Developmental Imaging and Biophysics Section, UCL Great Ormond Street Institute of Child Health, University College London, London, United Kingdom.
Epilepsy Unit, Department of Neurophysiology, Great Ormond Street Hospital, London, United Kingdom.
Ann Neurol. 2017 Aug;82(2):278-287. doi: 10.1002/ana.25003. Epub 2017 Aug 9.
Surgical treatment in epilepsy is effective if the epileptogenic zone (EZ) can be correctly localized and characterized. Here we use simultaneous electroencephalography-functional magnetic resonance imaging (EEG-fMRI) data to derive EEG-fMRI and electrical source imaging (ESI) maps. Their yield and their individual and combined ability to (1) localize the EZ and (2) predict seizure outcome were then evaluated.
Fifty-three children with drug-resistant epilepsy underwent EEG-fMRI. Interictal discharges were mapped using both EEG-fMRI hemodynamic responses and ESI. A single localization was derived from each individual test (EEG-fMRI global maxima [GM]/ESI maximum) and from the combination of both maps (EEG-fMRI/ESI spatial intersection). To determine the localization accuracy and its predictive performance, the individual and combined test localizations were compared to the presumed EZ and to the postsurgical outcome.
Fifty-two of 53 patients had significant maps: 47 of 53 for EEG-fMRI, 44 of 53 for ESI, and 34 of 53 for both. The EZ was well characterized in 29 patients; 26 had an EEG-fMRI GM localization that was correct in 11, 22 patients had ESI localization that was correct in 17, and 12 patients had combined EEG-fMRI and ESI that was correct in 11. Seizure outcome following resection was correctly predicted by EEG-fMRI GM in 8 of 20 patients, and by the ESI maximum in 13 of 16. The combined EEG-fMRI/ESI region entirely predicted outcome in 9 of 9 patients, including 3 with no lesion visible on MRI.
EEG-fMRI combined with ESI provides a simple unbiased localization that may predict surgery better than each individual test, including in MRI-negative patients. Ann Neurol 2017;82:278-287.
如果能正确定位和描述致痫区（EZ），则癫痫的外科治疗是有效的。在这里，我们使用同步脑电图-功能磁共振成像（EEG-fMRI）数据来获得 EEG-fMRI 和电源成像（ESI）图。然后评估它们的产量及其单独和联合（1）定位 EZ 和（2）预测癫痫发作结果的能力。
53 名耐药性癫痫儿童接受了 EEG-fMRI 检查。使用 EEG-fMRI 血流动力学反应和 ESI 对间期放电进行了定位。从每个个体测试（EEG-fMRI 全局最大值[GM]/ESI 最大值）和两个地图的组合（EEG-fMRI/ESI 空间交点）得出一个单一的定位。为了确定定位精度及其预测性能，将个体和组合测试的定位与假定的 EZ 和术后结果进行了比较。
53 例患者中有 52 例存在显著的图：47 例 EEG-fMRI，44 例 ESI，34 例两者均有。29 例患者的 EZ 特征良好；11 例 EEG-fMRI GM 定位正确的患者中有 26 例，17 例 ESI 定位正确的患者中有 22 例，11 例 EEG-fMRI 和 ESI 联合定位正确的患者中有 12 例。20 例患者中，EEG-fMRI GM 正确预测了 8 例癫痫发作结果，ESI 最大值正确预测了 13 例。9 例患者的 EEG-fMRI/ESI 联合区域完全预测了结果，其中 3 例 MRI 未见病变。
EEG-fMRI 联合 ESI 提供了一种简单的无偏置定位，其预测手术效果可能优于每个单独的测试，包括 MRI 阴性患者。
