Characterizing epileptic spikes as travelling waves using high-density EEG

科学摘要 全世界有6000多万人患有癫痫。据估计，30%到40%的癫痫患者 患有难治性癫痫。对于一些内科治疗无效的患者，可以提供手术治疗。而当 更好的癫痫自由率通常是通过综合切除手术实现的，它可能是 根据切除面积的不同，与术后缺陷相关。术前评估经常 涉及头皮脑电，在某些情况下，通过颅内监测技术可以帮助定位癫痫发作 起始区。然而，这些程序往往是非常有侵入性的，并有可能发生不良事件。高 密度脑电提供了一种潜在的非侵入性替代方法，可以准确地定位癫痫的发作区 病人。到目前为止，大多数高密度脑电研究都集中在头皮棘波功率的来源重建[1]。 并且没有利用EEG信号提供的高时间分辨率来显式地研究棘波 起源和传播动力学。在这里，我们建议使用我们团队以前开发的算法来 评估离散脑电事件的皮质传播，如睡眠慢波和纺锤波[2]，以 用高密度脑电表征癫痫棘波的起源和传播模式 录音。我们计划表征癫痫棘波起源和传播模式的重复性 在头皮水平和源空间，并将使用高密度脑电获得的结果与 使用在同一患者中进行的颅内记录获得。我们将通过以下方式补充这项研究 使用格兰杰因果关系和动态因果模型的方向性连通性评估。如果成功，这将是 该方法将为指导药物难治性癫痫的选择性切除手术提供有用的信息 患者，潜在地规避了侵入性研究的需要。 第1条

Scientific abstract Epilepsy affects over 60 millions of individuals worldwide. It is estimated that 30 to 40% of epilepsy patients have refractory epilepsy. Surgery may be offered to some patients who are refractory to medical therapy. While better seizure freedom rates are commonly achieved with comprehensive resective surgery, it may be associated with postoperative deficits depending on the area of resection. Pre-surgical evaluation often involves scalp EEG followed in some cases by intracranial monitoring techniques can help localizing seizure onset zone. These procedures are however are often very invasive and have potential for adverse events. High density EEG offers a potential non-invasive alternative to accurately localize epileptic onset zone in individual patients. To date, most high-density EEG studies focused on source reconstruction of scalp spike power [1] and did not take advantage of the high temporal resolution offered by EEG signals to explicitly study spike origin and propagation dynamics. Here we propose to use algorithms previously developed by our team to assess the cortical propagation of discrete EEG events such as sleep slow waves and spindles [2] to characterize the origins and propagation patterns of epileptic spikes captured using high-density EEG recordings. We plan to characterize reproducibility of origins and propagation patterns of epileptic spikes both at the scalp level and in source space, and compare the results obtained using high-density EEG to the ones obtained using intracranial recordings performed in the same patients. We will complement this study by directional connectivity assessments using Granger causality and Dynamic Causal Modeling. If successful, this approach will provide useful information to guide selective resection surgery in drug-refractory epileptic patients, potentially circumventing the need for invasive studies.   1