Selection of optimal AR spectral estimation method for EEG signals using Cramer-Rao bound

Electroencephalography is an essential clinical tool for the evaluation and treatment of neurophysiologic disorders related to epilepsy. Careful analyses of the electroencephalograph (EEG) records can provide valuable insight and improved understanding of the mechanisms causing epileptic disorders. The detection of epileptiform discharges in the EEG is an important element in the diagnosis of epilepsy. In this study, EEG signals recorded from 30 subjects were processed using autoregressive (AR) method and EEG power spectra were obtained. The parameters of autoregressive method were estimated by different methods such as Yule-Walker, covariance, modified covariance, Burg, least squares, and maximum likelihood estimation (MLE). EEG spectra were then used to analyze and characterize epileptiform discharges in the form of 3-Hz spike and wave complexes in patients with absence seizures. The variations in the shape of the EEG power spectra were examined in order to obtain medical information. These power spectra were then used to compare the applied methods in terms of their frequency resolution and determination of epileptic seizure. The Cramer-Rao bounds (CRB) were derived for the estimated AR parameters of the EEG signals and the performance evaluation of the estimation methods was performed using the CRB values. Finally, the optimal AR spectral estimation method for the EEG signals was selected according to the computed CRB values. According to the computed CRB values, the performance characteristics of the MLE AR method was found extremely valuable in EEG signal analysis.     

Heart rate variability during sleep in children with partial epilepsy

Alterations in autonomic control of cardiac activity in epileptic patients have been reported by several studies in the past, and both ictal and interictal modifications of heart rate regulation have been described. Alterations of autonomic control of cardiac activity can play an important role in sudden unexplained death in patients with epilepsy (SUDEP). However, the presence of specific changes in heart rate variability (HRV) during sleep, not correlated with seizures, has not been assessed in children with epilepsy; for this reason, we evaluated features of cardiac autonomic function during sleep without ictal epileptiform electroencephalogram (EEG) activity in a group of children with partial epilepsy. Eleven patients (five males and six females; mean age 11.5 years, SD: 3.65 years) affected by partial epilepsy were admitted to this study; 11 normal subjects (five males and six females; mean age 12.9 years, SD: 2.72 years) served as a control group. All subjects slept in the laboratory for two consecutive nights. The data were analyzed during the second night. Sleep was polygraphically recorded [including one electrocardiography (ECG) channel] and signals were digitally stored. A series of 5-min ECG epochs were chosen from each sleep stage, during periods without evident ictal epileptiform activity in the EEG. Electrocardiography signals were analyzed for automatic detection of R-waves and, subsequently, a series of time- and frequency-domain measures were calculated. Epileptic subjects tended to show an overall lower HRV in both time- and frequency-domain parameters, principally during rapid eye movement (REM) sleep and, to a lesser extent, during sleep stage 2. Among the different bands, this decrease was most evident for the high-frequency band (HF) absolute power. For this reason, the ratio between the low-frequency band (LF) and HF was always higher in epileptic patients than in normal controls and the difference was statistically significant during sleep stages 3 and/or 4 and REM sleep. Our results indicate that during sleep, a particular condition of basal modification in autonomic characteristics occurs (mostly during REM sleep) in partial epilepsy patients. This finding might represent an important factor contributing to the complex mechanism of SUDEP which takes place most often during sleep and supports the need of studying HRV specifically during this state in subjects with seizures.     

局限性癫痫患儿体感诱发电位特征及临床应用

目的：探讨局限性癫痫患儿体感诱发电位特征及应用价值。方法：依据头颅ＣＴ结果分组观测５１例局限性癫痫患儿及３０例正常儿童的双侧正中神经短潜伏期体感诱发电位（ＳＬＳＥＰ）各指标的差异，并与常规脑电图结果进行了比较。结果：与对照组相比，脑部ＣＴ未发现病变组Ｎ２０波幅在癫痫灶侧明显高于对侧；而癫痫灶侧大脑病变组的病灶侧Ｎ２０波幅却明显低于对侧。在部分脑电图未显示局灶性异常的病例也同样显示了这种差异。结论：对于局限性癫痫患儿进行正中神经ＳＬＳＥＰ检查可有助于早期判断症状性癫痫和特发性癫痫，在某种意义上它可能较常规脑电图检查具有更高的敏感性     

应用因果分析方法对癫痫发作间期头皮脑电信号进行致痫灶定侧

目的利用因果分析方法对癫痫发作间期的头皮脑电信号进行致痫灶定侧。方法在频域因果分析方法——自适应直接传递函数（ADTF）的基础上提出功率谱加权ADTF（psADTF）法,以给定频段内信号的功率谱对ADTF的标准化做加权,以适应不同癫痫波频域信息不同的特点。利用该方法对2组共30例患者的头皮脑电信号进行分析,含组1癫痫手术患者15例,组2门诊癫痫患者15例。其中组1患者共截取癫痫波样本数104个,组2患者共截取癫痫波样本数98个。结果组1患者通过psADTF分析对致痫灶定侧与手术侧一致的有96个,平均单个癫痫波致痫灶定侧准确率可达92％;组2患者通过psADTF分析对致痫灶定侧与专家判读结果一致的有94个,平均单个癫痫波致痫灶定侧准确率达96％。结论发作间期头皮脑电癫痫波信号的psADTF分析能够很好地辅助临床致痫灶定侧。     

Mise en correspondance de relations inter-structures lors de crises d'épilepsie

Matching of relationships between structures during epileptic seizures.The general context of this paper is the study of stereoelectroencephalographic signals (depth EEG) recorded during epileptic seizurs. A matching method is presented. It is aimed at quantifying (from seizure to seizure) the reproducibility of relations between channels estimated with the coherence function. On two recordings, we choose with the expert the signals on which the coherence functions are computed. A bias-correcting algorithm is then applied, and the corrected coherence is averaged on a set of frequencies to get an estimator with a lower variance and easy to interpret. The coherences are quantified and matched by extracting the least cost trace, from which we look for the vectors of coherence having a minimum number of identical coordinates. The method is applied on simulated signals built from scenarios that describe a subtype of temporal lobe epilepsy and then tested on real SEEG signals. Issues of this work are linked to: i) the automatic extraction of similarities in the evolution of relationships that take place between cerebral structures during seizures, and ii) the pointing out of recurrent scenarios in the organization of these seizures.     

Application of correlation dimension and pointwise dimension for non-linear topographical analysis of focal onset seizures

For many patients who are candidates for epilepsy surgery, non-invasive evaluation fails to provide sufficient information to permit surgical treatment. Since there are also definite risks and considerable costs associated with invasive procedures, new (non-invasive) techniques are required. This study provides empirical evidence that a non-linear approach applied to ictal surface electroencephalograms (EEGs) can help to delineate the area of seizure onset and may prove useful in complementing visual analysis of the EEG. Multichannel EEGs, recorded from eight patients with different drug-resistant localisation-related epilepsies, were analysed using the concept of correlation dimension and two extensions based on the pointwise dimension. The latter also provided results in cases where assessment of the correlation dimension was not feasible. Comparative values between 2 and 6 were accepted as the result of the algorithms, mostly 3-4 for the EEG channels strongly reflecting epileptic activity, and 4-6 for the other signals. The proportion of accepted pointwise values was usually 200-800% for strong epileptic EEG activity compared to the other data. The approach permitted the characterisation of the scalp area reflecting epileptic activity. The results obtained were in perfect concordance with those obtained during pre-surgical work-up and confirmed by the post-operative outcome.     

Mobile phone emission increases inter-hemispheric functional coupling of electroencephalographic alpha rhythms in epileptic patients

It has been reported that GSM electromagnetic fields (GSM-EMFs) of mobile phones modulate - after a prolonged exposure - inter-hemispheric synchronization of temporal and frontal resting electroencephalographic (EEG) rhythms in normal young and elderly subjects (Vecchio et al., 2007, 2010). Here we tested the hypothesis that this can be even more evident in epileptic patients, who typically suffer from abnormal mechanisms governing synchronization of rhythmic firing of cortical neurons. Eyes-closed resting EEG data were recorded in ten patients affected by focal epilepsy in real and sham exposure conditions. These data were compared with those obtained from 15 age-matched normal subjects of the previous reference studies. The GSM device was turned on (45 min) in the "GSM" condition and was turned off (45 min) in the other condition ("sham"). The mobile phone was always positioned on the left side in both patients and control subjects. Spectral coherence evaluated the inter-hemispheric synchronization of EEG rhythms at the following frequency bands: delta (about 2-4 Hz), theta (about 4-6 Hz), alpha1 (about 6-8 Hz), alpha2 (about 8-10 Hz), and alpha3 (about 10-12 Hz). The effects on the patients were investigated comparing the inter-hemispheric EEG coherence in the epileptic patients with the control group of subjects evaluated in the previous reference studies. Compared with the control subjects, epileptic patients showed a statistically significant higher inter-hemispheric coherence of temporal and frontal alpha rhythms (about 8-12 Hz) in the GSM than "Sham" condition. These results suggest that GSM-EMFs of mobile phone may affect inter-hemispheric synchronization of the dominant (alpha) EEG rhythms in epileptic patients. If confirmed by future studies on a larger group of epilepsy patients, the modulation of the inter-hemispheric alpha coherence due to the GSM-EMFs could have clinical implications and be related to changes in cognitive-motor function.     

Effects of Repetitive Transcranial Magnetic Stimulation on Spike Pattern and Topography in Patients with Focal Epilepsy

Repetitive Transcranial Magnetic Stimulation (rTMS) is a non-invasive method for brain stimulation. Group-studies applying rTMS in epilepsy patients aiming to decrease epileptic spike- or seizure-frequency have led to inconsistent results. Here we studied whether therapeutic trains of rTMS have detectable effects on individual spike pattern and/or frequency in patients suffering from focal epilepsy. Five patients with focal epilepsy underwent one session of rTMS online with EEG using a 6 Hz prime/1 Hz rTMS protocol (real and sham). The EEG was recorded continuously throughout the stimulation, and the epileptic spikes recorded immediately before (baseline) and after stimulation (sham and real) were subjected to further analysis. Number of spikes, spike-strength and spike-topography were examined. In two of the five patients, real TMS led to significant changes when compared to baseline and sham (decrease in spike-count in one patient, change in topography of the after-discharge in the other patient). Spike-count and topography remained unchanged the remaining patients. Overall, our results do not indicate a consistent effect of rTMS stimulation on interictal spike discharges, but speak in favor of a rather weak and individually variable immediate effect of rTMS on focal epileptic activity. The individuation of most effective stimulation patterns will be decisive for the future role of rTMS in epilepsies and needs to be determined in larger studies.     

Application of correlation dimension and pointwise dimension for non-linear topographical analysis of focal onset seizures

For many patients who are candidates for epilepsy surgery, non-invasive evaluation fails to provide sufficient information to permit surgical treatment. Since there are also definite risks and considerable costs associated with invasive procedures, new (non-invasive) techniques are required. This study provides empirical evidence that a non-linear approach applied to ictal surface electroencephalograms (EEGs) can help to delineate the area of seizure onset and may prove useful in complementing visual analysis of the EEG. Multichannel EEGs, recorded from eight patients with different drug-resistant localisation-related epilepsies, were analysed using the concept of correlation dimension and two extensions based on the pointwise dimension. The latter also provided results in cases where assessment of the correlation dimension was not feasible. Comparative values between 2 and 6 were accepted as the result of the algorithms, mostly 3–4 for the EEG channels strongly reflecting epileptic activity, and 4–6 for the other signals. The proportion of accepted pointwise values was usually 200–800% for strong epileptic EEG activity compared to the other data. The approach permitted the characterisation of the scalp area reflecting epileptic activity. The results obtained were in perfect concordance with those obtained during pre-surgical work-up and confirmed by the post-operative outcome.     

基于相位斜率指数的脑电信号方向研究

目的将相位斜率指数(phase-slope index,PSI)算法应用在健康及癫痫患者的脑电数据上,分析导联间信息流的方向和强度,拟为临床医学研究癫痫提供借鉴方法。方法首先在相位斜率指数理论方法的基础上研究了不同脑区间的信息流向的因果关系,并取合适的阈值2作为有效结果;然后对来自军区总医院采集的健康人和癫痫患者的最新脑电数据进行分析,用t检验方法考量健康数据和癫痫数据所得到的有效结果百分比是否具有显著性差异。结果左右区没有明确的方向,前后区在alpha频段有清晰的从前到后的方向;在宽带方法(所有频率)下,健康人数据所得到的有效结果百分比与癫痫患者数据所得到的有效结果百分比的差异无统计意义;在窄带方法(alpha频段)下,健康人数据所得到的有效结果百分比要高于癫痫患者数据所得到的有效结果百分比。结论基于相位斜率指数的脑电信号研究方法得出两种信号的有效结果百分比的差异有统计意义,对临床医学分析癫痫数据具有借鉴意义。     

一种基于改进经验模态分解的癫痫脑电识别新方法

在临床癫痫诊断过程中,为了提高癫痫脑电的识别率,能在癫痫发作前期就预测到癫痫疾病,其特征波的提取至关重要。针对这一问题,提出将平行延拓与镜像延拓相结合来改进EMD算法。首先,使用平行延拓的方法,在原始脑电信号的左、右端点处分别预测出一个极值;然后,使用基于镜像延拓的EMD方法,对信号进行镜像延拓,以避免经验模态分解过程中的端点效应;最后,采用支持向量机进行信号的分类识别。算法验证数据取自德国伯恩大学癫痫研究中心的脑电数据库,其中50例是正常脑电信号、50例是癫痫发作间期的脑电信号。实验研究表明：该方法对总测试脑电信号的识别率达到94%。其中,正常脑电信号和癫痫脑电信号的独立识别率均为94%,比传统EMD算法处理后的脑电识别率提高了5%,可见该方法可以有效地预测癫痫脑电。     

局限性癫痫脑电时间序列的分形维数计算比较

为探索应用非线性动力学理论诊断癫痫病的新方法，对局限性癫痫病患者脑电时间序列进行了相关维数（Ｄｃ）和波形维数（Ｄｗ）的计算比较。观察到，痫性导联脑电的分形维数多低于对侧导联的值；相关维数Ｄｃ的相对变化量较波形维数Ｄｗ大；而波形维数Ｄｗ对痫性棘波敏感；结果提示，脑电时间序列的分形维数有可能成为诊断癫痫的特征参数，值得进一步深入研究。     

Dipole tracing in childhood epilepsy with special reference to Rolandic epilepsy

To assess the clinical applicability of dipole tracing in childhood epilepsy, the location and stability of electric source generator of focal spikes seen in EEG of epileptic children were investigated using this method. The patients were divided into 3 groups; benign childhood epilepsy with centrotemporal spikes (BCECS, Group A, n = 14), other types of epilepsy with centrotemporal spikes (Group B, n = 15) and epilepsy with focal spikes in other areas (Group C, n = 13). The spike dipole in each group was analyzed using dipole tracing method. The following results were obtained. (1) The spikes of BCECS were characterized by constantly stable dipoles, compared to those of the other types of childhood epilepsy. The spikes of epileptic children with mental retardation mostly lacked the stability of dipoles. These seemed to suggest that the stability of dipoles was closely related to the prognosis and pathophysiology of epilepsy. (2) The dipoles of BCECS were localized strictly in the Rolandic area. Dipole tracing from EEG spikes was considered to contribute to the elucidation of the pathophysiology of childhood epilepsy.     

Negative BOLD Response to Interictal Epileptic Discharges in Focal Epilepsy

In EEG–fMRI studies, BOLD responses related to interictal epileptic discharges (IEDs) are most often the expected positive response (activation) but sometimes a surprising negative response (deactivation). The significance of deactivation in the region of IED generation is uncertain. The aim of this study was to determine if BOLD deactivation was caused by specific IED characteristics. Among focal epilepsy patients who underwent 3T EEG–fMRI from 2006 to 2011, those with negative BOLD having a maximum t-value in the IED generating region were selected. As controls, subjects with maximum activation in the IED generating region were selected. We established the relationship between the type of response (activation/deactivation) and (1) presence of slow wave in the IEDs, (2) lobe of epileptic focus, (3) occurrence as isolated events or bursts, (4) spatial extent of the EEG discharge. Fifteen patients with deactivation and 15 with activation were included. The IEDs were accompanied by a slow wave in 87 % of patients whose primary BOLD was a deactivation and only in 33 % of patients with activation. In the deactivation group, the epileptic focus was more frequently in the posterior quadrant and involved larger cortical areas, whereas in the activation group it was more frequently temporal. IEDs were more frequently of long duration in the deactivation group. The main factor responsible for focal deactivations is the presence of a slow wave, which is the likely electrographic correlate of prolonged inhibition. This adds a link to the relationship between electrophysiological and BOLD activities.     

样本熵及在脑电癫痫检测中的应用

对癫痫进行检测和预报具有非常重要的临床意义。首先采用近似熵对癫痫患者的脑电信号进行分析，探索运用近似熵进行癫痫检测的可行性。针对近似熵存在的问题，选用一种与近似熵类似但精度更好的方法一样本熵，并同近似熵就在癫痫脑电信号中的应用进行了比较分析。结果显示癫痫发作时脑电信号的近似熵和样本熵均明显低于发作前和发作后。样本熵的变化幅度明显大于近似熵，样本熵的变化幅度相对于近似熵提高了约10％～25％。     

Activation of the Glutamatergic System of the Medial Septal Area Accelerates Epileptogenesis

The effects of injection of L-glutamate into the medial septal area on epileptogenesis associated with stimulation of the perforant pathway were studied in conscious guinea pigs. EEG field potentials were recorded in the hippocampus and medial septal area in the initial state and on formation of epileptic foci in animals given physiological saline (controls) or L-glutamate (glutamate group) into the medial septal area. Administration of glutamate enhanced the theta rhythm in both structures. During kindling in animals given glutamate, there was significantly earlier onset of the convulsive state and secondary epileptiform discharges than in controls. As the epileptic focus formed, sharp weakening of theta oscillations was seen in all animals, though onset was much earlier in the glutamate group. The correlation between the hippocampal and septal EEG decreased sharply in animals given glutamate, while changes in controls were minor. These results are of value in deciphering the mechanisms of temporal epilepsy.     

术中颅内电极脑电监测在联合术式治疗顽固性癫痫中的应用

目的：探讨术中颅内电极脑电监测在癫痫外科手术中的应用方法及价值。方法：回顾分析难治性癫痫患者78例术中采用颅内电极脑电监测,在病灶切除前对致痫灶区行皮层电极描记,记录有无痼样放电和范围,对痫灶部位进一步精确定位。采用联合手术方法进行病灶切除,之后再次或多次描记,以判断痼样放电有无减少或消失,对有痫样放电的皮层再次进行软膜下横切术。术后常规应用抗癫痫药物。结果：78例患者,术前均在预定的致痫灶局部记录到棘波、棘慢波、尖波和尖慢波;致痢灶切除后即时监测结果显示其中有7例痫样放电完全消失,56例痫样放电明显减少、背景波幅降低,15例在病灶周围仍散在有明显痫样放电,经再次处理后减少,所有患者均无神经功能损害加重。78例患者均随访6个月至4年,多数患者生活质量提高、精神状态改善。其中发作消失51例（65％）,发作明显减少（75％以上）16例（21％）,发作减少（50％以上）7例（9％）,发作无变化4例（5％）。总有效率为95％。结论：在癫痫外科手术中运用颅内电极进行脑电监测,能进一步明确致痂灶部位、监测异常波出现的范围及异常程度,指导手术中正确切除致痫灶,在癫痫外科治疗中具有一定的实用价值。     

惊厥患儿的脑电图分析

目的：探讨癫痫、急性病毒性脑炎、热性惊厥（FC）与脑电图（EEG）关系。方法：记录166例8个月至6岁患儿EEG。结果：首次EEG异常率达90％，三天后复查EEG异常率达60％，在出现癫痫症状的急性病毒性脑炎患儿的EEG描记中，以低至中波幅慢波增多，呈弥散性或局限性异常者居多，出现癫痫样放电占50％，由热性惊厥引起的EEG变化，三天后复查EEG大都好转，部分患儿多次EEG检查出现发作性尖波，6岁后发热引起惊厥的患儿，最后诊断为癫痫。结论：在小儿发热引起的惊厥中，EEG的检查是一项重要的临床检查项目，对了解脑功能状况、排除病毒性脑炎、癫痫、热性惊厥有重要的参考作用。     

利用广义复杂度对癫痫病发作进行预测的方法研究

目的：通过用广义复杂度的运算方法,分析癫痫病患者发病期间的脑电图,计算出这一段脑电图数据的复杂度,通过对复杂度数据的分析,对癫痫病的发作进行预测。进而提高癫痫病患者的生活质量,降低癫痫病患者的生命危险。方法：将癫痫病发病的脑电图的图像信号还原成数字信号,计算出该数字信号各个区间的广义复杂度值。得出癫痫病发病前脑电图数据复杂度的变化状态,对癫痫病患者发病进行预测。结果：计算出21名癫痫病患者在发病前后共一个小时的脑电图数字信号的广义复杂度值,通过观察21名癫痫病患者在癫痫病发病期间的脑电图的数字信号的广义复杂度变化情况,对癫痫病的发作进行预测（对2l名病人癫痫病发作的数据进行预测研究,平均准确率可达94.3%）。结论：本文提出了一种基于广义复杂度的癫痫发作预测方法,通过对21名癫痫病患者的脑电图信号的实验和计算,结果表明,该方法可以有效地预测到癫痫发作。     

癫痫发作时的脑电图研究

脑电图已成为癫痫诊断中不可缺少的手段,特别是发作时的脑电图为癫痫诊断的基本要素,本文对22例癫痫发作时的脑电图进行了研究,分析了发作时的癫痫放电及发作前后的波形演变,对癫痫的分类及治疗有着重要的意义。