基于高频颅内脑电的致痫区定位和网络分析

目的通过癫痫发作期颅内高频脑电(60~90 Hz)的分析,实现一种基于非参数统计模型的致痫区定位方法,并研究致痫区内外的脑电信号相干性。方法首先构造高频脑电归一化指标——高频能量指数,再用该指标量化各导联的致痫性,并定位致痫区,最后分析致痫区内外电极的幅度相干和相位相干性。结果分析了8例有良好术后效果的癫痫患者共1252个电极触点的颅内脑电数据,以临床癫痫灶定位结果作为参考,高频能量指数平均敏感性29.57%、特异性96.67%。8位患者的致痫电极之间的平均相干性显著高于非致痫电极之间的相干性。结论构造了颅内脑电高频能量指标,并用非参数统计模型定位致痫区,特异性高于已有研究,能够帮助医生在癫痫手术规划中寻找关键相关致痫电极。

Epileptogenic zone localization and network analysis based on high frequency intracranial EEG during seizure

Objective To localize epileptogenic zone and to analyze its network coherence using high frequency intracranial EEG during seizure.Methods First, a normalized high frequency energy index (HFEI) was calculated using intracranial EEG.Then, the index was used to quantify the epileptogenicity of each electrode.Finally, the wavelet coherence was estimated between epileptogenic and non-epileptogenic electrodes.Results The HFEI of intracranial EEG (1252 contacts) from 8 patients with good postoperative outcome were extracted and those electrodes with HFEI above the threshold were identified as epileptogenic electrodes.Compared with physician's localization results, the sensitivity of our method was 29.57%, and specificity was 96.67%.Average coherence between epileptic electrodes of the 8 patients selected by HFEI was significantly higher than that between non epileptic electrodes.Conclusions As a nonparametric statistical method, the HFEI index we propose is successfully used to localize epileptogenic network, with higher specificity than previous studies, which meets the need of identifying the key epileptic electrodes during pre-surgical planning.