局灶性癫痫发作的脑电图特征及其与患者耐药的关系

Electroencephalography characteristics of focal epileptic seizures and their association with drug resistance

摘要:

目的 探讨局灶性癫痫发作的脑电图特征及其与患者耐药的关系。方法 纳入2020年1月至2023年4月北京市隆福医院收治的86例局灶性癫痫患者作为研究对象，年龄20～80岁，平均（59.37±5.63）岁。将患者分为耐药组（52例）和良好组（34例），另纳入45名健康志愿者作为对照组。观察癫痫发作的脑电图（EEG）波形特征和癫痫发作不同阶段的时间分布和线性特征。利用皮尔逊相关系数评价EEG导联间的同步程度。对比3组基线资料和各频段EEG网络参数，绘制脑网络P值连接图，分析EEG与患者耐药性的关系。结果 在0.5～60 Hz采样率下，癫痫患者EEG棘波最为典型；在250 Hz采样率下，癫痫发作间期、发作前期及发作期各采样190 s的EEG时域波形，发现棘波数随着癫痫发作状态的过渡呈现动态阶段性变化趋势，随着癫痫发作的推进逐渐增多。局灶性癫痫发作间期和发作期的导联间同步性存在显著差异，与发作间期相比，发作期强相关的通道数量增多。将对照组、耐药组和良好组患者基线资料进行对比，耐药组患者病程、发病频率及既往用药种类数高于良好组（P<0.05）。在θ频段和β频段，3组的聚类系数、特征路径长度、全局效率及局部效率差异均具有统计学意义（P<0.05）；在α频段，3组间特征路径长度和全局效率的差异具有统计学意义（P<0.05）。良好组和耐药组各频段，前头部及颞部耐药组短程连接增强。耐药组和对照组之间，β频段前头部耐药组长程连接增强，δ频段、θ频段和α频段前头部耐药组短程连接增强，各频段后头部耐药组短程连接均减弱。良好组和对照组比较，δ频段前头部良好组长程连接增强，各频段后头部良好组短程连接均减弱。结论 EEG信号特征分析可有效识别癫痫发作阶段。相较于对照组和良好组患者，耐药组患者的脑网络属性及拓扑结构发生了明显改变。耐药组患者的前头部脑网络连接增强。EEG网络研究可以帮助发现局灶性癫痫的耐药性。

Abstract:

Objective To investigate the electroencephalography （EEG） characteristics of focal epileptic seizures and their association with drug resistance.Methods A total of 86 patients with focal epilepsy who were admitted to Beijing Longfu Hospital from January 2020 to April 2023 were enrolled as subjects， with an age of 20-80 years and a mean age of 59.37±5.63 years. The patients were divided into drug resistance group with 52 patients and good response group with 34 patients， and 45 healthy volunteers were enrolled as control group. EEG waveform characteristics of epileptic seizures were observed， as well as the temporal distribution and linear characteristics of different stages of seizures. The Pearson correlation coefficient was used to evaluate the degree of synchronization between EEG leads. The three groups were compared in terms of baseline data and EEG network parameters of each frequency band， and the P-value connection map of brain network was drawn to analyze the association between EEG and drug resistance in patients.Results At the sampling rate of 0.5-60 Hz， spikes were the most typical epileptic discharge on the EEG of patients with epilepsy； at the sampling rate of 250 Hz， 190 s EEG time-domain waveforms were sampled in the interictal， preictal， and ictal periods of epilepsy， and it was found that the number of spikes showed a trend of dynamic phase change with the transition of epilepsy and gradually increased with the advance of epilepsy. There was a significant difference in lead synchronization between the interictal period and the ictal period， and compared with the interictal period， there was an increase in the number of strongly correlated channels the ictal period. Comparison of baseline data between the control group， the drug resistance group， and the good response group showed that compared with the good response group， the drug resistance group had a significantly longer course of disease， a significantly higher frequency of seizures， and a significantly higher number of types of previously used drugs （P<0.05）. In θ band and β band， there were significant differences in clustering coefficient， characteristic path length， global efficiency， and local efficiency between the three groups （P<0.05）； in α band， there were significant differences in characteristic path length and global efficiency between the three groups （P<0.05）. For each frequency band between the good response group and the drug resistance group， the short-range connection was enhanced in the frontal head and temporal regions of the drug resistance group. Between the drug resistance group and the control group， the long-range connection in the frontal head of the drug resistance group was enhanced in the β band， and the short-range connection in the frontal head of the drug resistance group was enhanced in the δ band， θ band， and α band； the short-range connection in the posterior head of the drug resistance group was weakened across all frequency bands. Between the good response group and the control group， the long-range connection in the frontal head of the good response group was enhanced in the δ band， and the short-range connection in the posterior head of the good response group was weakened across all frequency bands.Conclusions EEG signal feature analysis can effectively identify the stage of seizure. Compared with the control group and the good response group， the drug resistance group shows significant changes in the properties and topology of the brain network. Forehead brain network connectivity is enhanced in the drug resistance group. The research on EEG network can help identify drug resistance in focal epilepsy.