Effective connectivity relates seizure outcome to electrode placement in responsive neurostimulation 有效连接将癫痫发作结果与反应性神经刺激中的电极位置联系起来
(1)Katsuya Kobayashi,' Kenneth N. Taylor,' (C)Hossein Shahabi, (1)Balu Krishnan,' (1)Katsuya Kobayashi,' Kenneth N. Taylor,' (C)Hossein Shahabi, (1)Balu Krishnan,'Anand Joshi, Michael J. Mackow, ' Lauren Feldman, ' Omar Zamzam, Takfarinas Medani, Anand Joshi, Michael J. Mackow, ' Lauren Feldman, ' Omar Zamzam, Takfarinas Medani、 Juan Bulacio,' (1)Andreas V. Alexopoulos,' Imad Najm,' (1)William Bingaman,' Juan Bulacio、' (1)Andreas V. Alexopoulos、' Imad Najm、' (1)William BingamanRichard M. Leahy and (1)Dileep R. Nair' Richard M. Leahy 和 (1)Dileep R. Nair'
Abstract 摘要
Responsive neurostimulation is a closed-loop neuromodulation therapy for drug resistant focal epilepsy. Responsive neurostimulation electrodes are placed near ictal onset zones so as to enable detection of epileptiform activity and deliver electrical stimulation. There is no standard approach for determining the optimal placement of responsive neurostimulation electrodes. Clinicians make this determination based on presurgical tests, such as MRI, EEG, magnetoencephalography, ictal single-photon emission computed tomography and intracranial EEG. Currently functional connectivity measures are not being used in determining the placement of responsive neurostimulation electrodes. Cortico-cortical evoked potentials are a measure of effective functional connectivity. Cortico-cortical evoked potentials are generated by direct single-pulse electrical stimulation and can be used to investigate cortico-cortical connections in vivo. We hypothesized that the presence of high amplitude cortico-cortical evoked potentials, recorded during intracranial EEG monitoring, near the eventual responsive neurostimulation contact sites is predictive of better outcomes from its therapy. We retrospectively reviewed 12 patients in whom cortico-cortical evoked potentials were obtained during stereoelectroencephalography evaluation and subsequently underwent responsive neurostimulation therapy. We studied the relationship between cortico-cortical evoked potentials, the eventual responsive neurostimulation electrode locations and seizure reduction. Directional connectivity indicated by corticocortical evoked potentials can categorize stereoelectroencephalography electrodes as either receiver nodes/in-degree (an area of greater inward connectivity) or projection nodes/out-degree (greater outward connectivity). The follow-up period for seizure reduction ranged from 1.3-4.8 years (median 2.7) after responsive neurostimulation therapy started. Stereoelectroencephalography electrodes closest to the eventual responsive neurostimulation contact site tended to show larger in-degree cortico-cortical evoked potentials, especially for the early latency cortico-cortical evoked potentials period (10-60 ms period) in six out of 12 patients. Stereoelectroencephalography electrodes closest to the responsive neurostimulation contacts also had greater significant out-degree in the early corticocortical evoked potentials latency period than those further away . Additionally, significant correlation was noted between in-degree cortico-cortical evoked potentials and greater seizure reduction with responsive neurostimulation therapy at its most effective period . These findings suggest that functional connectivity determined by cortico-cortical evoked potentials may provide additional information that could help guide the optimal placement of responsive neurostimulation electrodes. 反应性神经刺激是一种治疗耐药性局灶性癫痫的闭环神经调节疗法。反应性神经刺激电极放置在癫痫发作区附近，以便检测癫痫样活动并进行电刺激。目前还没有标准方法来确定反应性神经刺激电极的最佳位置。临床医生根据手术前的检查，如核磁共振成像、脑电图、脑磁图、发作期单光子发射计算机断层扫描和颅内脑电图等，来确定最佳位置。目前，在确定神经刺激电极的位置时并未使用功能连接测量。皮层诱发电位是一种有效的功能连接测量方法。皮层诱发电位由直接单脉冲电刺激产生，可用于研究体内皮层与皮层的连接。我们假设，在颅内脑电图监测期间记录到的高振幅皮质-皮质诱发电位出现在最终有反应的神经刺激接触点附近，可预测其治疗效果。我们对在立体脑电图评估中获得皮质-皮质诱发电位并随后接受反应性神经刺激治疗的 12 名患者进行了回顾性研究。我们研究了皮质-皮质诱发电位、最终反应性神经刺激电极位置和癫痫发作减少之间的关系。皮质诱发电位显示的定向连通性可将立体脑电图电极划分为接收节点/内度（内向连通性较强的区域）或投射节点/外度（外向连通性较强的区域）。响应性神经刺激治疗开始后，癫痫发作减少的随访期为 1.3-4.8 年（中位数为 2.7 年）。最靠近最终反应性神经刺激接触点的立体脑电图电极往往显示出更大的皮质-皮质诱发电位内度，尤其是在皮质-皮质诱发电位早期潜伏期（10-60 毫秒），12 名患者中有 6 人显示出更大的皮质-皮质诱发电位内度。在皮质诱发电位早期潜伏期，距离神经刺激反应触点最近的立体脑电图电极 也比距离较远的电极有更大的显著偏离度 。此外，在反应性神经刺激疗法的最有效期内，皮质皮层诱发电位的内度与癫痫发作的减少程度之间存在明显的相关性 。 这些研究结果表明，由皮层-皮层诱发电位确定的功能连通性可提供更多信息，有助于指导神经刺激电极的最佳位置。
1 Charles Shor Epilepsy Center, Cleveland Clinic Foundation, Cleveland, OH 44195, USA 1 美国俄亥俄州克利夫兰 44195 克利夫兰诊所基金会查尔斯-肖尔癫痫中心
2 Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA 90007, USA 2 Ming Hsieh 美国加利福尼亚州洛杉矶市 90007 南加州大学电子与计算机工程系
Correspondence to: Dileep R. Nair, MD 通讯作者：Dileep R. Nair，医学博士Dileep R. Nair，医学博士
Charles Shor Epilepsy Center, Cleveland Clinic Foundation 查尔斯-肖癫痫中心，克利夫兰诊所基金会
Approximately of patients with epilepsy are drug resistant. Although epilepsy surgery is the most effective therapy for drug resistant focal epilepsy, not all are candidates for surgery. Patients with poorly localized epilepsy, multifocal epilepsy or epileptic foci that overlap with eloquent cortex are usually excluded from surgical resection. In addition, only of patients who undergo a resection attain seizure freedom. 约有 的癫痫患者具有耐药性。 虽然癫痫手术是治疗耐药性局灶性癫痫最有效的方法，但 ，并非所有患者都适合手术治疗。定位不清的癫痫患者、多灶性癫痫患者或癫痫灶与大脑皮质重叠的患者通常不能接受手术切除。 此外，只有 接受切除手术的患者能够摆脱癫痫发作。
Over the last two decades, neuromodulation therapies have been introduced for the treatment of epilepsy, including vagus nerve stimulation (VNS), responsive neurostimulation (RNS) and deep brain stimulation (DBS). RNS is a closed-loop neuromodulation therapy that delivers electrical stimuli directly to the ictal onset zone (IOZ) when epileptiform activity is detected. The RNS System has been proven safe and effective in drug resistant focal epilepsy. However, unlike VNS or DBS, RNS electrode placement is tailored to each patient's IOZ. Placement of RNS leads can be guided by several factors including MRI, EEG, magnetoencephalography (MEG), and ictal singlephoton emission computed tomography as well as intracranial EEG. RNS therapy can be directed to the IOZ, near the IOZ or target-relevant propagation networks. Functional connectivity (FC) measures are increasingly being studied to investigate epileptogenic brain networks. Some recent studies have introduced related strategies to identify biomarkers to prognosticate outcomes with RNS therapy, including network synchronizability (measured 在过去二十年里，神经调控疗法被引入癫痫治疗，包括迷走神经刺激（VNS）、 responsive neurostimulation (RNS) 和脑深部刺激（DBS）。 RNS 是一种闭环神经调控疗法，可在检测到癫痫样活动时直接向发作起始区 (IOZ) 输送电刺激。事实证明，RNS 系统对耐药性局灶性癫痫安全有效。 然而，与 VNS 或 DBS 不同的是，RNS 电极的放置是根据每位患者的 IOZ 量身定制的。RNS 导联的放置可由多种因素引导，包括核磁共振成像、脑电图、脑磁图（MEG）、发作期单光子发射计算机断层扫描以及颅内脑电图。RNS 治疗可针对 IOZ、IOZ 附近或目标相关传播网络。 功能连通性（FC）测量正越来越多地被用于研究致痫性大脑网络。 最近的一些研究引入了相关策略，以确定生物标志物来预示 RNS 治疗的结果，其中包括网络同步性（测量结果为
via intracranial EEG), functional connectivity (measured via MEG), structural connectivity (measured via tractography and the ability of brain networks to functionally reorganize (measured via chronic intracranial EEG). Khambhati et al. reported that the connectivity of interictal spikes may also play a role in predicting outcome of RNS therapy, suggesting that the mechanism for RNS involves network plasticity. However, FC measures using intracranial evoked responses are not currently being used to determine RNS lead location. 通过颅内脑电图测量）、 功能连通性（通过 MEG 测量）、 结构连通性（通过 tractography 测量） 以及大脑网络的功能重组能力（通过慢性颅内脑电图测量）。 Khambhati 等人 报告称，发作间期尖峰的连通性也可能在预测 RNS 治疗结果方面发挥作用，这表明 RNS 的机制涉及网络可塑性。然而，目前还没有使用颅内诱发反应的 FC 测量来确定 RNS 导联位置。
Several studies have shown disruption in connectivity and changes in network topology in patients with epilepsy. More recently, investigators suggested that FC, based on resting state functional MRI (rs-fMRI), could be used as a tool to impact decision-making for epilepsy surgery. analysis, assessed by rs-fMRI pre-treatment connectivity, was shown to predict response to repetitive transcranial magnetic stimulation (rTMS) in patients with drug resistant depression. is defined as a statistical dependency in neurophysiologic measurements between spatially remote areas. Effective functional connectivity ascribes causal relationships between nodes in a network using models to add weighted directionality. Intracranial electrical cortical stimulation induced evoked potentials can provide a direct method to evaluate effective connectivity in vivo in patients with epilepsy. 有几项研究显示，癫痫患者的连通性中断和网络拓扑结构发生了变化。 最近，研究人员提出，基于静息状态功能磁共振成像（rs-fMRI）的功能连通性可用作影响癫痫手术决策的工具。 分析表明，通过 rs-fMRI 治疗前连通性评估，可预测耐药抑郁症患者对重复经颅磁刺激（rTMS）的反应。 定义为空间遥远区域之间神经生理测量的统计依赖性。 有效的功能连通性通过模型添加加权方向性来描述网络中节点之间的因果关系。颅内皮质电刺激诱发电位可提供一种直接方法来评估癫痫患者体内的有效连接性。
Recordings of local evoked potentials induced by direct electrical cortical stimulation, termed 'direct cortical responses,' were first performed by who found a local surface negative potential evoked by electrical stimulation of the cortical surface in various species. Single-pulse electrical stimulation (SPES) induces evoked potentials, termed cortico-cortical evoked potentials (CCEPs), which can be used to trace corticocortical connections in vivo CCEPs have been extensively employed to evaluate the cortico-cortical networks associated with various normal brain functions and to evaluate cortical excitability and connectivity associated with areas of epileptogenicity. Keller et al. reported that both indegree (the total number of times stimulation of any region evokes a significant CCEP at the region of interest) and outdegree (the total number of significant CCEPs observed when the region of interest is stimulated) CCEPs were higher in the seizure onset zone (SOZ) than outside SOZ, suggesting that the more epileptogenic area has larger and stronger corticocortical connectivity. We hypothesized that the area with larger in-degree and out-degree CCEPs could be a hub of the brain and could be a good target for RNS therapy via association fibre through which CCEPs travel. We tested this hypothesis by analysing the amplitude of intracranially recorded in-degree and out-degree CCEPs and examined its correlation with RNS outcomes. In this study, we asked the question, could CCEPs be used to optimize targets for RNS neuromodulation? In the analysis of CCEPs, the out-degree of a region represents the number of significant CCEPs elicited following stimulation of the site of interest, while the in-degree refers to the total number of significant CCEPs elicited at the site of interest upon stimulation of all other sites. These out-degree and in-degree 首先记录了皮层直接电刺激诱发的局部诱发电位，称为 "皮层直接反应"，他发现在不同物种中，皮层表面电刺激会诱发局部表面负电位。单脉冲电刺激（SPES）可诱发诱发电位，称为皮质-皮质诱发电位（CCEPs），可用于追踪体内皮质连接 ，CCEPs 已被广泛用于评估与各种正常脑功能相关的皮质-皮质网络 ，以及评估与致痫区相关的皮质兴奋性和连接性。 Keller 等人 报告说，发作起始区（SOZ）的内度（刺激任何区域引起相关区域显著 CCEP 的总次数）和外度（刺激相关区域时观察到的显著 CCEP 的总次数）CCEP 均高于 SOZ 以外的区域，这表明致痫性更强的区域具有更大和更强的皮质连接性。我们假设，CCEPs 内度和外度较大的区域可能是大脑的一个枢纽，并可能通过 CCEPs 穿过的关联纤维成为 RNS 治疗的良好靶点。我们通过分析颅内记录的度内和度外 CCEPs 的振幅来验证这一假设，并研究其与 RNS 治疗结果的相关性。在这项研究中，我们提出了这样一个问题：CCEPs 可以用来优化 RNS 神经调控的目标吗？在 CCEPs 分析中，一个区域的 "外度 "代表刺激相关部位后引起的显著 CCEPs 的数量，而 "内度 "指的是刺激所有其他部位后在相关部位引起的显著 CCEPs 的总数。 这些外度和内度
CCEPs reflect the directional flow of information in the brain. Since seizures propagate through cortico-cortical networks, CCEPs is a unique approach to study directional connectivity of epileptogenic networks. As in-degree CCEPs of a given cortical site reflect an area of hypersynchrony, and out-degree CCEPs can be a measure of ictal propagation, we were interested in analysing both measures for the optimal placement of RNS electrodes. We reviewed the CCEPs performed in patients during stereoelectroencephalography (SEEG) evaluation who later went on to have RNS therapy. We investigated the correlations between CCEPs during SEEG and prior to the RNS placement, the eventual RNS contact location and outcomes from RNS therapy. In this study, we retrospectively investigated the degree of connectivity, as measured by CCEPs, at the point of the RNS contact placement in two scenarios: (i) the distances from the SEEG recording sites of CCEPs to the eventual closest RNS contacts for in-degree CCEPs; and (ii) distance from the SEEG stimulus sites for CCEPs to the eventual closest RNS contacts for out-degree CCEPs. Our primary hypothesis was that there exists a greater degree of both in-degree and out-degree connectivity at the point of RNS contact in patients who benefit from RNS therapy. Both in-degree and out-degree CCEPs have been reported higher in the SOZ than outside SOZ, suggesting that the more epileptogenic area has larger and stronger cortico-cortical connectivity. In-degree CCEPs may reflect the degree of hypersynchronous brain activity. Those brain regions associated with high indegree CCEPs could be influenced by neuronal or electrical activity from distant sources. This could be one explanation for why neuromodulation therapy with a wide variety of stimulation targets can still modulate the epileptic network. In addition, we have reported that by stimulating the ictal onset zone, significant out-degree CCEPs were observed within the epileptic network. Such a finding could support the notion that FC should be considered during RNS contact placement with the goal of modulating the epileptic network as a whole. CCEPs 反映了大脑中信息流动的方向。由于癫痫发作是通过皮层-皮层网络传播的，CCEPs 是研究致痫网络定向连接性的一种独特方法。由于特定皮层部位的内度 CCEPs 反映了一个超同步区域，而外度 CCEPs 则可作为癫痫发作传播的衡量标准，因此我们有兴趣分析这两种衡量标准，以确定 RNS 电极的最佳位置。我们回顾了在立体脑电图（SEEG）评估期间对后来接受 RNS 治疗的患者进行的 CCEP。我们研究了 SEEG 期间和 RNS 放置之前的 CCEPs、最终 RNS 接触位置和 RNS 治疗结果之间的相关性。在这项研究中，我们回顾性地调查了两种情况下通过 CCEPs 测定的 RNS 接触点位置的连接程度：(i) 度内 CCEPs 的 SEEG 记录点到最终最近 RNS 接触点的距离；(ii) 度外 CCEPs 的 SEEG 刺激点到最终最近 RNS 接触点的距离。我们的主要假设是，从 RNS 治疗中获益的患者在 RNS 接触点处存在更大程度的 "度内 "和 "度外 "连接。有报告称，SOZ 内的度内和度外 CCEPs 均高于 SOZ 外，这表明致痫区具有更大、更强的皮质-皮质连通性。 度内 CCEPs 可反映大脑活动的不同步程度。那些与高度内CCEPs相关的脑区可能受到来自远处的神经元或电活动的影响。这也可以解释为什么神经调控疗法使用多种刺激靶点仍能调节癫痫网络。此外，我们还报道了通过刺激癫痫发作区，在癫痫网络内观察到了显著的外度 CCEP。 这一发现支持了这样一种观点，即在放置 RNS 接触点时应考虑 FC，以达到调节整个癫痫网络的目的。
Materials and methods 材料和方法
Patients 患者
We studied 12 patients (six females) with drug resistant focal epilepsy who underwent CCEPs recordings during their presurgical SEEG evaluation and subsequently underwent RNS therapy (NeuroPace, Inc., Mountain View, CA). The median age was 26 years (range: 18-60) at time of RNS implantation, median age of epilepsy onset was 13 years (range: ) (Supplementary Table 1). Prior to their SEEG evaluation, patients underwent non-invasive testing including: scalp video-EEG, MEG, brain MRI, -fluorodeoxyglucose positron emission tomography (PET), ictal single-photon emission computerized tomography (SPECT) and neuropsychological examinations. All patients were discussed in the comprehensive patient management conferences at our centre before and after their SEEG evaluation that resulted 我们研究了 12 名耐药局灶性癫痫患者（6 名女性），他们在手术前 SEEG 评估期间接受了 CCEPs 记录，随后接受了 RNS 治疗（NeuroPace, Inc.）植入 RNS 时的中位年龄为 26 岁（范围：18-60 岁），癫痫发病的中位年龄为 13 岁（范围： ）（补充表 1）。在进行 SEEG 评估之前，患者接受了非侵入性检查，包括：头皮视频脑电图（EEG）、脑电图（MEG）、脑磁共振成像（MRI）、 -氟脱氧葡萄糖正电子发射断层扫描（PET）、发作期单光子发射计算机断层扫描（SPECT）和神经心理学检查。所有患者在接受 SEEG 评估前后都在本中心的患者综合管理会议上进行了讨论，结果如下
in a decision to offer RNS therapy. Discussion at our post-SEEG patient management conference included suggestions as to where the RNS electrodes should be placed for each patient and was based solely on the ictal patterns recorded during SEEG. The planning of the location of the RNS lead was informed by IOZ SEEG location. The implanted RNS System includes a programmable neurostimulator connected to 1-4 depth electrodes or subdural strip leads with each lead containing four electrodes. Figure 1 shows a flowchart of the steps involved in the analysis. The use of RNS depth electrodes versus subdural strips was assessed individually based on location of the IOZ during patient management conference. This retrospective study protocol was approved by the Cleveland Clinic institutional review board. 决定是否提供 RNS 治疗。我们在 SEEG 后患者管理会议上讨论的内容包括建议为每位患者放置 RNS 电极的位置，这些建议完全基于 SEEG 期间记录的发作模式。RNS 导联位置的规划参考了 IOZ SEEG 位置。植入式 RNS 系统包括一个与 1-4 个深度电极或硬膜下带状导线相连的可编程神经刺激器，每个导线包含四个电极。图 1 显示了分析步骤的流程图。在患者管理会议上，根据 IOZ 的位置对使用 RNS 深度电极还是硬膜下导联线进行了单独评估。这项回顾性研究方案已获得克利夫兰诊所机构审查委员会的批准。
CCEPs recording during the SEEG evaluation 在 SEEG 评估期间记录的 CCEP
Among the 12 patients, nine underwent bilateral and three unilateral SEEG electrode implantation (two left hemisphere, one right hemisphere). The median number of implanted SEEG electrodes was 13 (range: 7-16). The median number of pairs of stimulus sites for CCEPs was 11.5 (range 3-45) and the median number of SEEG contacts implanted used for CCEPs recording was 150 (range: 65-185). The median number of contacts on middle temporal gyrus (MTG) for out-degree CCEPs analysis was 13.5 (range: 2-20) and that on superior temporal gyrus (STG) was 11 (range: 226). The configurations of SEEG electrodes and RNS electrodes are shown in Talairach space and on a presurgical 3D-MRI in Fig. 2. The 12 patients were classified as follows: three mesial temporal lobe epilepsy (mesial TLE), seven neocortical TLE, one temporo-parietal lobe epilepsy (T-PLE) and one occipital lobe epilepsy (OLE) according to the results of SEEG evaluation (Supplementary Table 1). Only one patient (Patient 4) underwent a resective epilepsy surgery of the right posterior basal temporal region before RNS therapy. 12 名患者中，9 人接受了双侧 SEEG 电极植入术，3 人接受了单侧 SEEG 电极植入术（2 人左半球，1 人右半球）。植入 SEEG 电极的中位数为 13 个（范围：7-16）。用于 CCEPs 的刺激点对数中位数为 11.5 对（范围：3-45），用于 CCEPs 记录的 SEEG 触点植入数量中位数为 150 个（范围：65-185）。用于度外 CCEPs 分析的颞中回（MTG）触点数量中位数为 13.5 个（范围：2-20），颞上回（STG）触点数量中位数为 11 个（范围：226）。图 2 显示了 SEEG 电极和 RNS 电极在 Talairach 空间和术前 3D-MRI 上的配置。根据 SEEG 评估结果（补充表 1），12 名患者被分类如下：3 名颞叶中叶癫痫（中叶 TLE）、7 名新皮层 TLE、1 名颞顶叶癫痫（T-PLE）和 1 名枕叶癫痫（OLE）。只有一名患者（患者 4）在接受 RNS 治疗前接受了右侧后基底颞区癫痫切除手术。
Depth electrodes for SEEG evaluation were made of platinum (AdTech, Integra or PMT) and implanted using the Talairach stereotactic method based on the results of presurgical non-invasive evaluation in each patient. 用于 SEEG 评估的深度电极由铂金（AdTech、Integra 或 PMT）制成，根据每位患者的术前无创评估结果，采用 Talairach 立体定向法植入。
The methodology of CCEPs evaluation has been previously reported. In brief, CCEPs were recorded towards the end of the SEEG evaluation when patients are restarted on their antiseizure medications. Using a constant-current stimulation device (Grass S88, Astro-Med, Inc., RI, USA), SPES at consisting of square wave pulses was repetitively applied at with alternating polarity through a pair of adjacent SEEG contacts in the cortices. Trials of 60 SPESs were delivered for each stimulus site (two trials of 30 SPESs). Reponses were recorded at a sampling rate of (EEG-1200, Nihon Kohden, Tokyo, Japan) and bandpass filtered at . The reference electrode was placed on the skin at the vertex region. CCEPs 评估方法此前已有报道。 简而言之，CCEPs 是在 SEEG 评估即将结束、患者重新开始服用抗癫痫药物时记录的。使用恒流刺激装置（Grass S88，Astro-Med, Inc., RI, USA），在 重复施加由 方波脉冲组成的 SPES， ，极性交替地通过大脑皮层中一对相邻的 SEEG 触点。对每个刺激点进行 60 次 SPES 试验（两次试验 30 个 SPES）。反应以 的采样率记录（EEG-1200，Nihon Kohden，日本东京），并以 进行带通滤波。参比电极置于顶点区域的皮肤上。
Identification of locations of RNS contacts and the correlation with SEEG contacts 确定 RNS 接触点的位置以及与 SEEG 接触点的相关性
The locations of the contacts relative to brain anatomy were determined by co-registering the preoperative MRI of the brain with the head CT scan for SEEG electrode locations and CT scan for RNS electrode locations. We have recently developed a semi-automated identification process of anatomical labelling for intracranial electrode contacts, details of which are described in Tayler et al. In brief, the preoperative MRI of each patient was imported to BrainSuite and an anatomical segmentation was performed based on the USCBrain atlas. The SEEG contacts locations are determined from the CT using the Curry software (Compumedics, NeuroScan Laboratories, Charlotte, NC, USA), and the outputs are combined to localize and automatically assign anatomical labels to each contact. The contact locations of RNS were identified in the same manner. The distances between each SEEG and each RNS contact were calculated using an in-house MATLAB script (the Mathworks, Inc., Natick, MA, USA). BrainSuite and BrainStorm software were used to confirm the locations of the contacts of SEEG and RNS electrodes within the brain volumes. 通过术前脑部核磁共振成像与头部 CT 扫描（SEEG 电极位置）和 CT 扫描（RNS 电极位置）的共同对比，确定接触点与脑部解剖结构的相对位置。我们最近开发了一种半自动化的颅内电极触点解剖标记识别流程，详情请见 Tayler 等人的文章 。简而言之，每位患者的术前 MRI 都被导入 BrainSuite ，并根据 USCBrain 图集进行解剖学分割。 使用 Curry 软件（Compumedics，NeuroScan Laboratories，Charlotte，NC，USA）从 CT 上确定 SEEG 触点位置，并将输出结果结合起来，对每个触点进行定位并自动分配解剖标签。RNS 的触点位置也是以同样的方式确定的。使用内部 MATLAB 脚本（Mathworks, Inc.）BrainSuite 和 BrainStorm 软件 用于确认 SEEG 和 RNS 电极在脑体积内的接触位置。
Directional connectivity analysis using CCEPs 利用 CCEP 进行定向连通性分析
The process of CCEPs analysis was performed with BrainStorm using the method described by our group. After importing the raw SEEG data, excluding epochs contaminated with artefacts, we obtained CCEPs by averaging the remaining epochs time-locked to the stimulation. Baseline was taken from 100 to before stimulation. We categorized the CCEPs stimulation site and response site by their anatomical location. The anatomical location of CCEPs performed during presurgical SEEG evaluation was retrospectively compared to the ensuing locations of the RNS contacts. We analysed whether SEEG recording and stimulus sites of CCEPs close to the ensuing RNS contacts were associated with large in-degree [the sum of the potentials in the region of interest evoked by stimulating all other pairs of SEEG contacts (Fig. 1 bottom, left)] and outdegree CCEPs [the sum of potentials from the region of interest, namely, the degree of the output from the stimulus site (Fig. 1 bottom, right)] in patients with good RNS outcomes. CCEPs 分析过程是通过 BrainStorm 使用我们小组描述的方法进行的。 在导入 SEEG 原始数据并排除受伪影污染的历时后，我们对与刺激时间锁定的剩余历时进行平均，从而获得 CCEPs。基线取自刺激前 100 到 。我们将 CCEPs 的刺激部位和反应部位按其解剖位置进行分类。我们将手术前 SEEG 评估中进行的 CCEPs 解剖位置与随后的 RNS 接触位置进行了回顾性比较。我们分析了在 RNS 效果良好的患者中，SEEG 记录和 CCEPs 刺激位置靠近随后的 RNS 接触点是否与较大的内度 CCEPs [刺激所有其他成对 SEEG 接触点诱发的感兴趣区电位总和（图 1 左下方）] 和外度 CCEPs [感兴趣区电位总和，即刺激位置输出的程度（图 1 右下方）] 有关。
To evaluate directional connectivity, we analysed indegree and out-degree CCEPs. For in-degree CCEPs, we investigated the CCEPs responses seen in the SEEG contact closest to the ensuing RNS contact. The in-degree CCEPs are the sum of the potentials in the region of interest evoked by stimulating all other pairs of SEEG contacts (Fig. 1 bottom, left). For in-degree CCEPs, the analysis included responses in all regions of cortices stimulated. For out-degree CCEPs, we analysed the CCEPs responses in just two regions 为了评估定向连接性，我们分析了 "内度 "和 "外度 "CCEPs。 对于程度内 CCEPs，我们研究了在最靠近随后 RNS 接触的 SEEG 接触中看到的 CCEPs 反应。度内 CCEPs 是刺激所有其他成对 SEEG 触点所诱发的相关区域电位的总和（图 1 左下方）。对于程度内 CCEPs，分析包括所有受刺激皮层区域的反应。对于外度 CCEPs，我们只分析了两个区域的 CCEPs 反应
Patient selection 病人选择
Retrospective study 回顾性研究
Refractory epilepsy undergoing presurgical evaluation with SEEG 使用 SEEG 进行手术前评估的难治性癫痫
calculate distances between each SEEG contact and each RNS contact 计算每个 SEEG 触点和每个 RNS 触点之间的距离
CCEP analysis CCEP 分析
In-degree CCEP 学位内 CCEP
(all regions) (所有地区）
O SEEG electrode O SEEG 电极
SEEG contact (contact of interest) SEEG 联系人（相关联系人）
Stimulus site for CCEP 社区教育项目刺激站点
CCEP to the contact of interest 将 CCEP 发送给相关联系人
Out-degree CCEP 学位外 CCEP
(only on MTG/STG) (仅适用于 MTG/STG）
Recording site for out-degree CCEP on STG STG 上的学位外 CCEP 记录站点
Recording site for out-degree CCEP on MTG MTG 上的学位外 CCEP 记录站点
CCEP from the stimulus site of interest 相关刺激点的 CCEP
Figure I Flowchart of whole steps of recording and analyses. The functional anatomically-guided stacked-area (FAST) graph displays the sum of the responses at the particular area or SEEG contact to visualize and investigate 'in-degree' and 'out-degree' CCEPs. The 'in-degree' CCEPs indicate the sum of the potentials on the region of interest evoked by stimulating any pairs of SEEG contacts, namely, the degree of the input to the recording site. The 'out-degree' CCEPs indicate the sum of potentials from the region of interest, namely, the degree of output from the stimulus site. In this study, for the standardization of the CCEPs across patients, we focused on the 'out-degree' CCEPs from each stimulus site on only middle temporal gyrus (MTG) and superior temporal gyrus (STG), which were the only two common anatomical locations across the 12 patients. CCEPs, cortico-cortical evoked potential; SEEG, stereoelectroencephalography; RNS, responsive neurostimulation; FAST graph, functional anatomically-guided stacked-area graph; MTG, middle temporal gyrus; STG, superior temporal gyrus. 图 I 记录和分析整个步骤的流程图。功能解剖导向叠加区（FAST）图显示特定区域或 SEEG 接触点的反应总和，以直观显示和研究 "度内 "和 "度外 "CCEPs。度内 "CCEPs 表示刺激任何一对 SEEG 触点在感兴趣区域诱发的电位总和，即记录点的输入度。out-degree "CCEPs 表示来自感兴趣区域的电位总和，即从刺激部位输出的程度。在本研究中，为了对不同患者的 CCEPs 进行标准化，我们只关注了每个刺激部位的 "外度 "CCEPs，即颞中回（MTG）和颞上回（STG），这是 12 名患者中仅有的两个共同的解剖位置。CCEPs：皮质诱发电位；SEEG：立体脑电图；RNS：反应性神经刺激；FAST图：功能解剖学引导的叠加区域图；MTG：颞中回；STG：颞上回。
Figure 2 Configurations of SEEG and RNS electrodes. Configurations of SEEG and RNS electrodes are shown in anatomical labels in Talairach space (top for each patient) and on a presurgical 3D-MRI (bottom for each patient), except that for Patient II whose is shown in Fig. 3. IOZs were defined by the experts' review of ictal SEEG. The red lines on the electrode maps indicate the vertical line passing the anterior commissure (VAC line, between and ), the vertical line passing the posterior commissure (VPC line, between and ) and the line (a horizontal line between 8 and 9) in Talairach space to visualize and standardize the electrode locations of SEEG. Abbreviations: the conventions are same as for Fig. I. 图 2 SEEG 和 RNS 电极的配置。SEEG 和 RNS 电极的配置显示在 Talairach 空间的解剖标签上（每个患者的顶部）和手术前 3D-MRI 上（每个患者的底部），但患者 II 除外，其配置显示在图 3 中。IOZ 是由专家对发作期 SEEG 进行审查后确定的。电极图上的红线表示通过前神经丘的垂直线（VAC 线，位于 和 之间）、通过后神经丘的垂直线（VPC 线，位于 和 之间）以及 Talairach 空间中的 线（位于 8 和 9 之间的水平线），以直观和标准化 SEEG 的电极位置。缩写：与图 I 一致。
A
SEEG electrode SEEG 电极
SEEG electrode (IOZ) SEEG 电极（IOZ）
RNS electrode RNS 电极
Stimulus site for CCEP 社区教育项目刺激站点
Patient 11 病人 11
Recording site for out-degree CCEP on STG STG 上的学位外 CCEP 记录站点
Recording site for out-degree CCEP on MTG MTG 上的学位外 CCEP 记录站点
B
Figure 3 Example of CCEPs FAST graph and a sign permutation test. (A) Configurations of SEEG and RNS electrodes in a representative patient (Patient II). (Top) SEEG electrodes (white circle), those judged as ictal onset zone (IOZ) (red/solid circle) and RNS electrodes (blue/filled square) are shown in anatomical labels in Talairach space. Two red vertical lines and one red horizontal line indicate VAC, VPC and AC-PC lines, respectively, as in Fig. 2. (Bottom) The SEEG and RNS electrodes and stimulus sites (yellow/electric mark) are displayed on a presurgical (before SEEG evaluation) 3D-MRI. The light green (STG) and the light blue (MTG) areas represent the recording sites for out-degree CCEPs. (B) An example of CCEPs FAST graph recorded from the SEEG contacts on MTG. (Top) A total of seven pairs of SEEG contacts were stimulated for the out-degree CCEPs recording in this patient. Each graph indicates the stacked area plots of the rectified out-degree CCEPs waveforms presenting the sum of the amplitudes on the left and right MTGs. The horizontal axis indicates the analysed time window ( -100 to from the stimulation). The area above 0 and below 0 suggests the sum in the left MTG and that in the right MTG, respectively. (C) An example of a sign permutation test. The FAST graph of the sum of the out-degree CCEPs on MTG by stimulation of a pair of SEEG contacts on Lt MTG (highlighted with a rectangle in red in Fig. 2B) was evaluated statistically for each time point. The amplitude in each epoch was randomly multiplied by either or and then the epochs were averaged. This procedure was repeated 5000 times for each response (grey waveforms), and they were compared with the original CCEPs waveform (blue and black waveform) to show the data-driven distribution for the statistical analysis. For the multiple comparisons, we used a false discovery rate (FDR, ). The blue parts indicate the significant time points and the black parts the time points without significance. IOZ, ictal onset zone; Lt, left; Rt, right; Amyg, amygdala; PoCG, postcentral gyrus; SMG, 图 3 CCEPs FAST 图和符号排列检验示例。(A) 一名代表性患者（患者 II）的 SEEG 和 RNS 电极配置。(上图）SEEG 电极（白色圆圈）、被判定为发作起始区（IOZ）的电极（红色/实心圆圈）和 RNS 电极（蓝色/填充方形）以 Talairach 空间的解剖标记显示。两条红色垂直线和一条红色水平线分别表示 VAC、VPC 和 AC-PC 线，如图 2 所示。 (下图) SEEG 和 RNS 电极和刺激部位（黄色/电标记）显示在手术前（SEEG 评估前）的 3D-MRI 上。浅绿色（STG）和浅蓝色（MTG）区域代表度外 CCEP 的记录点。(B) 从 MTG 上 SEEG 触点记录的 CCEPs FAST 图示例。(上图）该患者共有七对 SEEG 触点受到刺激，以记录外度 CCEPs。每张图表都显示了整流后的度外 CCEPs 波形的叠加区域图，呈现了左右 MTG 上的振幅总和。横轴表示分析的时间窗口（刺激后 -100 至 ）。高于 0 和低于 0 的区域分别表示左侧 MTG 和右侧 MTG 的总和。(C) 符号置换测试示例。通过刺激左侧 MTG 上的一对 SEEG 触点（图 2B 中用红色矩形突出显示），对 MTG 上的外度 CCEPs 总和的 FAST 图进行统计评估。每个时间点的振幅随机乘以 或 ，然后取平均值。每个反应（灰色波形）重复此过程 5000 次，并与原始 CCEPs 波形（蓝色和黑色波形）进行比较，以显示统计分析的数据驱动分布。对于多重比较，我们使用了错误发现率（FDR， ）。蓝色部分表示有意义的时间点，黑色部分表示无意义的时间点。IOZ，发作开始区；Lt，左侧；Rt，右侧；Amyg，杏仁核；PoCG，中央后回；SMG、
supramarginal gyrus; AG, angular gyrus. Other conventions are same as for Figs I and 2. AG，角回。其他约定与图 I 和图 2 相同。
that were common to all patients, MTG and STG, following stimulation of site closest to the ensuing RNS contact. The out-degree CCEPs are the sum of potentials from the region of interest, namely, the degree of the output from the stimulus site (Fig. 1 bottom, right). 所有患者（MTG 和 STG）在刺激与随后的 RNS 接触最接近的部位后，都会出现相同的 CCEPs。外度 CCEPs 是相关区域电位的总和，即刺激部位输出的程度（图 1 右下方）。
In performing our CCEPs analysis, we generated functional anatomically-guided stacked-area (FAST) graphs of CCEPs responses. Details of the FAST graph have been previously reported by our group. In brief, area plots of rectified CCEPs waveforms are stacked to present the sum of responses, either across all contacts for a single stimulation pair (out-degree), or at a single contact across multiple stimulations (in-degree). An example of the FAST graph in a representative patient (Patient 11) is shown in Fig. 3. The locations of SEEG and RNS electrodes and stimulus sites for CCEPs in this patient are displayed in Fig. 3A. In previous studies recorded with subdural electrodes, CCEPs were typically composed of two negative potentials (early and late ), and the latency of N1 and N2 usually ranged from and , respectively. However, the CCEPs in patients with SEEG demonstrate more complex waveforms, likely due to variable contact locations within the six layers of 在进行 CCEPs 分析时，我们生成了 CCEPs 反应的功能解剖导向叠加区图（FAST）。关于 FAST 图形的详细信息，我们的研究小组之前已经做过报告。 简而言之，将整流 CCEPs 波形的区域图进行堆叠，以显示单个刺激对的所有触点的反应总和（out-degree），或多个刺激的单个触点的反应总和（in-degree）。图 3 显示了一位代表性患者（患者 11）的 FAST 图示例。图 3A 显示了 SEEG 和 RNS 电极的位置以及该患者 CCEPs 的刺激点。在之前使用硬膜下电极记录的研究中，CCEP 通常由两个负电位（早期 和晚期 ）组成，N1 和 N2 的潜伏期通常分别为 和 。 然而，SEEG 患者的 CCEP 波形更为复杂，这可能是由于六层脑膜的接触位置不同所致。
cerebral cortex and white matter. In order to overcome this complexity, we used the sum of the absolute value of the CCEPs response, separated into early , middle and late latency periods (Figs 1 and . 大脑皮层和白质。 为了克服这种复杂性，我们使用了 CCEPs 反应绝对值的总和，将其分为早期 、中期 和晚期潜伏期 （图 1 和 ）。
For in-degree CCEPs, we calculated the CCEPs FAST graph at each SEEG contact from all stimulation sites. The responses for each SEEG contact were sorted in descending RMS order and ranked for the early, middle and late latency periods in each patient. SEEG contacts in both grey and white matters were included in the in-degree CCEPs. In our experience, CCEPs performed during SEEG can show meaningful responses in electrodes within both white and grey matters. To specify the responses of each location including white matter, we adopted a bipolar montage for the analyses of CCEPs and corresponding FAST graphs in the localization of in-degree CCEPs. For out-degree CCEPs, we grouped similar stimulation and responses in the only two common anatomical locations across patients (MTG and STG). The out-degree CCEPs FAST graphs from MTG/STG were recorded using a referential montage (referenced the scalp vertex region), since the possibility of far field responses was eliminated by sampling only the contacts in this cortical region of interest in all patients. In this study, we confirmed the area showing hyperperfusion in ictal SPECT based on our previous report demonstrating strong connectivity between the ictal onset zone and hyperperfused regions in ictal SPECT Eleven out of 12 patients underwent an ictal SPECT examination during preoperative evaluation. The rate of significant out-degree CCEPs between patients with and without ictal hyperperfusion on MTG/STG in subtraction ictal SPECT co-registered with MRI (SISCOM) was compared. 对于同度 CCEPs，我们计算了所有刺激部位每个 SEEG 接触点的 CCEPs FAST 图。每个 SEEG 接触点的反应按 RMS 降序排序，并对每位患者的早期、中期和晚期潜伏期进行排序。灰质和白质中的 SEEG 触点都包含在同度 CCEP 中。根据我们的经验，在 SEEG 期间进行的 CCEP 可在白质和灰质中的电极上显示有意义的反应。为了明确包括白质在内的每个位置的反应，我们采用了双极蒙太奇来分析 CCEPs 和相应的 FAST 图来定位度内 CCEPs。 对于度外 CCEPs，我们将患者仅有的两个共同解剖位置（MTG 和 STG）的类似刺激和反应分组。来自 MTG/STG 的外度 CCEPs FAST 图使用参照蒙太奇（以头皮顶点区域为参照）进行记录，因为在所有患者中只对该皮质感兴趣区域的触点进行取样，从而消除了远场反应的可能性。在本研究中，我们根据之前的报告确认了在发作期 SPECT 中出现高灌注的区域，该报告显示发作期起始区与发作期 SPECT 中的高灌注区域之间具有很强的连通性 12 名患者中有 11 名在术前评估时接受了发作期 SPECT 检查。我们比较了在与核磁共振成像（SISCOM）共同注册的减影发作期SPECT检查中，MTG/STG有发作期高灌注和无发作期高灌注的患者之间有明显度外CCEP的比率。
Outcome of RNS therapy RNS 治疗的结果
Seizure outcomes in patients undergoing epilepsy surgery are typically reported using the Engel classification or International League Against Epilepsy (ILAE) classification. For RNS seizure outcomes, the response to neuromodulation changes over time. Therefore, the percentage seizure reduction at one time point to the baseline seizure frequency pre-RNS therapy is more clinically meaningful. We applied a modified percentage outcome scale based on the seizure reduction from pre-RNS seizure frequency. These seizure outcome categories consisted of: worsened (scale -1), no change (scale 0 ), seizure reduction (scale 1), seizure reduction (scale 2 ), seizure reduction (scale 3), seizure reduction (scale 4 ) and seizure reduction (scale 5). The follow-up period for evaluation ranged from 1.3-4.8 years (median 2.7) after RNS therapy was initiated. In two patients (#3 and #10), we could not accurately determine the seizure outcome due to the patient's inability to report their seizure frequency reliably. Of the two patients seizure reduction; scale 4, one had antiseizure medication (ASM) reduced and one had no change; one patient seizure reduction; scale 3 ) had no ASM change; four patients seizure reduction; scale 2), three had no change and one had ASM increased; three patients seizure reduction; scale 1 ), two had no change and one had ASM increased. 接受癫痫手术的患者的癫痫发作结果通常采用恩格尔分类 或国际抗癫痫联盟（ILAE）分类进行报告。 对于 RNS 癫痫发作结果，对神经调控的反应会随着时间的推移而改变。因此，一个时间点的癫痫发作减少百分比与 RNS 治疗前的基线癫痫发作频率相比更具临床意义。 我们根据与 RNS 治疗前发作频率相比发作减少的百分比，对结果进行了修改。这些癫痫发作结果类别包括：恶化（评分-1）、无变化（评分 0）、 癫痫发作减少（评分 1）、 癫痫发作减少（评分 2）、 癫痫发作减少（评分 3）、 癫痫发作减少（评分 4）和 癫痫发作减少（评分 5）。在开始接受 RNS 治疗后，随访评估期为 1.3-4.8 年（中位数为 2.7 年）。在两名患者（3 号和 10 号）中，由于患者无法可靠地报告其癫痫发作频率，我们无法准确确定其癫痫发作结果。在两名患者 癫痫发作减少；量表 4 中，一名患者的抗癫痫药物（ASM）减少，一名患者没有变化；一名患者 癫痫发作减少；量表 3 ）的 ASM 没有变化；四名患者 癫痫发作减少；量表 2 ）中，三名患者的 ASM 没有变化，一名患者的 ASM 增加；三名患者 癫痫发作减少；量表 1 ）中，两名患者的 ASM 没有变化，一名患者的 ASM 增加。
Grouping of SEEG contacts according to the distance to the ensuing closest RNS contact 根据与随后最近的 RNS 联系人的距离对 SEEG 联系人进行分组
We classified the SEEG contacts (for stimulation and recording in CCEPs) into four groups based on distance to the closest ensuing RNS contact: Group 1 (G1: 0-5 mm), Group 2 (G2: 5-10 mm), Group 3 (G3: 10-20 mm) and Group 4 (G4: ). Due to the number of stimulation pairs being smaller than those for the recording, we combined G3 and G4 as G3-4 (>10 mm from the closest RNS contact) for the analyses of out-degree CCEPs. 我们根据与最近的 RNS 接触点的距离将 SEEG 接触点（用于 CCEPs 的刺激和记录）分为四组：第 1 组（G1：0-5 毫米）、第 2 组（G2：5-10 毫米）、第 3 组（G3：10-20 毫米）和第 4 组（G4： ）。由于刺激对的数量少于记录对的数量，我们将 G3 和 G4 合并为 G3-4（距离最近的 RNS 接触点大于 10 毫米），用于分析度外 CCEP。
Statistical analysis 统计分析
For each latency period (early, middle and late), we compared the ranks according to the highest amplitude over the time window of in-degree CCEPs in each group described above using a Mann-Whitney test a -value of was considered significant for multiple comparisons (G1 versus G2, G1 versus G3 and G1 versus G4) by Bonferroni correction]. We standardized the CCEPs per patient in this way because the distribution and amplitude of CCEPs are different for each patient. We then calculated the in-degree CCEPs ratios for G1, G2 and G3 relative to G4 by using the highest amplitude of the sum of CCEPs. The mean of the highest amplitude of the sum of CCEPs in G1, G2 and G3 was divided by that in G4 for the early, middle and late latency periods in each patient. The correlations between the in-degree CCEPs ratios (G1/G4, G2/G4 and G3/G4 ratios) and the outcome of RNS therapy were evaluated by the Pearson correlation coefficient (a -value of was considered significant for each for multiple comparisons by Bonferroni correction). We examined the seizure reduction rate every visit. Variations of the outcome of RNS therapy are known to occur based on programming changes to both detection and stimulation parameters. Thus, we adopted the score at the time of the greatest seizure reduction after RNS therapy for each patient in this study. 对于每个潜伏期（早期、中期和晚期），我们使用 Mann-Whitney ，根据上述各组阶内 CCEPs 在时间窗内的最高振幅进行了等级比较， ， ，经 Bonferroni 校正后认为多重比较（G1 与 G2、G1 与 G3 和 G1 与 G4）具有显著性]。由于每位患者的 CCEPs 分布和振幅不同，因此我们以这种方式对每位患者的 CCEPs 进行了标准化处理。然后，我们使用 CCEPs 总和的最高振幅计算 G1、G2 和 G3 相对于 G4 的度内 CCEPs 比率。将 G1、G2 和 G3 中 CCEPs 最高振幅之和的平均值除以 G4 中早、中、晚潜伏期的 CCEPs 最高振幅之和的平均值。度内 CCEPs 比值（G1/G4、G2/G4 和 G3/G4 比值）与 RNS 治疗结果之间的相关性通过皮尔逊相关系数进行评估（经 Bonferroni 校正后，多重比较中每项的 - 值 均被视为显著）。我们对每次就诊的癫痫发作减少率进行了检查。众所周知，RNS 治疗的结果会因检测和刺激参数的程序更改而发生变化。 因此，在本研究中，我们采用了每位患者接受 RNS 治疗后癫痫发作减少最多时的评分。
When investigating the out-degree CCEPs, we performed a sign permutation test to extract statistically significant CCEPs for each stimulus site in the aforementioned FAST graph based on previously established methods. The details of the sign permutation test are further described by our group. In order to evaluate significance, the amplitude in each epoch was randomly multiplied by either +1 or -1 and the epochs were averaged. This procedure was repeated 5000 times for each response to generate a data-driven null distribution. This was then compared with the original CCEPs waveform for statistical analysis (Fig. 3C). For multiple comparison correction, we used a false discovery rate correction protocol (FDR, ). CCEPs were deemed 在研究外度 CCEPs 时，我们根据以前建立的方法，对上述 FAST 图中的每个刺激点进行了符号置换检验，以提取具有统计意义的 CCEPs。 关于符号置换检验的细节，我们的研究小组有进一步的说明。 为了评估显著性，我们随机将每个时程的振幅乘以+1或-1，然后取平均值。每个反应重复这一过程 5000 次，以生成数据驱动的空分布。然后将其与原始 CCEPs 波形进行比较，以进行统计分析（图 3C）。对于多重比较校正，我们使用了错误发现率校正协议（FDR， ）。CCEPs 被认为是
significant when at least 1 time point showed significance in the sign permutation test. The rate of significant out-degree CCEPs, namely the number of stimulus pairs producing significant CCEPs divided by the number of all stimulus pairs, was then calculated for MTG and STG in each patient. For group comparisons of the rate of significant out-degree CCEPs between patients with and without ictal hyperperfusion on MTG/STG in SISCOM, we applied a MannWhitney test [a -value of was considered significant for multiple comparisons (early, middle and late latency periods on MTG/STG) by Bonferroni correction]. A Wilcoxon Signed-rank test was performed for the comparison between G1 and G3-4 for out-degree CCEPs (a -value of was considered significant). 当至少有 1 个时间点在符号置换测试中显示出显著性时，则为显著。然后计算每名患者 MTG 和 STG 的显著度外 CCEPs 率，即产生显著 CCEPs 的刺激对数量除以所有刺激对数量。为了对 SISCOM 中 MTG/STG 上有和没有发作性高灌注的患者之间的显著度外 CCEPs 率进行分组比较，我们采用了 MannWhitney 检验[通过 Bonferroni 校正， - 值为 则认为多重比较（MTG/STG 的早期、中期和晚期潜伏期）具有显著性]。对 G1 和 G3-4 之间的阶外 CCEP 比较进行了 Wilcoxon Signed-rank 检验（ -值为 即为显著）。
An additional analysis was done for out-degree CCEPs and is available in Supplementary Material. 对学位外 CCEP 进行了补充分析，见补充材料。
Results 成果
In-degree CCEPs to the closest RNS contacts 与最近的区域联络网联系的内向 CCEP
Following SEEG evaluation, the electrodes for RNS therapy were implanted with up to four electrodes of which only two can be connected: one electrode in one patient, two electrodes in six patients (bilaterally in three patients), three electrodes in three patients (bilaterally in one patient) and four electrodes in one patient. Figure 4 represents the ranks of all contacts sorted by the values of in-degree CCEPs in each group relative to their rank among all contacts in a representative patient (Patient 11). The rank of G1 from the closest RNS contact) was significantly higher than that of G4 from the closest RNS contact) for the early, middle and late latency periods , for all) and that of G3 from the closest RNS contact) for the early latency period , whereas there were no differences between G1 and G2 or between G1 and G3 (for the middle and late latency periods) . The in-degree CCEPs were significantly higher in G1 than G4 in six patients for the early and middle latency periods ( , for all), and in five patients for the late latency periods , for all). The in-degree CCEPs were also significantly higher in G1 than G3 in four patients for the middle and late latency periods , for all) and in five patients for the early latency period , for all). There was no difference between G1 and G2 for all the latency periods (early, middle and late) in any of the patients . This analysis could not be assessed in two out of 12 patients due to lack of CCEPs data in proximity to the eventual RNS contact. The results of the statistical tests for all patients are summarized in Table 1. SEEG 评估后，用于 RNS 治疗的电极最多植入四个，其中只能连接两个：一名患者植入一个电极，六名患者植入两个电极（三名患者为双侧），三名患者植入三个电极（一名患者为双侧），一名患者植入四个电极。图 4 显示了所有触点的等级，按每组中的同度 CCEPs 值相对于其在具有代表性的患者（患者 11）的所有触点中的等级进行排序。在早期、中期和晚期潜伏期，G1 from the closest RNS contact) 的排名均明显高于 G4 from the closest RNS contact) 的排名 ，在早期潜伏期，G3 from the closest RNS contact) 的排名 ，而 G1 和 G2 之间以及 G1 和 G3 之间（在中期和晚期潜伏期）没有差异 。在早期和中期潜伏期，G1 中有 6 名患者的度内 CCEPs 明显高于 G4（ ，全部），在晚期潜伏期，有 5 名患者的度内 CCEPs 明显高于 G4（ ，全部）。在中潜伏期和晚潜伏期，G1 级有 4 名患者的度内 CCEPs 明显高于 G3 级 ，所有患者均是如此），在早潜伏期，有 5 名患者的度内 CCEPs 明显高于 G3 级 ，所有患者均是如此）。在所有潜伏期（早期、中期和晚期），G1 和 G2 在所有患者中均无差异 。由于缺乏与最终 RNS 接触附近的 CCEPs 数据，无法对 12 名患者中的两名进行此项分析评估。表 1 汇总了所有患者的统计检测结果。
Correlation between in-degree CCEPs ratio and RNS outcomes 学位内 CCEPs 比率与 RNS 结果之间的相关性
Figure 5 shows the correlations between the in-degree CCEPs ratios (for early, middle and late latency periods) 图 5 显示了度内（早期、中期和晚期）CCEPs 比率之间的相关性。
Table I Comparison of in-degree CCEPs between Group I (0-5 mm from RNS contacts) and Groups 2-4 表 I 第一组（距离 RNS 触点 0-5 毫米）与第二至第四组的同度 CCEP 比较
Patient
Group I versus 第 I 组 与
Group
Group I versus Group 3 第 I 组 与第 3 组
Group I versus Group 4 第 I 组 与第 4 组
Early
Middle
Late
Early
Middle
Late
Early
Middle
Late
I
0.021
0.040
0.029
2
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
3
0.400
0.200
0.400
0.643
0.143
0.143
0.148
0.895
0.876
4
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
5
0.128
0.248
0.248
6
0.868
0.616
0.525
7
0.029
0.396
0.672
0.051
0.786
0.197
0.439
8
0.042
0.148
0.057
0.058
9
0.842
1.000
1.000
0.482
0.750
0.841
0.619
0.431
0.436
10
0.081
0.938
0.815
0.118
0.104
0.548
0.042
0.113
II
0.357
0.699
0.809
0.180
0.213
12
0.404
0.525
0.404
0.274
0.904
0.659
0.195
0.224
0.576
n.a., not available. A P-value of was considered significant for all the statistical tests in the comparison between the Group I and Groups 2-4 for multiple comparison with Bonferroni correction. n.a.，不详。经 Bonferroni 校正后，在第一组和第二至第四组之间进行多重比较的所有统计检验中，P 值 均被视为具有显著性。
A
B
In-degree CCEP ratio (G2/G4) 学位内 CCEP 比率（G2/G4）
C
In-degree CCEP ratio (G1/G4) 学位内 CCEP 比率（G1/G4）
Late latency period 后期潜伏期
In-degree CCEP ratio (G1/G4) 学位内 CCEP 比率（G1/G4）
Figure 5 Correlation of the in-degree CCEPs ratio with the outcome value of RNS therapy. A significant correlation was seen between the in-degree CCEPs ratios of GI/G4 and the outcome for the early latency period ( , while there were no correlations for middle and late latency periods ( , for both) (A). Individual data points on these plots for panel are shown in Supplementary Table 2. There were no correlations between the in-degree CCEPs ratios of G2/G4 or G3/G4 and the outcome (early, middle and late latency periods) ( , for all) (B and ). The Pearson correlation coefficient was used for all the statistical tests. 图 5 程度内 CCEPs 比率与 RNS 治疗结果值的相关性。GI/G4 的阶内 CCEPs 比率与早期潜伏期（ ，而中期和晚期潜伏期（ ，两者均为）的结果之间存在明显的相关性（A）。 面板图中的各个数据点见补充表 2。G2/G4或G3/G4的阶内CCEPs比率与结果（早期、中期和晚期潜伏期）之间没有相关性（ ，全部）（B和 ）。所有统计检验均采用皮尔逊相关系数。
A
Early latency period 早期潜伏期
B Middle latency period B 中间潜伏期
C Late latency period C 延迟期
Figure 6 Significant out-degree CCEPs and the correlation with the distance between stimulus site and the closest RNS contact. Each data represents the ratio of significant out-degree CCEPs on MTG/STG, namely the number of stimulus pairs producing significant CCEPs divided by the number of all stimulus pairs. The stimulus sites with distances of to the closest RNS contacts ( : 0-5 mm) showed more significant out-degree early latency period than those with distances more than to the closest RNS contacts ( : . There was no difference in the significant out-degree CCEPs between GI and G3-4 for middle or late latency period , for both). A Wilcoxon Signed-rank test was used for all statistical tests. The numbers for each group for early, middle and late latency periods are 16. 图 6 显著的外度 CCEPs 以及与刺激点和最近的 RNS 接触点之间距离的相关性。每个数据都代表了 MTG/STG 上显著的度外 CCEPs 比例，即产生显著 CCEPs 的刺激对数量除以所有刺激对数量。与最近的 RNS 接触点距离为 的刺激点（ : 0-5 mm）相比，与最近的 RNS 接触点距离超过 的刺激点（ : 。在 和晚潜伏期 ，GI 和 G3-4 之间的 CCEP 显著偏离度没有差异。）所有统计检验均采用 Wilcoxon Signed-rank 检验。每组早、中、晚潜伏期的人数均为 16 人。
and RNS seizure outcomes. There was a significant correlation between the early latency period of the in-degree CCEPs ratios of and outcome . This result implies that high in-degree CCEPs early latency period ratios correlate with better outcome of RNS therapy, whereas there were no significant correlations for middle and late latency periods ( ) (Fig. 5A). No significant correlations were seen between the in-degree CCEPs ratios of G2/G4 or G3/G4 and the outcome for early, middle and late latency periods ( , for all) (Fig 5B and C). 和 RNS 癫痫发作结果。度内 CCEPs 早期潜伏期比率 与结果 之间存在明显的相关性。这一结果意味着，高阈值 CCEPs 早期潜伏期比率与更好的 RNS 治疗结果相关，而中期和晚期潜伏期 ( ) 则无显著相关性（图 5A）。G2/G4或G3/G4的度内CCEPs比率与早期、中期和晚期潜伏期的疗效均无明显相关性( , for all)（图5B和C）。
Ratio of stimulus sites producing significant out-degree CCEPs on MTG/STG 在 MTG/STG 上产生明显外度 CCEP 的刺激点比率
In all patients, FAST graphs for the out-degree CCEPs on MTG/STG were constructed and a sign permutation test was performed. The ratios of stimulus sites with significant out-degree CCEPs on MTG and STG to all stimulus sites for early, middle and late latency periods were calculated in each patient. There was no association of significant outdegree CCEPs (either early, middle or late latency period) to the epilepsy classification. Five patients (Patients 3, 5, 6, 7, 10) showed hyperperfusion on MTG/STG in SISCOM. There was no difference in the rate of significant out-degree CCEPs on MTG/STG between patients with or without SISCOM hyperperfusion on MTG/STG for early ( 0.792 ), middle or late latency period . 在所有患者中，绘制了 MTG/STG 上外度 CCEP 的 FAST 图，并进行了符号排列检验。计算了每位患者早、中、晚潜伏期在 MTG 和 STG 上有显著度外 CCEPs 的刺激点与所有刺激点的比率。无论是早期、中期还是晚期潜伏期，明显偏离度 CCEP 与癫痫分类均无关联。五名患者（患者 3、5、6、7、10）在 SISCOM 中的 MTG/STG 上出现高灌注。在早期（ 0.792）、中期 或晚期潜伏期 ，MTG/STG 上有无 SISCOM 过度灌注的患者在 MTG/STG 上出现明显程度外 CCEP 的比率没有差异。
Out-degree CCEPs to the closest RNS contacts 将 CCEP 外派到最近的区域网络联系点
The ratios of stimulus sites with significant out-degree during the early latency period over MTG/STG was compared to all stimulus sites based on the distance from SEEG CCEPs stimulus site to the closest RNS contact (grouped by G1 and G3-4). CCEPs in the G1 group (0-5 mm to the closest RNS contacts) showed more significant out-degree during the early latency period than those in G3-4 (>10 to the closest RNS contacts) ( , a Wilcoxon Signed-rank test) (Fig. 6A). There was no difference in the significant outdegree CCEPs between the stimulus site in G1 and that in G3-4 for middle or late latency period , for both (Fig. 6B and C). We could only perform limited analysis of the seizure outcomes relating to out-degree CCEPs as described in Supplementary Material as the regions of brain sampled and stimulated varied among the limited number of patients. A modest negative correlation was observed between the out-degree and the outcome in the late latency period, suggesting that better outcomes were associated with smaller outdegree CCEPs, when the electrodes in these two distance groups (G1 and G2) were stimulated (Supplementary Fig. 1). Also, we found a larger out-degree in patients with poor outcome (outcome scale 1), especially in the middle and late latency periods, for all three distance groups, excluding the late latency period in G2 (Supplementary Fig. 2). 根据 SEEG CCEPs 刺激点到最近 RNS 接触点（按 G1 和 G3-4 分组）的距离，比较了 MTG/STG 与所有刺激点相比，在早期潜伏期具有显著偏离度的刺激点比率。在早期潜伏期，G1 组（距离最近的 RNS 接触点 0-5 mm）的 CCEPs 比 G3-4 组（距离最近的 RNS 接触点 >10 ）的 CCEPs 表现出更明显的失度（ ，Wilcoxon Signed-rank 检验）（图 6A）。在潜伏期中期或晚期 ，G1 和 G3-4 的刺激位点之间的显著外度 CCEP 没有差异， （图 6B 和 C）。由于取样和刺激的大脑区域在有限的患者中各不相同，我们只能对 "补充材料 "中描述的与程度外 CCEP 相关的癫痫发作结果进行有限的分析。在晚潜伏期，我们观察到外度数与结果之间存在适度的负相关，这表明在刺激这两个距离组（G1 和 G2）的电极时，较好的结果与较小的外度数 CCEP 相关（补充图 1）。此外，我们还发现，在所有三个距离组中，除 G2 的晚期潜伏期外，预后较差（预后量表 1）的患者，尤其是在中晚期潜伏期，CCEP 的失度较大（补图 2）。
Discussion 讨论
Correlation of CCEPs and RNS CCEP 与 RNS 的相关性
outcomes 结果
In this study, we analysed CCEPs, during SEEG evaluations, in patients who subsequently underwent RNS therapy. CCEPs, a measure of evoked effective connectivity, enables an assessment of in vivo directional connectivity. Directional connectivity derived by CCEPs can be categorized as in-degree measures, representing the total number of significant responses evoked at a region of interest upon stimulation of all other sites, and out-degree CCEPs, 在本研究中，我们分析了接受 RNS 治疗的患者在 SEEG 评估期间的 CCEPs。CCEPs 是一种诱发有效连通性的测量方法，可用于评估体内定向连通性。 通过 CCEPs 得出的定向连通性可分为度内测量和度外 CCEPs，度内测量代表的是刺激所有其他部位时在感兴趣区域诱发的显著反应的总数、
representing the number of significant responses following stimulation of region of interest. We investigated the correlation between CCEPs and clinical characteristics including RNS contact locations and seizure outcomes from RNS therapy. The main findings of our study are as follows: (i) significant in-degree early latency CCEPs were seen at sites close to the eventual RNS electrode placement (Fig. 4A); (ii) a significant correlation was found between in-degree (in the early CCEPs latency period) and RNS seizure outcome, based on the distance from the eventual RNS contact location (Fig. 5A); and (iii) CCEPs stimulation sites within 0-5 mm of the eventual RNS electrodes showed more significant outdegree during the early latency period than those more than away (Fig. 6A). 代表刺激相关区域后出现明显反应的次数。我们研究了 CCEPs 与临床特征（包括 RNS 接触位置）和 RNS 治疗后癫痫发作结果之间的相关性。我们研究的主要发现如下：(i) 在靠近最终 RNS 电极放置位置的部位出现了明显的同度早期潜伏 CCEPs（图 4A）；(ii) 根据与最终 RNS 接触位置的距离，发现同度（CCEPs 早期潜伏期）与 RNS 癫痫发作结果之间存在明显的相关性（图 5A）；以及 (iii) 根据与最终 RNS 接触位置的距离，发现同度（CCEPs 早期潜伏期）与 RNS 癫痫发作结果之间存在明显的相关性（图 5B）。5A）；(iii) 在早期潜伏期，距离最终 RNS 电极 0-5 毫米范围内的 CCEPs 刺激点比距离超过 的刺激点显示出更明显的失度（图 6A）。
RNS electrode locations were placed based on expert analysis of SEEG IOZ localization. The CCEPs findings were not considered by the expert when determining the location of the RNS electrodes. The analysis of evoked functional connectivity in these patients, particularly the early latency period CCEPs, correlated well with clinically determined RNS electrode locations (both in-degree and out-degree) and degree of seizure reduction (only in-degree). This finding suggests that in addition to the standard clinical analyses of the IOZ, investigation of the effective connectivity by CCEPs could help guide the placement of electrodes for RNS therapy as an objective measure of the epileptogenic network. RNS 电极的位置是根据专家对 SEEG IOZ 定位的分析确定的。专家在确定 RNS 电极位置时并未考虑 CCEPs 结果。这些患者的诱发功能连接分析，尤其是早期潜伏期的 CCEPs，与临床确定的 RNS 电极位置（内度和外度）和癫痫发作减少程度（仅内度）密切相关。这一发现表明，除了对 IOZ 进行标准的临床分析外，通过 CCEPs 对有效连接性进行调查有助于指导 RNS 治疗的电极位置，从而客观衡量致痫网络。
Mechanisms of RNS therapy RNS 治疗的机制
The observation of both acute as well as delayed therapeutic effects of RNS suggests that there are likely multiple mechanisms of actions. Examples of the acute effect of electrical stimulation include in vitro amplitude reduction of epileptiform activity, aborting after discharges and suppression of phase locked gamma-frequencies. A recent study has shown that acute stimulation related reduction in IEEG spectral power was associated with reductions in clinical seizure frequency. Such acute effects could be stimulus dependent and involve changes to neurotransmitters regulating the balance of excitation and inhibition. Electrical stimulation has also been shown to influence neuronal activity acutely distant from the site of stimulation. Progressive improvement in seizure control over years suggests that there are likely chronic effects related to electrical stimulation. Furthermore, one study found that modulation of the epileptic network over time appeared to correlate with changes noted in regions remote to the sites of stimulation. In a recent study focusing on the chronic network re-organization, patients with the greatest therapeutic benefit of RNS therapy showed progressive, frequency-dependent re-organization of interictal functional connectivity. In relation to the epileptic network, subregions of the cerebral cortex can be characterized based on the degree of incoming and outgoing connections. A key finding of our study is that RNS electrodes tend to be placed in regions of the cerebral cortex that could be described as major receiver nodes (receivers of influence) of large scale cortico-cortical influence. This conclusion is based on the correlations seen with early latency period in-degree CCEPs with eventual RNS electrode locations and seizure reduction. This finding is consistent with the observation that modulation of the epileptic network in regions distant to the site of stimulation may be important to the mechanisms by which responsive neurostimulation exerts its effects. 对 RNS 急性和延迟治疗效果的观察表明，其作用机制可能是多重的。 电刺激的急性效应包括癫痫样活动的体外振幅减弱、 ，放电后终止 以及锁相伽马频率的抑制。 最近的一项研究表明，与急性刺激相关的 IEEG 频谱功率降低与临床癫痫发作频率降低有关。 这种急性效应可能依赖于刺激，并涉及调节兴奋和抑制平衡的神经递质的变化。 也有研究表明，电刺激会对远离刺激部位的神经元活动产生急性影响。 多年来癫痫发作控制的逐步改善表明，电刺激很可能会产生慢性影响。 此外，一项研究发现，随着时间的推移，癫痫网络的调节似乎与远离刺激部位的区域的变化相关。 在最近一项关注慢性网络重组的研究中，RNS 治疗获益最大的患者表现出发作间期功能连接的渐进式、频率依赖性重组。 就癫痫网络而言，大脑皮层亚区的特征可基于传入和传出连接的程度。 我们研究的一个重要发现是，RNS 电极往往位于大脑皮层的一些区域，而这些区域可被描述为大规模皮质-皮质影响的主要接收节点（影响接收器）。这一结论是基于早期潜伏期同度 CCEP 与最终 RNS 电极位置和癫痫发作减少之间的相关性得出的。这一发现与以下观察结果一致，即在远离刺激部位的区域对癫痫网络的调节可能对反应性神经刺激发挥其作用的机制非常重要。
Stimulus frequency may impact the therapeutic effect of electrical stimulation. Various stimulation frequencies have been studied for their potential neuromodulatory effects in epilepsy. Low-frequency SPES ) also employed in CCEPs recordings has been shown to have inhibitory effects on ongoing epileptiform activity. High-frequency electrical stimulation has been shown to desynchronize neuronal activities. We studied the evoked effective connectivity using stimulation, while RNS therapy typically utilizes 100 stimulation. Modelling changes in cortical excitability with repetitive stimulation (higher than ) can also be used to understand evoked connectivity profiles, although RNS does not seek to study cortico-cortical connectivity using this high-frequency stimulation. Keller et al. showed that singlepulse evoked connectivity profiles provided a high degree of accuracy and discriminability of the cortical regions that exhibited changes in excitability to repetitive stimulation. In our study, we showed that single-pulse evoked connectivity profiles can also inform degree of seizure reduction in patients undergoing RNS (which typically utilizes higher frequency stimulation bursts). Greater in-degree CCEP, as a ratio distance, correlated with greater seizure reduction. A possible explanation of this finding is that these receiver node regions (higher in-degree) are more capable of being modulated by electrical stimulation than those regions of the brain with less in-degree properties. 刺激频率可能会影响电刺激的治疗效果。人们研究了各种刺激频率对癫痫的潜在神经调节作用。 低频 SPES ) 也用于 CCEPs 记录，已被证明对正在进行的癫痫样活动有抑制作用。 高频电刺激可使神经元活动失同步。 我们使用 刺激来研究诱发的有效连接，而 RNS 治疗通常使用 100 刺激。 模拟皮层兴奋性在重复刺激（高于 ）下的变化也可用于了解诱发的连通性概况，尽管 RNS 并不寻求使用这种高频刺激来研究皮层-皮层连通性。 Keller 等人的研究 表明，单脉冲诱发的连通性图谱对 重复刺激下兴奋性发生变化的皮层区域具有高度的准确性和可辨别性。在我们的研究中，我们发现单脉冲诱发连接图谱也能告知接受 RNS（通常使用较高频率的脉冲刺激）的患者癫痫发作的减少程度。作为比值距离，程度更高的 CCEP 与癫痫发作减少程度相关。对这一发现的一种可能解释是，这些接收节点区域（内度较高）比那些内度属性较低的大脑区域更能受到电刺激的调节。
Assessing RNS electrode locations 评估 RNS 电极位置
Utilizing functional connectivity as a method for determining the appropriate settings of RNS therapy, including assessing potential sites of RNS electrodes, has previously not been established to the best of our knowledge. After the initial publications of CCEPs that evaluated various functional networks, CCEPs connectivity measures were used to study the epileptogenic network. SPES of the IOZ during seizure resulted in larger early CCEPs responses (within after stimulation) than those during the interictal period, suggesting that the early latency period of CCEPs could be a marker of cortical excitability at the epileptic focus. separate study comparing the early latency period of CCEPs at IOZ and non-IOZ regions showed that CCEPs at IOZ produced significantly higher amplitudes than non-IOZ regions. These accentuated CCEPs amplitudes, near IOZ, could reflect an increased cortical excitability associated with epileptogenicity. CCEPs connectivity was also studied in the propagation network of epileptic seizures. CCEPs evoked gamma band activity differed between early versus late ictal spread regions. CCEPs were also shown to have strong correlations between the evoked connectivity profiles 据我们所知，利用功能连通性作为确定 RNS 治疗的适当设置（包括评估 RNS 电极的潜在位置）的方法以前尚未建立。在最初发表评估各种功能网络的 CCEPs 之后， CCEPs 连接性测量被用于研究致痫网络。在癫痫发作期间对 IOZ 进行 SPES 会导致比发作间期更大的早期 CCEPs 反应（刺激后 内），这表明 CCEPs 的早期潜伏期可能是癫痫灶皮质兴奋性的标志。 比较 IOZ 和非 IOZ 区域 CCEPs 早期潜伏期的单独研究显示，IOZ 区域的 CCEPs 产生的振幅明显高于非 IOZ 区域。 在 IOZ 附近，CCEPs 振幅增大，这可能反映了与致痫性相关的皮层兴奋性增高。我们还研究了癫痫发作传播网络中的 CCEPs 连接性。CCEPs诱发的伽马带活动在早期和晚期发作扩散区域有所不同。 研究还显示，CCEPs 的诱发连接图谱与癫痫发作的传播网络之间具有很强的相关性。
A
Figure 7 Schema of interpretations of in-degree and out-degree CCEPs in relation to RNS therapy. (A) The in-degree CCEPs recorded from contacts close to the RNS contact (GI) were larger than those recorded from the contacts relatively far from the RNS contact (G4) in six patients for early latency period. The ratio of amplitudes of early latency period in GI/G4 significantly correlated with the outcome of RNS therapy, suggesting that the recording sites involved in the network for the epileptic activities would be important for the RNS therapy. 图 7 与 RNS 治疗有关的度内和度外 CCEPs 的解释图。(A) 在六名患者中，靠近 RNS 接触点（GI）的触点记录到的度内 CCEPs 在早期潜伏期比远离 RNS 接触点（G4）的触点记录到的 CCEPs 大。GI/G4 早期潜伏期振幅的比值与 RNS 治疗的结果明显相关，这表明癫痫活动网络中涉及的记录点对 RNS 治疗非常重要。
(B) The out-degree CCEPs by stimulating the sites with distances of to the closest RNS contacts (GI) presented more significant early latency period than those with distances more than (G3-4), implying that the CCEPs by stimulating the sites close to the RNS contacts reflect the cortical excitability at the site of stimulation and/or epileptic networks considering the pathological meanings of the early latency period of CCEPs. (B) 与距离最近的 RNS 接触点（ ）（GI）相比，刺激距离超过 （G3-4）的部位所产生的外度 CCEPs 的早期潜伏期更为显著，这意味着考虑到 CCEPs 早期潜伏期的病理意义，刺激靠近 RNS 接触点的部位所产生的 CCEPs 反映了刺激部位的皮层兴奋性和/或癫痫网络。
to those areas with ictal hyperperfusion in SPECT. These studies strongly suggest that CCEPs could serve as markers of cortical excitability in epileptic networks. In our study, the MTG/STG regions were uniformly sampled by SEEG in all 12 patients. For this reason, we could study the out-degree of this region across patients. There was no difference in significant out-degree CCEPs with either epilepsy syndrome or ictal SPECT hyperperfusion. Although ictal SPECT hyperfusion is likely a measure of and seizure propagation, it is not a measure of directional connectivity. However, the early latency period out-degree CCEPs were significant in 与 SPECT 中出现发作性高灌注的区域一致。 这些研究有力地表明，CCEPs 可作为癫痫网络中皮质兴奋性的标记。在我们的研究中，所有12名患者的MTG/STG区域均通过SEEG取样。因此，我们可以研究不同患者该区域的外度。无论是癫痫综合征还是发作期 SPECT 高灌注，CCEP 的显著外展度均无差异。虽然发作期 SPECT 高灌注可能是 和癫痫传播的一种测量方法， ，但它并不是定向连接的一种测量方法。然而，早期潜伏期外度CCEPs在以下情况中具有显著意义
MTG/STG only when the RNS electrodes were placed within of this region. This suggests that RNS electrode locations are targeted towards regions that serve as projection nodes within the brain. 只有当 RNS 电极放置在该区域 范围内时，MTG/STG 才会出现异常。这表明，RNS 电极的位置是针对作为脑内投射节点的区域的。
Summary 摘要
In this study, we employed measures of in-degree and outdegree CCEPs and showed that both in-degree and out-degree responses correlate with the distances between the contacts of 在这项研究中，我们采用了度内（in-degree）和度外（out-degree）CCEP 的测量方法，结果表明，度内和度外反应都与接触点之间的距离相关。
interest and the eventual RNS contact location. The in-degree CCEPs recorded from the contacts close to the eventual RNS contact tended to be larger compared with those recorded from the contacts further away. There were significant differences between ( from the closest RNS contact) and G4 ( ) in six out of 10 patients for CCEPs early latency period. The ratio of amplitudes of early latency period in G1/G4 showed a significant correlation with the outcome of RNS therapy. This is the first study showing correlations between RNS outcomes and effective connectivity measures using CCEPs (Fig. 7A) based on RNS electrode placement. We also demonstrated that locations planned by the clinical expert for RNS electrodes serve as receiver nodes within the brain. In a group analysis, the sites close to the eventual RNS contact location (G1 with distances of to the closest RNS contacts) showed a larger number of significant out-degree CCEPs during the early latency period compared to those regions further away from the eventual RNS contact location (G3-4 with distances > . Based on this finding, the location of RNS contacts could also reflect the cortical regions that serve as projection nodes to provide influence over wide regions of brain beyond the site of stimulation (Fig. 7B). These results of out-degree CCEPs do not show correlation with seizure outcome of RNS therapy but suggested that RNS electrodes might have been implanted in areas with higher out-degree to the MTG/STG. Additional analyses for out-degree CCEPs (Supplementary Material), although limited, showed a larger out-degree in patients with poor outcome (outcome scale 1), especially in the middle and late latency periods, for all three distance groups, excluding the late latency period in G2. In our previous study, we found that focal cortical dysplasia (FCD) type I, generally associated with worse epilepsy surgery outcomes, exhibited more widespread and pronounced out-degree CCEPs. We also found that in patients with FCD type II, which has a better outcome after epilepsy surgery, had a more temporally and spatially restricted out-degree of CCEPs, suggesting a more focal epilepsy. For our current study, these findings imply that patients with poor outcomes from RNS therapy may show a pattern akin to that observed in FCD type I, characterized by a larger epileptic network. Thus, we can conjecture that those patients who responded positively to RNS likely had more focal epilepsy as well. Furthermore, the association between out-degree CCEPs and RNS outcomes was only seen in the late period of CCEPs. The pathophysiology of generation of the late CCEPs response is postulated to involve cortico-thalamo-cortical connections. It may be possible that the extent of involvement of subcortical networks in a given epileptic network might also affect RNS outcomes. 和最终的 RNS 联系地点。与距离最终 RNS 联系点较远的联系点相比，距离最终 RNS 联系点较近的联系点所记录的 CCEP 的度内值往往较大。就 CCEP 早期潜伏期而言，10 名患者中有 6 名患者的 （距离最近的 RNS 接触点为 ）和 G4（ ）之间存在明显差异。G1/G4 早期潜伏期振幅的比值与 RNS 治疗的结果有明显的相关性。这是第一项基于 RNS 电极位置，利用 CCEPs（图 7A）显示 RNS 治疗结果与有效连通性测量之间相关性的研究。我们还证明，临床专家规划的 RNS 电极位置可作为大脑内的接收节点。在分组分析中，与距离最终 RNS 接触位置较远的区域（G3-4，与最近 RNS 接触点的距离 > ）相比，靠近最终 RNS 接触位置的区域（G1，与最近 RNS 接触点的距离为 ）在早期潜伏期显示出更多的显著度外 CCEP。基于这一发现，RNS 接触点的位置也可能反映了作为投射节点的皮层区域，这些区域对刺激部位以外的广泛脑区产生影响（图 7B）。这些度外 CCEPs 结果并未显示与 RNS 治疗的癫痫发作结果相关，但表明 RNS 电极可能植入了与 MTG/STG 度外较高的区域。额外的外度数 CCEPs 分析（补充材料）虽然有限，但显示在所有三个距离组中，结果不佳（结果量表 1）的患者的外度数较大，尤其是在潜伏期中期和晚期，但不包括 G2 的潜伏期晚期。在我们之前的研究中， ，我们发现局灶性皮质发育不良（FCD）I型通常与较差的癫痫手术预后相关，表现出更广泛和明显的CCEP外度。我们还发现，癫痫手术后疗效较好的 FCD II 型患者的 CCEPs 在时间和空间上更受限制，这表明他们患的是局灶性癫痫。就我们目前的研究而言，这些发现意味着 RNS 治疗效果不佳的患者可能会表现出与 FCD I 型患者类似的模式，其特点是癫痫网络更大。因此，我们可以推测，那些对 RNS 反应积极的患者可能也有更多的局灶性癫痫。此外，程度外 CCEP 与 RNS 结果之间的关联仅出现在 CCEP 的晚期。据推测，产生晚期 CCEPs 反应的病理生理学涉及皮质-眼球-皮质连接。 皮层下网络在特定癫痫网络中的参与程度可能也会影响 RNS 的结果。
As described, the RNS electrodes were mainly implanted close to the IOZ, and thus other clinical measures of the IOZ are important in consideration of the RNS electrode location. However, this study suggests that functional connectivity exhibited by CCEPs could also help guide the placement of electrodes for RNS therapy. Since CCEPs can visualize the interictal effective connectivity in individual patients, their use to determine the targets for the RNS therapy appears reasonable when patients undergo SEEG evaluation before RNS therapy. While invasive evaluations may not be necessary in all the patients who will undergo implantation of RNS electrodes, presurgical invasive evaluation including IOZ and CCEPs analyses could be considered to better localize the RNS contact locations. 如前所述，RNS 电极主要植入在靠近 IOZ 的位置，因此 IOZ 的其他临床测量对于考虑 RNS 电极的位置非常重要。然而，本研究表明，CCEPs 显示的功能连接性也有助于指导 RNS 治疗电极的位置。由于 CCEPs 可以直观地显示个别患者发作间期的有效连接，因此当患者在接受 RNS 治疗前接受 SEEG 评估时，使用 CCEPs 来确定 RNS 治疗的目标似乎是合理的。虽然并非所有接受 RNS 电极植入的患者都需要进行侵入性评估，但可以考虑在手术前进行侵入性评估，包括 IOZ 和 CCEPs 分析，以便更好地定位 RNS 接触位置。
Clinical implications and limitations 临床意义和局限性
We acknowledge that this is a retrospective study with its inherent limitations. We showed that only in-degree CCEPs were correlated with favourable outcome of RNS therapy. The number of patients was limited and we could not analyse the seizure outcomes relating to out-degree CCEPs as the regions of brain sampled and stimulated varied among patients. This study was limited to patients who required an invasive evaluation for epilepsy. Many patients receiving RNS therapy do so without the need for an invasive evaluation therefore the results of this study would not be pertinent to that group of patients. The placement of SEEG electrodes was driven by the hypothesis of the epileptogenic network based on non-invasive data. A number of additional brain regions were sampled that were outside the ictal onset and propagation networks, however there is an inherent spatial limitation using SEEG. The limited number of regions of brain sampled by CCEPs led us to perform a group analysis for the out-degree CCEPs since an individual analysis could not be performed. We focused on the out-degree CCEPs on only the MTG and STG to standardize the responses among patients, who had variable probable epileptogenic zones and CCEPs stimulus sites. In the out-degree CCEPs analysis, evoked potentials from MTG and STG would be expected to be higher in amplitude if the stimulation site were near this region compared to those stimulation sites that were far away due to direct orthodromic propagation within cortico-cortical pathways. This could be an alternative explanation for this finding rather than it being representative of a significant projection node in the epileptogenic network. For the in-degree CCEPs analyses, we included the SEEG contacts in both grey and white matters, while most of the previous CCEPs studies during SEEG evaluations focused only on cortical responses. The clinical interpretation and significance of CCEPs recorded from white matter have not been clarified. One study, using subdural rather than SEEG electrodes, did show that subcortico-cortical evoked potentials could be elicited by stimulating the white matter. Epileptic activity can affect the integrity of white matter, which is one proposed mechanism involved in the emergence of cognitive and psychiatric co-morbidities. Therefore, we considered that it would be reasonable to include all SEEG contacts for the in-degree CCEPs analyses. Another limitation is that rather than using connectivity analysis for analysing distance between contacts, we used Euclidian distance because we do not have the connectivity analyses of these patients, such as a connectivity map or diffusion tensor image (DTI) for all the patients. The results we have presented in the current study show correlations between CCEPs and 我们承认这是一项回顾性研究，有其固有的局限性。我们的研究表明，只有程度内 CCEP 与 RNS 治疗的良好结果相关。由于取样和刺激的脑部区域因人而异，我们无法分析与程度外 CCEPs 相关的癫痫发作结果。这项研究仅限于需要对癫痫进行侵入性评估的患者。许多接受 RNS 治疗的患者无需进行侵入性评估，因此本研究的结果与这部分患者无关。SEEG 电极的放置是基于非侵入性数据的致痫网络假设。研究人员对发作开始和传播网络之外的多个脑区进行了取样，但 SEEG 存在固有的空间限制。由于 CCEPs 取样的脑区数量有限，我们无法进行单个分析，因此对程度外 CCEPs 进行了分组分析。由于患者的可能致痫区和 CCEPs 刺激部位各不相同，因此我们仅对 MTG 和 STG 的外度 CCEPs 进行了重点分析，以统一患者的反应。在外度 CCEPs 分析中，如果刺激部位靠近 MTG 和 STG，那么由于皮质-皮质通路内的直接正交传播，与那些距离较远的刺激部位相比，MTG 和 STG 的诱发电位预计振幅会更高。这可能是这一发现的另一种解释，而不是它代表了致痫网络中的一个重要投射节点。在度内CCEPs分析中，我们将灰质和白质中的SEEG接触点都包括在内，而之前在SEEG评估期间进行的CCEPs研究大多只关注皮质反应。白质记录的 CCEPs 的临床解释和意义尚未明确。一项使用硬膜下电极而非 SEEG 电极的研究显示，刺激白质可诱发皮层下诱发电位。 癫痫活动会影响白质的完整性，而这正是导致认知和精神并发症的一个拟议机制。 因此，我们认为将所有 SEEG 接触点纳入程度内 CCEPs 分析是合理的。另一个局限是，我们没有使用连接分析来分析接触点之间的距离，而是使用了欧几里得距离，因为我们没有这些患者的连接分析，如所有患者的连接图或弥散张量图像（DTI）。我们在本次研究中提出的结果显示，CCEP 与
locations of RNS contacts likely reflecting the effective connectivity within the epileptogenic network. These measures were also correlated with favourable RNS treatment outcomes. The results of this study would indicate that CCEPs could provide additional insights in selecting appropriate stimulus sites for RNS therapy. Future investigations involving a larger number of patients and a greater number of CCEPs stimulation sites would be of interest and a prospective trial would be needed to clarify our findings in the future. RNS 接触点的位置可能反映了致痫网络内的有效连接。这些指标还与良好的 RNS 治疗结果相关。这项研究的结果表明，CCEPs 可以为选择合适的刺激点进行 RNS 治疗提供更多启示。未来涉及更多患者和更多 CCEPs 刺激点的研究将是有意义的，并且需要一项前瞻性试验来澄清我们的研究结果。
Conclusions 结论
Optimal electrode placement for RNS therapy is an important consideration in patients. In this study, evoked effective connectivity measures using CCEPs showed that the closer the distances of RNS contacts to the recording sites or stimulus sites of CCEPs, the larger the early latency in-degree CCEPs (receiver node) and significant early latency outdegree CCEPs (projection node). Electrophysiological characteristics of the epileptogenic network guiding the placement of RNS electrodes may very well play an important role in determining seizure outcomes with RNS therapy. 对患者而言，RNS 治疗的最佳电极位置是一个重要的考虑因素。在这项研究中，使用 CCEPs 进行的诱发有效连通性测量显示，RNS 接触点与 CCEPs 记录点或刺激点的距离越近，早期潜伏期内度 CCEPs（接收节点）和显著早期潜伏期外度 CCEPs（投射节点）就越大。指导 RNS 电极放置的致痫网络的电生理特征很可能在决定 RNS 治疗的癫痫发作结果方面发挥重要作用。
Supplementary material 补充材料
Supplementary material is available at Brain Communications online. 补充材料可在 "大脑通讯 "网上查阅。
Funding 资金筹措
Research reported in this publication was supported in part by the National Institutes of Health under awards R01 NS089212 (for D.R.N. and R.M.L.) and R01 NS074980 and R01 EB026299 (for R.M.L.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. 本刊物所报道的研究部分由美国国立卫生研究院（National Institutes of Health）的R01 NS089212（D.R.N.和R.M.L.）以及R01 NS074980和R01 EB026299（R.M.L.）资助。内容仅代表作者个人观点，不代表美国国立卫生研究院的官方观点。
Competing interests 竞争利益
The authors report no competing interests. 作者未报告任何利益冲突。
Data availability 数据可用性
The de-identified data in this study is available from the corresponding author upon reasonable request and completion of a formal data sharing agreement. 本研究中的去标识化数据可向通讯作者索取，但需提出合理要求并签订正式的数据共享协议。
References 参考资料
Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med. 2000;342(5):314-319. Kwan P, Brodie MJ.难治性癫痫的早期识别。N Engl J Med.2000;342(5):314-319.
Englot DJ, Chang EF. Rates and predictors of seizure freedom in resective epilepsy surgery: An update. Neurosurg Rev. 2014;37(3):389-404. Englot DJ, Chang EF.切除性癫痫手术中癫痫发作自由的比率和预测因素：更新。Neurosurg Rev. 2014;37(3):389-404.
Jehi L, Silveira DC, Bingaman W, Najm I. Temporal lobe epilepsy surgery failures: Predictors of seizure recurrence, yield of reevaluation, and outcome following reoperation. Neurosurg. 2010;113:1186-1194. Jehi L、Silveira DC、Bingaman W、Najm I. 颞叶癫痫手术失败：颞叶癫痫手术失败：癫痫复发的预测因素、重新评估的结果以及再次手术后的预后。 神经外科》，2010;113:1186-1194。
Wiebe S, Blume WT, Girvin JP, Eliasziw M; Effectiveness and Efficiency of Surgery for Temporal Lobe Epilepsy Study Group. A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med. 2001;345(5):311-318. Wiebe S、Blume WT、Girvin JP、Eliasziw M；颞叶癫痫手术治疗效果与效率研究小组。颞叶癫痫手术的随机对照试验。N Engl J Med.2001;345(5):311-318.
Engel J Jr, Wiebe S, French J, et al. Practice parameter: Temporal lobe and localized neocortical resections for epilepsy. Epilepsia. 2003;44(6):741-751. Engel Jr, Wiebe S, French J, et al：癫痫的颞叶和局部新皮质切除术。Epilepsia.2003;44(6):741-751.
Dwivedi R, Ramanujam B, Chandra PS, et al. Surgery for drug-resistant epilepsy in children. N Engl J Med. 2017;377: 1639-1647. Dwivedi R、Ramanujam B、Chandra PS 等：儿童耐药性癫痫的手术治疗。N Engl J Med.2017;377: 1639-1647.
Morrell MJ; RNS System in Epilepsy Study Group. Responsive cortical stimulation for the treatment of medically intractable partial epilepsy. Neurology. 2011;77(13):1295-1304. Morrell MJ；RNS 系统治疗癫痫研究小组。用于治疗药物难治性部分性癫痫的反应性皮层刺激。神经病学》。2011;77(13):1295-1304.
Geller EB. Responsive neurostimulation: Review of clinical trials and insights into focal epilepsy. Epilepsy Behav. 2018;88S:11-20. 盖勒 EB.反应性神经刺激：临床试验回顾及对局灶性癫痫的见解。癫痫行为。2018;88S:11-20.
Fisher R, Salanova V, Witt T, et al. Electrical stimulation of the anterior nucleus of thalamus for treatment of refractory epilepsy. Epilepsia. 2010;51(5):899-908. Fisher R, Salanova V, Witt T, et al. 丘脑前核电刺激治疗难治性癫痫。Epilepsia.2010;51(5):899-908.
Salanova V, Witt T, Worth R, et al. Long-term efficacy and safety of thalamic stimulation for drug-resistant partial epilepsy. Neurology. 2015;84(10):1017-1025. Salanova V、Witt T、Worth R 等：丘脑刺激治疗耐药部分性癫痫的长期疗效和安全性。神经病学》。2015;84(10):1017-1025.
Heck CN, King-Stephens D, Massey AD, et al. Two-year seizure reduction in adults with medically intractable partial onset epilepsy treated with responsive neurostimulation: Final results of the RNS System Pivotal trial. Epilepsia. 2014;55(3):432-441. Heck CN、King-Stephens D、Massey AD 等.接受反应性神经刺激治疗的药物难治性部分发作癫痫成人患者两年癫痫发作减少情况：RNS 系统关键性试验的最终结果。Epilepsia.2014;55(3):432-441.
Bergey GK, Morrell MJ, Mizrahi EM, et al. Long-term treatment with responsive brain stimulation in adults with refractory partial seizures. Neurology. 2015;84(8):810-817. Bergey GK、Morrell MJ、Mizrahi EM 等：对难治性部分性癫痫发作成人进行反应性脑部刺激的长期治疗。神经病学》。2015;84(8):810-817.
Nair DR, Laxer KD, Weber PB, et al. Nine-year prospective efficacy and safety of brain-responsive neurostimulation for focal epilepsy. Neurology. 2020;95(9):e1244-56. Nair DR、Laxer KD、Weber PB 等：脑响应神经刺激治疗局灶性癫痫的九年前瞻性疗效和安全性。神经病学》。2020;95(9):e1244-56.
Ma BB, Fields MC, Knowlton RC, et al. Responsive neurostimulation for regional neocortical epilepsy. Epilepsia. 2020;61(1): 96-106. Ma BB、Fields MC、Knowlton RC 等：针对区域性新皮质癫痫的反应性神经刺激。Epilepsia.2020;61(1):96-106.
Ma BB, Rao VR. Responsive neurostimulation: Candidates and considerations. Epilepsy Behav. 2018;88:388-395. Ma BB, Rao VR.反应性神经刺激：候选者和考虑因素。Epilepsy Behav.2018;88:388-395.
Stefan H, da Silva FH L. Epileptic neuronal networks: Methods of identification and clinical relevance. Front Neurol. 2013;4:8. Stefan H, da Silva FH L. 癫痫神经元网络：识别方法和临床意义。前沿神经学》。2013;4:8.
van Diessen E, Diederen SJ, Braun KP, Jansen FE, Stam CJ. Functional and structural brain networks in epilepsy: What have we learned? Epilepsia. 2013;54:1855-1865. van Diessen E, Diederen SJ, Braun KP, Jansen FE, Stam CJ.癫痫的大脑功能和结构网络：我们学到了什么？Epilepsia.2013;54:1855-1865.
Bartolomei F, Lagarde S, Wendling F, et al. Defining epileptogenic networks: Contribution of SEEG and signal analysis. Epilepsia. 2017;58:1131-1147. Bartolomei F, Lagarde S, Wendling F, et al：SEEG 和信号分析的贡献。Epilepsia.2017;58:1131-1147.
Scheid BH, Bernabei JM, Khambhati AN, et al. Intracranial electroencephalographic biomarker predicts effective responsive neurostimulation for epilepsy prior to treatment. Epilepsia. 2022;63(3): 652-662. Scheid BH、Bernabei JM、Khambhati AN 等：颅内脑电图生物标志物可预测癫痫治疗前有效的神经刺激反应。Epilepsia.2022;63(3):652-662.
Fan JM, Lee AT, Kudo K, et al. Network connectivity predicts effectiveness of responsive neurostimulation in focal epilepsy. Brain Сотmиn. 2022;4(3)::сас104. Fan JM、Lee AT、Kudo K 等：网络连通性预测局灶性癫痫中响应性神经刺激的有效性。Brain Сотmиn.2022;4(3)::сас104.
Charlebois CM, Anderson DN, Johnson KA, et al. Patient-specific structural connectivity informs outcomes of responsive neurostimulation for temporal lobe epilepsy. Epilepsia. 2022;63(8):2037-2055. Charlebois CM、Anderson DN、Johnson KA 等：颞叶癫痫反应性神经刺激的患者特异性结构连接性结果。Epilepsia.2022;63(8):2037-2055.
Khambhati AN, Shafi A, Rao VR, Chang EF. Long-term brain network reorganization predicts responsive neurostimulation outcomes for focal epilepsy. Sci Transl Med. 2021;13(608):eabf6588. Khambhati AN、Shafi A、Rao VR、Chang EF。长期脑网络重组可预测局灶性癫痫的神经刺激疗效。Sci Transl Med.2021;13(608):eabf6588.
Ponten SC, Bartolomei F, Stam CJ. Small-world networks and epilepsy: Graph theoretical analysis of intracerebrally recorded mesial temporal lobe seizures. Clin Neurophysiol. 2007;118:918-927. Ponten SC, Bartolomei F, Stam CJ.小世界网络与癫痫：脑内记录颞叶中段癫痫发作的图论分析。Clin Neurophysiol.
Reijneveld JC, Ponten SC, Berendse HW, Stam CJ. The application of graph theoretical analysis to complex networks in the brain. Clin Neurophysiol. 2007;118:2317-2331. Reijneveld JC, Ponten SC, Berendse HW, Stam CJ.图论分析在大脑复杂网络中的应用。Clin Neurophysiol.
Bullmore E, Sporns O. Complex brain networks: Graph theoretical analysis of structural and functional systems. Nat Rev Neurosci. 2009;10:186-198. Bullmore E, Sporns O. 复杂大脑网络：结构和功能系统的图论分析。Nat Rev Neurosci.2009;10:186-198.
Stam CJ, van Straaten EC. Go with the flow: Use of a directed phase lag index (dPLI) to characterize patterns of phase relations in a large-scale model of brain dynamics. Neuroimage. 2012;62: 1415-1428. Stam CJ, van Straaten EC.随波逐流：使用定向相位滞后指数（dPLI）描述大规模大脑动力学模型中的相位关系模式。神经影像。2012;62: 1415-1428.
Bandt SK, Besson P, Ridley B, et al. Connectivity strength, time lag structure and the epilepsy network in resting-state fMRI. Neuroimage Clin. 2019;24:102035. Bandt SK、Besson P、Ridley B 等：静息态 fMRI 中的连接强度、时滞结构和癫痫网络。神经影像临床》。2019;24:102035.
Boerwinkle VL, Mirea L, Gaillard WD, et al. Resting-state functional MRI connectivity impact on epilepsy surgery plan and surgical candidacy: Prospective clinical work. J Neurosurg Pediatr. 2020; 20:1-8. Boerwinkle VL、Mirea L、Gaillard WD 等：静息态功能磁共振成像连接性对癫痫手术计划和手术候选资格的影响：前瞻性临床工作。J Neurosurg Pediatr.
Ge R, Downar J, Blumberger DM, Daskalakis ZJ, Vila-Rodriguez F. Functional connectivity of the anterior cingulate cortex predicts treatment outcome for rTMS in treatment-resistant depression at 3-month follow-up. Brain Stimul. 2020;13:206-214. Ge R、Downar J、Blumberger DM、Daskalakis ZJ、Vila-Rodriguez F. 前扣带回皮层的功能连接可预测经颅磁刺激治疗耐药抑郁症 3 个月随访的疗效。2020;13:206-214.
Friston KJ. Functional and effective connectivity: A review. Brain Connect. 2011;1:13-36. Friston KJ.功能和有效连接：综述。Brain Connect.2011;1:13-36.
Yaffe RB, Borger P, Megevand P, et al. Physiology of functional and effective networks in epilepsy. Clin Neurophysiol. 2015;126: 227-236. Yaffe RB, Borger P, Megevand P, et al. 癫痫的功能和有效网络生理学。Clin Neurophysiol.
Adrian ED. The spread of activity in the cerebral cortex. J Physiol. 1936;88:127-161. Adrian ED.大脑皮层活动的传播。J Physiol.
Matsumoto R, Nair DR, LaPresto E, et al. Functional connectivity in the human language system: A cortico-cortical evoked potential study. Brain. 2004;127(10):2316-2330. Matsumoto R, Nair DR, LaPresto E, et al. 人类语言系统的功能连接：皮质诱发电位研究。脑。2004;127(10):2316-2330.
Matsumoto R, Nair DR, LaPresto E, Bingaman W, Shibasaki H, Lüders HO. Functional connectivity in human cortical motor system: A cortico-cortical evoked potential study. Brain. 2007; 130(1):181-197. Matsumoto R, Nair DR, LaPresto E, Bingaman W, Shibasaki H, Lüders HO.人类皮层运动系统的功能连接：皮质-皮质诱发电位研究。脑。2007; 130(1):181-197.
Lacruz ME, García Seoane JJ, Valentin A, Selway R, Alarcón G. Frontal and temporal functional connections of the living human brain. Eur J Neurosci. 2007;26(5):1357-1370. Lacruz ME, García Seoane JJ, Valentin A, Selway R, Alarcón G. 活体人脑的额叶和颞叶功能连接。Eur J Neurosci.2007;26(5):1357-1370.
Greenlee JD, Oya H, Kawasaki H, et al. Functional connections within the human inferior frontal gyrus. J Comp Neurol. 2007; 503(4):550-559. Greenlee JD, Oya H, Kawasaki H, et al. 人类额叶下回的功能连接。J Comp Neurol.2007; 503(4):550-559.
Matsumoto R, Kinoshita M, Taki J, et al. In vivo epileptogenicity of focal cortical dysplasia: A direct cortical paired stimulation study. Epilepsia. 2005;46(11):1744-1749. Matsumoto R, Kinoshita M, Taki J, et al. 局灶性皮质发育不良的体内致痫性：皮质直接配对刺激研究。Epilepsia.2005;46(11):1744-1749.
Iwasaki M, Enatsu R, Matsumoto R, et al. Accentuated corticocortical evoked potentials in neocortical epilepsy in areas of ictal onset. Epileptic Disord. 2010;12(4):292-302. Iwasaki M, Enatsu R, Matsumoto R, et al. 新皮质癫痫发作区皮质诱发电位增强。Epileptic Disord.2010;12(4):292-302.
Enatsu R, Jin K, Elwan S, et al. Correlations between ictal propagation and response to electrical cortical stimulation: A cortico-cortical evoked potential study. Epilepsy Res. 2012; 101(1-2):76-87. Enatsu R、Jin K、Elwan S 等人.发作传播与皮质电刺激反应之间的相关性：皮质诱发电位研究。2012; 101(1-2):76-87.
Tousseyn S, Krishnan B, Wang ZI, et al. Connectivity in ictal single photon emission computed tomography perfusion: A corticocortical evoked potential study. Brain. 2017;140(7):1872-1884. Tousseyn S、Krishnan B、Wang ZI 等：发作期单光子发射计算机断层扫描灌注的连接性：皮质诱发电位研究。Brain.2017;140(7):1872-1884.
Zhang N, Zhang B, Rajah GB, et al. The effectiveness of corticocortical evoked potential in detecting seizure onset zones. Neurol Res. 2018;40:480-490. Zhang N, Zhang B, Rajah GB, et al. 皮层诱发电位检测癫痫发作起始区的有效性。Neurol Res. 2018;40:480-490.
Guo ZH, Zhao BT, Toprani S, et al. Epileptogenic network of focal epilepsies mapped with cortico-cortical evoked potentials. Clin Neurophysiol. 2020;131:2657-2666. Guo ZH，Zhao BT，Toprani S，et al.临床神经生理学》，2020;131:2657-2666。
Shahabi H, Taylor K, Hirfanoglu T, et al. Effective connectivity differs between focal cortical dysplasia types I and II. Epilepsia. 2021; 62:2753-2765 Shahabi H, Taylor K, Hirfanoglu T, et al. 局灶性皮质发育不良 I 型和 II 型的有效连接性不同。Epilepsia.2021; 62:2753-2765
van Blooijs D, Leijten FSS, van Rijen PC, Meijer HGE, Huiskamp GJM. Evoked directional network characteristics of epileptogenic tissue derived from single pulse electrical stimulation. Hum Brain Марр. 2018;39(11):4611-4622. van Blooijs D, Leijten FSS, van Rijen PC, Meijer HGE, Huiskamp GJM.单脉冲电刺激致痫组织的诱发定向网络特征。Hum Brain Марр.2018;39(11):4611-4622.
Keller CJ, Honey CJ, Entz L, et al. Corticocortical evoked potentials reveal projectors and integrators in human brain networks. J Neurosci. 2014;34(27):9152-9163. Keller CJ, Honey CJ, Entz L, et al. 皮层诱发电位揭示了人脑网络中的投射器和整合器。J Neurosci.2014;34(27):9152-9163.
Keller CJ, Honey CJ, Mégevand P, Entz L, Ulbert I, Mehta AD. Mapping human brain networks with cortico-cortical evoked potentials. Philos Trans R Soc Lond B Biol Sci. 2014;369(1653): 20130528. Keller CJ, Honey CJ, Mégevand P, Entz L, Ulbert I, Mehta AD.用皮质诱发电位绘制人脑网络图。Philos Trans R Soc Lond B Biol Sci：20130528.
Gonzalez-Martinez J, Mullin J, Vadera S, et al. Stereotactic placement of depth electrodes in medically intractable epilepsy. J Neurosurg. 2014;120:639-644. Gonzalez-Martinez J, Mullin J, Vadera S, et al. 医学难治性癫痫的立体定向深度电极放置。J Neurosurg.
Enatsu R, Bulacio J, Nair DR, Bingaman W, Najm I, Gonzalez-Martinez J. Posterior cingulate epilepsy: Clinical and neurophysiological analysis. J Neurol Neurosurg Psychiatry. 2014;85(1):44-50. Enatsu R、Bulacio J、Nair DR、Bingaman W、Najm I、Gonzalez-Martinez J. 后扣带回癫痫：临床和神经生理学分析。J Neurol Neurosurg Psychiatry.2014;85(1):44-50.
Taylor K, Joshi AA, Hirfanoglu T, et al. Validation of semiautomated anatomically labeled SEEG contacts in a brain atlas for mapping connectivity in focal epilepsy. Epilepsia Open. 2021;6: 493-503. Taylor K, Joshi AA, Hirfanoglu T, et al. 半自动解剖标记的 SEEG 接触点在脑图谱中的验证，用于绘制局灶性癫痫的连接图。Epilepsia Open.2021;6: 493-503.
Shattuck DW, Leahy RM. BrainSuite: An automated cortical surface identification tool. Med Image Anal. 2002;6(2):129-142. Shattuck DW, Leahy RM.BrainSuite：皮质表面自动识别工具。医学图像分析。2002;6(2):129-142.
Joshi AA, Choi S, Liu Y, et al. A hybrid high-resolution anatomical MRI atlas with sub-parcellation of cortical gyri using resting fMRI. J Neurosci Methods. 2022;374:109566. Joshi AA, Choi S, Liu Y, et al. 利用静息 fMRI 对皮层回进行子划分的混合高分辨率解剖 MRI 图谱。J Neurosci Methods.2022;374:109566.
Taylor K, Joshi AA, Li J, et al. The FAST graph: A novel framework for the anatomically-guided visualization and analysis of cortico-cortical evoked potentials. Epilepsy Res. 2020;161: 106264. Taylor K, Joshi AA, Li J, et al：用于解剖学引导的皮质诱发电位可视化和分析的新框架。癫痫研究》，2020; 161: 106264。
Entz L, Tóth E, Keller CJ, et al. Evoked effective connectivity of the human neocortex. Hum Brain Mapp. 2014;35(12): 5736-5753. Entz L, Tóth E, Keller CJ, et al. 人类新皮层的诱发有效连接。人类脑图。2014;35(12):5736-5753.
Prime D, Rowlands D, O'Keefe S, Dionisio S. Considerations in performing and analyzing the responses of cortico-cortical evoked potentials in stereo-EEG. Epilepsia. 2018;59(1):16-26. Prime D、Rowlands D、O'Keefe S、Dionisio S.在立体电子脑电图中执行和分析皮质皮质诱发电位反应的考虑因素。Epilepsia.2018;59(1):16-26.
Dickey AS, Alwaki A, Kheder A, Willie JT, Drane DL, Pedersen NP. The referential montage inadequately localizes corticocortical evoked potentials in stereoelectroencephalography. J Clin Neurophysiol. 2022;39(5):412-418. Dickey AS, Alwaki A, Kheder A, Willie JT, Drane DL, Pedersen NP.立体脑电图中皮质诱发电位的参考蒙太奇定位不足。J Clin Neurophysiol.
Engel J Jr, Van Ness PC, Rasmussen TB, Ojemann LM. Outcome with respect to epileptic seizures. In: Engel J Jr, ed. Surgical treatment of the epilepsies. Raven Press; 1993:609-621. Engel J Jr、Van Ness PC、Rasmussen TB、Ojemann LM。与癫痫发作有关的结果。In：Engel J Jr, ed..癫痫的外科治疗。Raven Press; 1993:609-621.
Wieser HG, Blume WT, Fish D, et al. Proposal for a new classification of outcome with respect to epileptic seizures following epilepsy surgery. Epilepsia. 2001;42(2):282-286. Wieser HG、Blume WT、Fish D 等：《关于癫痫手术后癫痫发作结果新分类的建议》。Epilepsia.2001;42(2):282-286.
Pantazis D, Fang M, Qin S, Mohsenzadeh Y, Li Q, Leahy RM. Decoding the orientation of contrast edges from MEG evoked and induced responses. Neuroimage. 2018;180(Pt A):267-279. Pantazis D, Fang M, Qin S, Mohsenzadeh Y, Li Q, Leahy RM.从 MEG 诱发和诱导反应解码对比边缘的方向。Neuroimage.2018;180(Pt A):267-279.
Thomas GP, Jobst BC. Critical review of the responsive neurostimulator system for epilepsy. Med Devices (Auckl). 2015;8:405-411. Thomas GP, Jobst BC.癫痫反应性神经刺激器系统点评。Med Devices (Auckl).2015;8:405-411.
Durand D. Electrical stimulation can inhibit synchronized neuronal activity. Brain Res. 1986;382(1):139-144. Durand D. 电刺激可抑制同步神经元活动。脑研究，1986；382（1）：139-144。
Motamedi GK, Lesser RP, Miglioretti DL, et al. Optimizing parameters for terminating cortical afterdischarges with pulse stimulation. Epilepsia. 2002;43(8):836-846. Motamedi GK、Lesser RP、Miglioretti DL 等人.用脉冲刺激终止皮质后放电的优化参数.Epilepsia.2002;43(8):836-846.
Sohal VS, Sun FT. Responsive neurostimulation suppresses synchronized cortical rhythms in patients with epilepsy. Neurosurg Clin N Am. 2011;22(4):481-488. Sohal VS, Sun FT.反应性神经刺激抑制癫痫患者的同步皮质节律。Neurosurg Clin N Am.2011;22(4):481-488.
Rønborg SN, Esteller R, Tcheng TK, et al. Acute effects of brainresponsive neurostimulation in drug-resistant partial onset epilepsy. Clin Neurophysiol. 2021;132(6):1209-1220. Rønborg SN、Esteller R、Tcheng TK 等：脑响应神经刺激对耐药部分发作性癫痫的急性影响。临床神经生理学》，2021;132(6):1209-1220。
Liang F, Isackson PJ, Jones EG. Stimulus-dependent, reciprocal up- and downregulation of glutamic acid decarboxylase and calmodulin-dependent protein kinase II gene expression in rat cerebral cortex. Exp Brain Res. 1996;110:163-174. Liang F, Isackson PJ, Jones EG.大鼠大脑皮层中谷氨酸脱羧酶和 钙调素依赖性蛋白激酶 II 基因表达的刺激依赖性相互上调和下调。Exp Brain Res. 1996;110:163-174.
Yu T, Wang X, Li Y, et al. High-frequency stimulation of anterior nucleus of thalamus desynchronizes epileptic network in humans. Brain. 2018;141:2631-2643. Yu T, Wang X, Li Y, et al. 丘脑前核的高频刺激可使人类癫痫网络失同步。Brain.2018;141:2631-2643.
Kokkinos V, Sisterson ND, Wozny TA, Richardson RM. Association of closed-loop brain stimulation neurophysiological features with seizure control among patients with focal epilepsy. JAMA Neurol. 2019;76:800-808. Kokkinos V、Sisterson ND、Wozny TA、Richardson RM。闭环脑刺激神经生理学特征与局灶性癫痫患者发作控制的关系。JAMA Neurol.2019;76:800-808.
Penfield W, Jasper H. Electrocorticography, In: Epilepsy and the functional anatomy of the human brain. Little Brown; 1954:363-365. Penfield W, Jasper H. Electrocorticography, In：癫痫与人脑功能解剖学》。Little Brown; 1954:363-365.
Jefferys JG. Influence of electric fields on the excitability of granule cells in guinea-pig hippocampal slices. J Physiol. 1981;319: 143-152. Jefferys JG.电场对豚鼠海马切片颗粒细胞兴奋性的影响J Physiol.
Velasco M, Velasco F, Velasco AL, et al. Subacute electrical stimulation of the hippocampus blocks intractable temporal lobe seizures and paroxysmal EEG activities. Epilepsia. 2000;41(2): 158-169. Velasco M, Velasco F, Velasco AL, et al. 亚急性海马电刺激可阻断顽固性颞叶癫痫发作和阵发性脑电图活动。Epilepsia.2000;41(2):158-169.
Kinoshita M, Ikeda A, Matsumoto R, et al. Electric stimulation on human cortex suppresses fast cortical activity and epileptic spikes. Epilepsia. 2004;45(7):787-791. Kinoshita M, Ikeda A, Matsumoto R, et al. 电刺激人体皮层可抑制快速皮层活动和癫痫尖峰。Epilepsia.2004;45(7):787-791.
Kinoshita M, Ikeda A, Matsuhashi M, et al. Electric cortical stimulation suppresses epileptic and background activities in neocortical epilepsy and mesial temporal lobe epilepsy. Clin Neurophysiol. 2005;116(6):1291-1299. Kinoshita M, Ikeda A, Matsuhashi M, et al.皮层电刺激抑制新皮层癫痫和颞叶中叶癫痫的癫痫活动和背景活动。临床神经生理学》，2005；116（6）：1291-1299。
Lundstrom BN, Van Gompel J, Britton J, et al. Chronic subthreshold cortical stimulation to treat focal epilepsy. JAMA Neurol. 2016;73(11):1370-1372. Lundstrom BN、Van Gompel J、Britton J 等：慢性阈下皮层刺激治疗局灶性癫痫。JAMA Neurol.2016;73(11):1370-1372.
Kerezoudis P, Grewal SS, Stead M, et al. Chronic subthreshold cortical stimulation for adult drug-resistant focal epilepsy: Safety, feasibility, and technique. J Neurosurg. 2018;129(2):533-543. Kerezoudis P、Grewal SS、Stead M 等：慢性阈下皮层刺激治疗成人耐药局灶性癫痫：安全性、可行性和技术J Neurosurg. 2018;129(2):533-543.
Chiang S, Khambhati AN, Wang ET, Vannucci M, Chang EF, Rao VR. Evidence of state-dependence in the effectiveness of responsive neurostimulation for seizure modulation. Brain Stimul. 2021;14(2):366-375. Chiang S, Khambhati AN, Wang ET, Vannucci M, Chang EF, Rao VR.反应性神经刺激对癫痫发作调节效果的状态依赖性证据。2021;14(2):366-375.
Yamamoto J, Ikeda A, Satow T, et al. Low-frequency electric cortical stimulation has an inhibitory effect on epileptic focus in mesial temporal lobe epilepsy. Epilepsia. 2002;43(5):491-495. Yamamoto J, Ikeda A, Satow T, et al. 低频皮质电刺激对颞叶中叶癫痫的癫痫灶有抑制作用。Epilepsia.2002;43(5):491-495.
Yamamoto J, Ikeda A, Kinoshita M, et al. Low-frequency electric cortical stimulation decreases interictal and ictal activity in human epilepsy. Seizure. 2006;15(7):520-527. Yamamoto J, Ikeda A, Kinoshita M, et al. Low-frequency electric cortical stimulation decreases interictal and ictal activity in human epilepsy.Seizure.2006;15(7):520-527.
Khambhati AN, Kahn AE, Costantini J, et al. Functional control of electrophysiological network architecture using direct neurostimulation in humans. Netw Neurosci. 2019;3(3):848-877. Khambhati AN、Kahn AE、Costantini J 等人利用直接神经刺激对电生理网络结构进行功能控制。Netw Neurosci.2019;3(3):848-877.
Lega B, Dionisio S, Flanigan P, et al. Cortico-cortical evoked potentials for sites of early versus late seizure spread in stereoelectroencephalography. Epilepsy Res. 2015;115:17-29. Lega B, Dionisio S, Flanigan P, et al. 立体脑电图中发作早期与晚期扩散部位的皮质皮层诱发电位。癫痫研究》，2015;115:17-29。
Van Paesschen W. Ictal SPECT. Epilepsia. 2004;45(Suppl 4):35-40. Van Paesschen W. Ictal SPECT.Epilepsia.2004;45(Suppl 4):35-40.
Yamao Y, Matsumoto R, Kunieda T, et al. Intraoperative dorsal language network mapping by using single-pulse electrical stimulation. Hum Brain Mapp. 2014;35(9):4345-4361. Yamao Y, Matsumoto R, Kunieda T, et al. 利用单脉冲电刺激绘制术中背侧语言网络图。Hum Brain Mapp.2014;35(9):4345-4361.
Hatton SN, Huynh KH, Bonilha L, et al. White matter abnormalities across different epilepsy syndromes in adults: An ENIGMA-Epilepsy study. Brain. 2020;143:2454-2473. Hatton SN、Huynh KH、Bonilha L 等：成人不同癫痫综合征的白质异常：ENIGMA-癫痫研究。脑。2020;143:2454-2473.
Received September 12, 2022. Revised September 06, 2023. Accepted February 19, 2024. Advance access publication February 22, 2024 2022 年 9 月 12 日收到。2023 年 9 月 6 日修订。2024 年 2 月 19 日接受。2024 年 2 月 22 日提前获取出版
(C) The Author(s) 2024. Published by Oxford University Press on behalf of the Guarantors of Brain. (C) 作者 2024 年。由牛津大学出版社代表大脑担保人出版。
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 这是一篇根据知识共享署名许可协议 ( https://creativecommons.org/licenses/by/4.0/) 条款发布的开放获取文章，该协议允许在适当引用原作的前提下，在任何媒体上不受限制地再利用、传播和复制。
