Iannotti Giannina R, Grouiller Frédéric, Centeno Maria, Carmichael David W, Abela Eugenio, Wiest Roland, Korff Christian, Seeck Margitta, Michel Christoph, Pittau Francesca, Vulliemoz Serge
Functional Brain Mapping Lab, Faculty of Medicine, University of Geneva, Geneva, Switzerland.
Department of Radiology and Medical Informatics, University Hospital of Geneva, Geneva, Switzerland.
Epilepsia. 2016 Jul;57(7):1086-96. doi: 10.1111/epi.13400. Epub 2016 May 6.
Epilepsy is increasingly considered as the dysfunction of a pathologic neuronal network (epileptic network) rather than a single focal source. We aimed to assess the interactions between the regions that comprise the epileptic network and to investigate their dependence on the occurrence of interictal epileptiform discharges (IEDs).
We analyzed resting state simultaneous electroencephalography-functional magnetic resonance imaging (EEG-fMRI) recordings in 10 patients with drug-resistant focal epilepsy with multifocal IED-related blood oxygen level-dependent (BOLD) responses and a maximum t-value in the IED field. We computed functional connectivity (FC) maps of the epileptic network using two types of seed: (1) a 10-mm diameter sphere centered in the global maximum of IED-related BOLD map, and (2) the independent component with highest correlation to the IED-related BOLD map, named epileptic component. For both approaches, we compared FC maps before and after regressing out the effect of IEDs in terms of maximum and mean t-values and percentage of map overlap.
Maximum and mean FC maps t-values were significantly lower after regressing out IEDs at the group level (p < 0.01). Overlap extent was 85% ± 12% and 87% ± 12% when the seed was the 10-mm diameter sphere and the epileptic component, respectively.
Regions involved in a specific epileptic network show coherent BOLD fluctuations independent of scalp EEG IEDs. FC topography and strength is largely preserved by removing the IED effect. This could represent a signature of a sustained pathologic network with contribution from epileptic activity invisible to the scalp EEG.
癫痫越来越被视为一种病理性神经网络(癫痫网络)的功能障碍,而非单一的病灶源。我们旨在评估构成癫痫网络的各区域之间的相互作用,并研究它们对发作间期癫痫样放电(IEDs)发生的依赖性。
我们分析了10例耐药性局灶性癫痫患者的静息态同步脑电图-功能磁共振成像(EEG-fMRI)记录,这些患者具有多灶性IED相关的血氧水平依赖(BOLD)反应,且在IED区域有最大t值。我们使用两种类型的种子计算癫痫网络的功能连接(FC)图:(1)以IED相关BOLD图的全局最大值为中心的直径10毫米的球体;(2)与IED相关BOLD图相关性最高的独立成分,称为癫痫成分。对于这两种方法,我们在去除IEDs的影响前后,根据最大和平均t值以及图重叠百分比比较FC图。
在组水平上去除IEDs后,最大和平均FC图t值显著降低(p < 0.01)。当种子为直径10毫米的球体和癫痫成分时,重叠程度分别为85%±12%和87%±12%。
参与特定癫痫网络的区域显示出与头皮脑电图IEDs无关的连贯BOLD波动。通过去除IED效应,FC拓扑结构和强度在很大程度上得以保留。这可能代表了一个持续的病理性网络的特征,其中癫痫活动对头皮脑电图不可见但有贡献。