Department of Neurology, School of Medicine, Emory University, Atlanta, Georgia, USA.
Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina, Charleston, South Carolina, USA.
Epilepsia. 2024 Nov;65(11):3362-3375. doi: 10.1111/epi.18128. Epub 2024 Sep 21.
Intracranial EEG can identify epilepsy-related networks in patients with focal epilepsy; however, the association between network organization and post-surgical seizure outcomes remains unclear. Hubness serves as a critical metric to assess network organization by identifying brain regions that are highly influential to other regions. In this study, we tested the hypothesis that favorable post-operative seizure outcomes are associated with the surgical removal of interictal network hubs, measured by the novel metric "Resection-Hub Alignment Degree (RHAD)."
We analyzed Phase II interictal intracranial EEG from 69 patients with epilepsy who were seizure-free (n = 45) and non-seizure-free (n = 24) 1 year post-operatively. Connectivity matrices were constructed from intracranial EEG recordings using imaginary coherence in various frequency bands, and centrality metrics were applied to identify network hubs. The RHAD metric quantified the congruence between hubs and resected/ablated areas. We used a logistic regression model, incorporating other clinical factors, and evaluated the association of this alignment regarding post-surgical seizure outcomes.
There was a significant difference in RHAD in fast gamma (80-200 Hz) interictal network between patients with favorable and unfavorable surgical outcomes (p = .025). This finding remained similar across network definitions (i.e., channel-based or region-based network) and centrality measurements (Eigenvector, Closeness, and PageRank). The alignment between surgically removed areas and other commonly used clinical quantitative measures (seizure-onset zone, irritative zone, high-frequency oscillations zone) did not reveal significant differences in post-operative outcomes. This finding suggests that the hubness measurement may offer better predictive performance and finer-grained network analysis. In addition, the RHAD metric showed explanatory validity both alone (area under the curve [AUC] = .66) and in combination with surgical therapy type (resection vs ablation, AUC = .71).
Our findings underscore the role of network hub surgical removal, measured through the RHAD metric of interictal intracranial EEG high gamma networks, in enhancing our understanding of seizure outcomes in epilepsy surgery.
颅内脑电图可识别局灶性癫痫患者的癫痫相关网络；然而，网络组织与术后癫痫发作结果之间的关联尚不清楚。枢纽作用是通过识别对其他区域具有高度影响力的脑区来评估网络组织的关键指标。在这项研究中，我们通过一种新的度量标准“切除-枢纽对齐度（RHAD）”，检验了这样一个假设，即术后癫痫发作结果良好与切除术中的网络枢纽（通过间歇性颅内 EEG 高伽马网络中的 RHAD 度量来衡量）有关。
我们分析了 69 例癫痫患者的 II 期间歇性颅内 EEG 数据，这些患者在术后 1 年均无癫痫发作（n=45）和有癫痫发作（n=24）。使用虚相干性在不同频带构建颅内 EEG 记录的连通矩阵，并应用中心性度量来识别网络枢纽。RHAD 度量量化了枢纽与切除/消融区域之间的一致性。我们使用逻辑回归模型，结合其他临床因素，评估了这种对准度与术后癫痫发作结果的关联。
在快伽马（80-200Hz）间歇性网络中，有良好手术结果和不良手术结果的患者之间的 RHAD 存在显著差异（p=0.025）。这一发现在网络定义（即基于通道或基于区域的网络）和中心性测量（特征向量、接近度和 PageRank）上均相似。手术切除区域与其他常用的临床定量测量（发作起始区、刺激性区、高频振荡区）之间的对准度在术后结果方面没有显示出显著差异。这表明枢纽测量可能提供更好的预测性能和更精细的网络分析。此外，RHAD 度量单独（曲线下面积[AUC] = 0.66）和与手术治疗类型（切除与消融，AUC = 0.71）结合使用均具有解释有效性。
我们的发现强调了网络枢纽切除的作用，通过间歇性颅内 EEG 高伽马网络的 RHAD 度量来衡量，这有助于我们理解癫痫手术中的癫痫发作结果。
