of faces (P1) took place in the visual association cortex (lingual gyrus) around 120 ms post-stimulus. Next, structural encoding of facial features (N170) occurred between 160–200 ms in the inferior temporal/fusiform region, and perhaps early emotion processing (Vertex Positive Potential or VPP) in the amygdala and orbitofrontal cortex. Finally, cognitive analysis of facial expressions (P2) in the prefrontal cortex and emotional reactions in somatosensory areas were observed from about 230 ms onwards. The temporal sequence of cortical source activation in response to facial emotion processing was occipital, prefrontal, fusiform, parietal for young adolescents and occipital, limbic, inferior temporal, and prefrontal for older adolescents.

Conclusion

This is a first report of high-density ERP dipole source analysis in healthy adolescents which traces the sequence of neural activity within the first 500 ms of categorizing emotion from faces. Our spatio-temporal brain source models showed the presence of adult-like cortical networks for face processing in adolescents, whose functional specificity to different emotions appear to be not yet fully mature. Age-related differences in brain activation patterns illustrate the continued development and maturation of distinct neural systems for processing facial expressions during adolescence and possible changes in emotion perception, experience, and reaction with age.