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

We describe the use of random field and permutation methods to detect activation in cortically constrained maps of current density computed from MEG data. The methods are applicable to any inverse imaging method that maps event-related MEG to a coregistered cortical surface. These approaches also extend directly to images computed from event-related EEG data. We determine statistical thresholds that control the familywise error rate (FWER) across space or across both space and time. Both random field and permutation methods use the distribution of the maximum statistic under the null hypothesis to find FWER thresholds. The former methods make assumptions on the distribution and smoothness of the data and use approximate analytical solutions, the latter resample the data and rely on empirical distributions. Both methods account for spatial and temporal correlation in the cortical maps. Unlike previous nonparametric work in neuroimaging, we address the problem of nonuniform specificity that can arise without a Gaussianity assumption. We compare and evaluate the methods on simulated data and experimental data from a somatosensory-evoked response study. We find that the random field methods are conservative with or without smoothing, though with a 5 vertex FWHM smoothness, they are close to exact. Our permutation methods demonstrated exact specificity in simulation studies. In real data, the permutation method was not as sensitive as the RF method, although this could be due to violations of the random field theory assumptions.

Words have been found to elicit a negative potential at the scalp peaking at �400 ms that is strongly modulated by semantic context. The current study used whole-head magnetoencephalography (MEG) as male subjects read sentences ending with semantically congruous or incongruous words. Compared with congruous words, sentence-terminal incongruous words consistently evoked a large magnetic field over the left hemisphere, peaking at �450 ms. Source modeling at this latency with conventional equivalent current dipoles (ECDs) placed the N400m generator in or near the left superior temporal sulcus. A distributed solution constrained to the cortical surface suggested a sequence of differential activation, beginning in Wernicke’s area at �250 ms, spreading to anterior temporal sites at �270 ms, to Broca’s area by

We survey the field of magnetoencephalography (MEG) and electroencephalography (EEG) source estimation. These modalities offer the potential for functional brain mapping with temporal resolution in the millisecond range. However, the limited number of spatial measurements and the ill-posedness of the inverse problem present significant limits to our ability to produce accurate spatial maps from these data without imposing major restrictions on the form of the inverse solution. Here we describe approaches to solving the forward problem of computing the mapping from putative inverse solutions into the data space. We then describe the inverse problem in terms of low dimensional solutions, based on the equivalent current dipole (ECD), and high dimensional solutions, in which images of neural activation are constrained to the cerebral cortex. We also address the issue of objective assessment of the relative performance of inverse procedures by the free-response receiver operating characteristic (FROC) curve. We conclude with a discussion of methods for assessing statistical significance of experimental results through use of the bootstrap for determining confidence regions in dipole-fitting methods, and random field (RF) and permutation methods for detecting significant activation in cortically constrained imaging studies.

Distributed linear solutions of the EEG source localization problem are used routinely. Here we describe an approach based on the weighted minimum norm method that imposes constraints using anatomical and physiological information derived from other imaging modalities to regularize the solution. In this approach the hyperparameters controlling the degree of regularization are estimated using restricted maximum likelihood (ReML). EEG data are always contaminated by noise, e.g., exogenous noise and background brain activity. The conditional expectation of the source distribution, given the data, is attained by carefully balancing the minimization of the residuals induced by noise and the improbability of the estimates as determined by their priors.

Abstract not found

Abstract—By using signal processing techniques, an estimate of activity in the brain from the electro- or magneto-encephalogram (EEG or MEG) can be obtained. For a proper analysis, a test is required to indicate whether the model for brain activity fits. A problem in using such tests is that often, not all assumptions are satisfied, like the assumption of the number of shells in an EEG. In such a case, a test on the number of sources (model order) might still be of interest. A detailed analysis is presented of the Wald test for these cases. One of the advantages of the Wald test is that it can be used when not all assumptions are satisfied. Two different, previously suggested, Wald tests in electromagnetic source analysis (EMSA) are examined: a test on source amplitudes and a test on the closeness of source pairs. The Wald test is analytically studied in terms of alternative hypotheses that are close to the null hypothesis (local alternatives). It is shown that the Wald test is asymptotically unbiased, that it has the correct level and power, which makes it appropriate to use in EMSA. An accurate estimate of the Cramér–Rao bound (CRB) is required for the use of the Wald test when not all assumptions are satisfied. The sandwich CRB is used for this purpose. It is defined for nonseparable least squares with constraints required for the Wald test on amplitudes. Simulations with EEG show that when the sensor positions are incorrect, or the number of shells is incorrect, or the conductivity parameter is incorrect, then the CRB and Wald test are still good, with a moderate number of trials. Additionally, the CRB and Wald test appear robust against an incorrect assumption on the noise covariance. A combination of incorrect sensor positions and noise covariance affects the possibility of detecting a source with small amplitude. Index Terms—Approximate model, constrained optimization, Fisher information with constraints, model checking, parameter covariance, separable least squares, source localization. I.

■ Research on the spatio-temporal dynamics of visual object recognition suggests a recurrent, interactive model whereby an initial feedforward sweep through the ventral stream to prefrontal cortex is followed by recurrent interactions. However, critical questions remain regarding the factors that mediate the degree of recurrent interactions necessary for meaningful object recognition. The novel prediction we test here is that recurrent interactivity is driven by increasing semantic integration demands as defined by the complexity of semantic information required by the task and driven by the stimuli. To test this prediction, we recorded magnetoencephalography data while participants named living and nonliving objects during two naming tasks. We found that the spatio-temporal dynamics of neural activity were modulated by the level of semantic integration required. Specifically, source reconstructed time courses and phase synchronization measures showed increased recurrent interactions as a function of semantic integration demands. These findings demonstrate that the cortical dynamics of object processing are modulated by the complexity of semantic information required from the visual input. ■

iteria. Declarative Memory Observations of preserved and impaired memory in patients with amnesia indicate that the recall and recognition of facts and episodes, or declarative memory, is dependent on a particular subset of brain regions and can be disrupted selectively. How can we develop a better understanding of this selectivity? Indeed, one might pose the question: Why is declarative memory different from all other forms of memory? Here are four answers to this question: 1. Because declarative memory has distinct behavioral characteristics. 2. Because declarative memory has distinct subjective characteristics. 3. Because declarative memory has a distinct cognitive structure. 4. Because declarative memory has distinct neural substrates. Memory theorists tend to give one or another of these answers greater emphasis, as discussed further below. In any event, determi

Effective instructional methods are now an important national issue. We reviewed the current research techniques used in cognitive neuroscience and what is currently known about the neurocognition of reading and mathematics. We found that while the neurocognition aspects of reading and mathematics share common processes associated with language, certain aspects of semantics and comprehension are unique to reading and certain aspects of mathematics entail visual-spatial processing not observed during reading. We conclude that although significant advances have been made in the understanding of the underlying neurocognitive process in the last decade more research is needed before the neurosciences can make a direct contribution to instructional practice.