Clinical Neuroscience Center, The Seventh Affiliated Hospital of Sun Yat-sen University, Shenzhen, China.
Guangzhou Laboratory, Guangzhou, China.
Epilepsia. 2023 Jun;64(6):1541-1553. doi: 10.1111/epi.17587. Epub 2023 Mar 30.
We aim to determine whether automatically detected ripple rate (ADRR) of 10-min scalp electroencephalography (EEG) during slow-wave sleep can be a useful tool for rapid epilepsy differentiation and seizure activity assessment, and we analyze the clinical factors that may affect the scalp ripple rates.
We retrospectively included 336 patients who underwent long-term video-EEG with a sampling rate ≥1000 Hz, and three groups were established based on their final clinical diagnosis (non-epilepsy; non-active epilepsy [epilepsy being seizure-free for at least 1 year]; and active epilepsy [epilepsy with one or more seizures in the past year]). ADRRs between groups were compared and diagnostic thresholds set according to the maximum of Youden index with the receiver-operating characteristic curve.
The 336 patients comprised 49 non-epilepsy and 287 epilepsy patients (95 non-active epilepsy and 192 active epilepsy). The median ADRR of the epilepsy group was significantly greater than in the non-epilepsy group, with a diagnostic threshold of 4.25 /min (specificity 89.8%, sensitivity 47.74%, p<.001). Following stratification by age, the area under the curve was greatest in the 0-20 year subgroup, threshold 4.10 /min (specificity 100%, sensitivity 52.47%, p<.001). Regarding distinguishing active epilepsy from non-active epilepsy patients, the area under the curve was also greatest in patients 0-20 years of age, threshold 13.05/min (specificity 98.36%, sensitivity 35.64%, p<.001). Following stratification by epilepsy type, the diagnostic efficiency was best in children with developmental and epileptic encephalopathies/epileptic encephalopathies (DEEs/EEs) (threshold 5.20/min, specificity 100%, sensitivity 100%) and self-limited focal epilepsies (SeLFEs) (threshold 5.45/min, specificity 80%, sensitivity 100%). Multivariate analysis revealed that the influential factors of ADRRs were age, depth of epileptogenic lesion, and seizure frequency.
ADRR of scalp EEG can be a rapid and specific method to differentiate epilepsy and evaluate seizure activity. This method is especially suitable for young patients.
我们旨在确定 10 分钟头皮脑电图(EEG)慢波睡眠期间自动检测到的涟漪率(ADRR)是否可作为快速鉴别癫痫和评估发作活动的有用工具,并分析可能影响头皮涟漪率的临床因素。
我们回顾性纳入了 336 例接受了采样率≥1000Hz 的长程视频-EEG 检查的患者,并根据最终临床诊断将其分为三组(非癫痫;非活动性癫痫[癫痫无发作至少 1 年];活动性癫痫[过去 1 年中发作 1 次或多次])。比较各组之间的 ADRR,并根据接收者操作特征曲线的最大 Youden 指数设定诊断阈值。
336 例患者中包括 49 例非癫痫和 287 例癫痫患者(95 例非活动性癫痫和 192 例活动性癫痫)。癫痫组的 ADRR 中位数明显大于非癫痫组,诊断阈值为 4.25/min(特异性 89.8%,敏感性 47.74%,<.001)。按年龄分层后,0-20 岁亚组的曲线下面积最大,阈值为 4.10/min(特异性 100%,敏感性 52.47%,<.001)。对于区分活动性癫痫与非活动性癫痫患者,0-20 岁患者的曲线下面积也最大,阈值为 13.05/min(特异性 98.36%,敏感性 35.64%,<.001)。按癫痫类型分层后,发育性和癫痫性脑病/癫痫性脑病(DEEs/EEs)和自限性局灶性癫痫(SeLFEs)儿童的诊断效率最佳(阈值分别为 5.20/min、特异性 100%、敏感性 100%和阈值为 5.45/min、特异性 80%、敏感性 100%)。多变量分析显示,ADRR 的影响因素为年龄、致痫病灶深度和发作频率。
头皮 EEG 的 ADRR 可以是一种快速且特异的鉴别癫痫和评估发作活动的方法。该方法特别适用于年轻患者。