Integrating EEG/MEG and fMRI

我们早期的基础研究表明，在认知任务期间，MEG记录能够以与功能性MRI相当的方式定位大脑活动。具体来说，大脑皮层β频段功率的降低大体上与BOLD激活结果一致。然而，像MEG/EEG这样的电生理记录具有其他成像技术无法实现的细粒度时间动态。我们还发现，前扣带皮层的GABA （γ -氨基丁酸）浓度与静息时MEG β带功率的空间定位相关。我们扩大了匹配良好的受试者队列，并隔离了特定频带差异，特别是在高伽马范围（65至115 Hz）。

Our earlier basic studies have shown that MEG recordings during cognitive tasks have the ability to localize brain activity in comparable fashion to functional MRI. In specific, reduction in power in the beta frequency band at the cortex in general agrees with BOLD activation results. However, electrophysiological recordings such as the MEG/EEG have fine grained time dynamics not possible with other imaging techniques. We have also found that GABA (gamma-aminobutyric acid) concentration in the anterior cingulate cortex correlates with spatially localized resting MEG beta band power. We have expanded the cohort of well-matched subjects and isolated specific frequency band difference especially in the high gamma range (65 to 115 Hz). Differences in the degree of activation especially in frontal regions as indexed by beta desynchronization during a working memory task have been found between patients with schizophrenia compared to well siblings and healthy control volunteers. Previously we have seen that this activation reveals an interaction with genotype for the well-studied COMT marker. With groups of patients, siblings and controls matched for working memory task performance, patients show a distinct reduced DLPFC activation in apparent distinction to an increase in BOLD relative to task load. The MEG analysis isolates a working memory component that may reflect a different aspect of cortical processing. We have extended these results to examine differences in the high gamma band across groups. The relation of the different frequency bands to patterns of blood flow activation have revealed more specific targets for the neurophysiology underlying patient differences. We are now also comparing these differences in patients on and off anti-psychotic medication. It appears that medication may attenuate the PFC differences. We are also extending our genetic comparisons to examine modulation by SCN2A, a gene encoding the alpha-2 subunit of the sodium channel. Other studies have shown variation in cognitive performance across group, with MEG source localization showing significant differences in DLPFC and dACC for genotype by task interactions. Differences in network patterns and dynamics are key to understanding underlying pathology in clinical groups. Previously Bassett et al. have shown that functional network variations in patient groups can be related to behavioral outcomes on cognitive activities. We have found distinct patterns of the temporal sequence of brain regions involved in these memory tasks that show a variety of individual differences across subjects. We are continuing to utilize a variety of new measures to examine the flow of information in relation to cognitive task demand. We previously demonstrated that critical dynamics exist in resting MEG recordings and are now expanding this to compare patients with schizophrenia.