Division of Neurosurgery, Department of Surgery, Hospital for Sick Children, Toronto, Ontario, Canada;
J Neurophysiol. 2013 Nov;110(10):2475-83. doi: 10.1152/jn.00034.2013. Epub 2013 Sep 4.
Synchronization of neural oscillations is thought to integrate distributed neural populations into functional cell assemblies. Epilepsy is widely regarded as a disorder of neural synchrony. Knowledge is scant, however, regarding whether ictal changes in synchrony involving epileptogenic cortex are expressed similarly across various frequency ranges. Cortical regions involved in epileptic networks also exhibit pathological high-frequency oscillations (pHFOs, >80 Hz), which are increasingly utilized as biomarkers of epileptogenic tissue. It is uncertain how pHFO amplitudes are related to epileptic network connectivity. By calculating phase-locking values among intracranial electrodes implanted in children with intractable epilepsy, we constructed ictal connectivity networks and performed graph theoretical analysis to characterize their network properties at distinct frequency bands. Ictal data from 17 children were analyzed with a hierarchical mixed-effects model adjusting for patient-level covariates. Epileptogenic cortex was defined in two ways: 1) a hypothesis-driven method using the visually defined seizure-onset zone and 2) a data-agnostic method using the high-frequency amplitude of each electrode. Epileptogenic cortex exhibited a logarithmic decrease in interregional functional connectivity at high frequencies (>30 Hz) during seizure initiation and propagation but not at termination. At slower frequencies, conversely, epileptogenic cortex expressed a relative increase in functional connectivity. Our findings suggest that pHFOs reflect epileptogenic network interactions, yielding theoretical support for their utility in the presurgical evaluation of intractable epilepsy. The view that abnormal network synchronization plays a critical role in ictogenesis and seizure dynamics is supported by the observation that functional isolation of epileptogenic cortex at high frequencies is absent at seizure termination.
神经振荡的同步被认为是将分布式神经群体整合到功能细胞集合中。癫痫被广泛认为是一种神经同步障碍。然而,关于涉及致痫皮质的同步的发作期变化是否在各种频率范围内表现相似,我们知之甚少。参与癫痫网络的皮质区域也表现出病理性高频振荡(pHFO,>80 Hz),越来越多地被用作致痫组织的生物标志物。尚不清楚 pHFO 幅度与癫痫网络连通性有何关系。通过计算植入难治性癫痫儿童颅内电极之间的锁相值,我们构建了发作连接网络,并进行了图论分析,以在不同频率带中表征其网络特性。使用分层混合效应模型对 17 名儿童的发作数据进行了分析,该模型调整了患者水平的协变量。致痫皮质通过以下两种方式定义:1)使用视觉定义的发作起始区的假设驱动方法,2)使用每个电极的高频振幅的无数据方法。在发作起始和传播期间,高频(>30 Hz)时,致痫皮质的区域间功能连通性呈对数下降,但在终止时则没有。相反,在较慢的频率下,致痫皮质表达了功能连通性的相对增加。我们的研究结果表明,pHFO 反映了致痫网络的相互作用,为其在难治性癫痫的术前评估中的应用提供了理论支持。异常网络同步在发作发生和发作动力学中起着关键作用的观点得到了支持,因为在发作终止时,高频时致痫皮质的功能隔离不存在。