Recherche Images Maps Play YouTube Actualités Gmail Drive Plus »
Connexion
Les utilisateurs de lecteurs d'écran peuvent cliquer sur ce lien pour activer le mode d'accessibilité. Celui-ci propose les mêmes fonctionnalités principales, mais il est optimisé pour votre lecteur d'écran.

Brevets

  1. Recherche avancée dans les brevets
Numéro de publicationCN104545900 B
Type de publicationOctroi
Numéro de demandeCN 201410836452
Date de publication22 févr. 2017
Date de dépôt29 déc. 2014
Date de priorité29 déc. 2014
Autre référence de publicationCN104545900A
Numéro de publication201410836452.3, CN 104545900 B, CN 104545900B, CN 201410836452, CN-B-104545900, CN104545900 B, CN104545900B, CN201410836452, CN201410836452.3
Inventeurs殷涛, 王欣, 刘志朋
Déposant中国医学科学院生物医学工程研究所
Exporter la citationBiBTeX, EndNote, RefMan
Liens externes:  SIPO, Espacenet
一种基于配对样本t检验的事件相关电位分析方法 T one kind of event-related potentials analysis method based on paired-sample test Langue du texte original : Chinois
CN 104545900 B
Résumé  Langue du texte original : Chinois
一种基于配对样本T检验的事件相关电位分析方法:设计含两种刺激的脑电诱发实验,并利用脑电采集设备记录多个导联的头皮脑电信号,进行初步的预处理;提取两种刺激下的ERP信号;对两种刺激下的ERP信号进行配对样本T检验,确定具有显著差异的时段;计算两种刺激下的ERP信号在显著差异时段内的差异面积,并绘制脑地形图,确定差异脑区。 An event-related potential analysis based on paired samples T test: Evoked Potential design comprising two stimulation experiments, using a plurality of recording EEG scalp EEG acquisition device leads, initial pretreatment; extracting two ERP signal at the stimuli; on ERP signal under two stimulus paired samples T test, determined to have a period significantly different; calculated difference area ERP signal at two stimuli in the significant difference in time and draws brain mapping to determine the difference brain regions. 本发明确定了两种刺激下的ERP显著差异时段,并绘制了基于ERP波形差异面积的脑地形图,进而得到在显著差异时段内的差异脑区分布。 ERP of the present invention determines a significant difference in the two stimulation period, and the waveform drawn ERP brain mapping area difference based on the difference obtained and further significant difference in cerebral distribution period. 本发明对于信噪比较差、且单个成分并不明显的ERP研究具有重要意义,并且为ERP信号和自发脑电的剥离提供了新的思路。 Poor signal to noise ratio for the present invention, the individual components are not significant and important to study the ERP, and provides a new way for ERP EEG signals and peeling.
Revendications(3)  Langue du texte original : Chinois
  1. I. 一种基于配对样本T检验的事件相关电位分析方法,其特征在于,包括如下步骤: 1) 设计含两种刺激的脑电诱发实验,并利用脑电采集设备记录多个导联的头皮脑电信号,进行初步的预处理; 2) 提取两种刺激下的ERP信号;所述的提取两种以上刺激下的ERP信号,包括如下过程: (1)对初步预处理之后的整段头皮脑电信号进行分割,得到20个时长为4s的静息脑电片段和20个时长为Is的诱发脑电片段,其中,两种刺激对应的诱发脑电片段各10个; ⑵选取刺激呈现之前的200ms,即静息期的后200ms为基准脑电,并计算基准脑电的平均幅值,将每个诱发脑电片段减去基准脑电平均幅值,实现去基线操作; (3)分别对去基线后的两种诱发脑电片段进行叠加平均,得到每位被试者的每个导联在两种刺激下的ERP信号,分别表示为X= {Xijk}和Y= {Yijk},其中,i = l,2,……,见,见 I. Based on paired samples T-test analysis of the event-related potential, characterized by comprising the steps of: a) Design Evoked Potential comprising two stimulation experiments, using a plurality of recording scalp EEG leads acquisition device EEG, initial pretreatment; 2) extracting ERP signal at the two stimuli; extracting the ERP signals at two or more stimulus, comprising the following processes: (1) after initial pretreatment of the whole scalp EEG segmentation, 20 when resting EEG fragment length 4s and 20 when the length is induced EEG segments, wherein two stimuli evoked EEG segments each corresponding to 10; ⑵ selected before stimulation presentation of 200ms, i.e., after the period of 200ms resting EEG as a reference, and calculates the average amplitude of EEG reference, each segment of EEG evoked by subtracting the reference average EEG amplitude, baseline operation to achieve; (3), respectively, to the baseline of the two segments are overlapped and averaged evoked potentials, to give each lead in each subject's ERP both stimulation signals, denoted as X = {Xijk} and Y = {Yijk}, where, i = l, 2, ......, see, see =15, Ni是被试者数目;j = 1,2,......,犯,吣=32,犯是导联数目,1^=1,2,......,N3,N3 = 1024,N3是数据点数; 3) 对两种刺激下的ERP信号进行配对样本T检验,确定具有显著差异的时段; 所述的对两种刺激下的ERP信号进行配对样本T检验,是对每个导联中的每个数据点对应的ERP序列进行配对样本T检验,对于第j个导联第k个数据点,首先建立一个新变量Z = {Zijk} ,Zijk = Xijk-Yijk, i = l ,2,……,N1,计算新变量的均值 = 15, Ni is the number of subjects; j = 1,2, ......, guilty, Qin = 32, make the number of leads, 1 ^ = 1,2, ......, N3 , N3 = 1024, N3 is the number of data points; 3) of the ERP signal under two stimulus paired samples T test to determine significant differences having a period; ERP signal according to the two paired samples T test stimuli, is paired samples T test sequence ERP for each data point of each corresponding lead, the k-th data point for j-th leads, first create a new variable Z = {Zijk}, Zijk = Xijk-Yijk , i = l, 2, ......, N1, calculating a new mean variable
    Figure CN104545900BC00021
    和方差 And variance
    Figure CN104545900BC00022
    ,构造统计量 , Construction statistics
    Figure CN104545900BC00023
    检验tjk是否服从自由度为N1-I的T分布,并计算出对应的显著程度Pjk,若Pjk〈0.05,则序列{ Verify tjk subject to a degree of freedom of the N1-I T distribution, and calculate the corresponding PJK significant degree, if Pjk <0.05, then the sequence {
    Figure CN104545900BC00024
    具有显著性差异,即两种刺激下的第j个导联第k个数据点具有显著性差异,否则,两种刺激下的第j个导联第k个数据点不具有显著性差异; 所述的确定具有显著差异的时段,是在已确定两种刺激下的第j个导联第k个数据点是否具有显著性差异的基础上进行的,包含如下过程: (1) 对于第j个导联的N3个数据点,若存在10个以上的连续数据点k,使得序列 Significant difference, i.e. the k-th data point j th lead in both stimulated significant difference, otherwise, the j-th lead k-th data point in the two stimulation no significant difference; the period determined to have significant differences described below, is has been determined that the j th lead of the k-th data point in the two stimulation whether conducted on the basis significant differences on, comprising the following process: (1) for the j th N3 data points leads, if more than 10 consecutive data points k exists, such that the sequence
    Figure CN104545900BC00025
    和序列U具有显著性差异,那么这些连续的数据点k对应的时段就是第j个导联在两种刺激下的ERP信号具有显著差异的时段,若不存在10个以上的连续数据点k,使得序列…,和序列…,心1#}具有显著性差异,则第j 个导联在两种刺激下的ERP信号不具有显著差异时段; (2) 根据所有导联的显著差异时段分布,选择一个以上相对大的时段,使得尽可能包含多数导联的显著差异时段,所选的显著差异时段对应的数据点集标记为丨丨,其中,r=l, 2,…,m,m为所选的显著差异时段数,可选范围为{1,2,…,100},nr分别是每个显著差异时段所对应的数据点数,取值均大于10; 4) 计算两种刺激下的ERP信号在显著差异时段内的差异面积,并绘制脑地形图,确定差异脑区; 所述的计算两种刺激下的ERP信号在显著差异时段内的差异面积,是对每位被试者的每个导联均计算两种 Period and sequence U having significant difference, then the successive data points corresponding to k time period is the j-th lead having a significant difference in the ERP signals at two stimuli, if the presence of more than 10 consecutive data point k, such that the sequence of ..., and the sequence ..., heart 1 #} significant difference, the j-th lead ERP signal at both stimuli no significant difference period; (2) the distribution of significant differences according to time all leads, selecting one or more relatively large period of time, so as to contain significant difference period majority leads as possible, significant differences in the chosen slot corresponding to the set of data points marked as Shushu, wherein, r = l, 2, ..., m, m is the significant difference in the number of the selected period, the selectable range of {1,2, ..., 100}, nr are significantly different for each time period corresponding to the number of data points, the values were greater than 10; 4) calculated both stimuli ERP signal difference significant difference in the area of the period, and draw brain mapping, determine the difference brain; calculating the difference in the two areas ERP signal significant differences in the stimulation period, the subject is each of two kinds of each lead are calculated 激下的ERP信号在所选显著差异时段内的差异面积,其中,在显著差异时段 ERP difference signal in the area stimulated significant differences within the selected period of time, wherein a significant difference in the period
    Figure CN104545900BC00031
    内的两种刺激下的ERP信号的差异面积为 ERP area difference signal in both of stimulating
    Figure CN104545900BC00032
    =对所有被试者的数据进行叠加平均,得到第j个导联在所选显著差弁时段内网柙刺微卜的ERP信号差异面牙 = Overlapped and averaged data for all subjects, the difference signal obtained ERP tooth surface leads the j-th network within a selected period Bian significant difference pen barbed micro Bu
    Figure CN104545900BC00033
  2. 2. 根据权利要求1所述的一种基于配对样本T检验的事件相关电位分析方法,其特征在于,步骤1)所述的初步的预处理,是为去除头皮脑电信号记录过程中的低频漂移、高频干扰以及眼动电生理信号的干扰,对原始脑电信号进行变平均参考、0.5-lOHz带通滤波以及独立成分分析去眼电的预处理操作。 According to one of the claims 1 based on event-related potential analysis paired samples T test, wherein the initial pretreatment step 1), for the removal of the scalp EEG signal during recording the low frequency drift, high-frequency interference and an interference eye electrophysiological signals, the original EEG mean reference variable, 0.5-lOHz band-pass filtering and pre-processing operations independent component analysis to the eye power.
  3. 3. 根据权利要求1所述的一种基于配对样本T检验的事件相关电位分析方法,其特征在于,步骤4)中所述的绘制脑地形图、确定差异脑区,是对于每一个显著差异时段,均根据所有导联的两种刺激下的ERP信号差异面积绘制脑地形图,进而分析两种刺激所诱发ERP信号差异的空间分布状况,得到在显著差异时段内的主要激活脑区分布。 According to one of the claims 1 paired sample T test based on event-related potential analysis, wherein, rendering brain mapping step 4) was used, determining a difference brain, it is a significant difference for each period, according to both brain mapping drawing ERP signal in the two stimulation leads all the difference area, further analysis of the signal difference ERP both stimuli induced spatial distribution, to obtain the distribution of the major brain regions in the activation significant difference period.
Description  Langue du texte original : Chinois
一种基于配对样本T检验的事件相关电位分析方法 An analytical method based on paired-sample T test event-related potentials

技术领域 FIELD

[0001] 本发明涉及一种事件相关电位分析方法。 [0001] The present invention relates to an event-related potential analysis. 特别是涉及一种包含多种外界刺激、且单种刺激重复次数较少的基于配对样本T检验的事件相关电位分析方法。 More particularly it relates to an external stimulus, and a single stimulus was repeated based on the event-related potential analysis paired sample T test method for a smaller number.

背景技术 Background technique

[0002] 脑电信号是通过电极记录下来的脑细胞群的自发性、节律性电活动,根据是否含有外部刺激,可分为自发脑电(Electroencephalo_graph,EEG)和事件相关电位(Event-related Potentials ,ERP) 两种。 [0002] EEG signals recorded by the electrodes spontaneous, rhythmic electrical activity of brain cells population, according to whether they contain external stimulus, can be divided into EEG (Electroencephalo_graph, EEG) and event-related potentials (Event-related Potentials , ERP) two kinds. 事件相关电位是人们对特定刺激事件进行感知加工或执行某种认知任务时诱发出来的一种脑电信号,常用于反映刺激发生前后脑电位的变化情况,与大脑注意资源分配、客体记忆、思维决策、认知加工等相关。 Event-related potentials are people to kind of brain electrical signals induced perception out when processing or perform some cognitive tasks to a particular stimulus event, commonly used in potential changes in the brain before and after reflect irritation occurs, resource allocation and brain attention, memory object, decision-making related to thinking, cognitive processing. 由于ERP信号具有毫秒级的时间分辨率、良好的非侵入性,并且采集设备操作较简单,该信号在脑功能研究和脑疾病预诊方面都有颇多应用。 As the ERP signal having millisecond time resolution, good non-invasive, simple operation and acquisition device, the signal has a lot of applications in brain function and brain disease research aspect pre-diagnosis.

[0003] 多年来,ERP研究的一个重大难点就是与自发脑电的剥离。 [0003] Over the years, a major difficulty ERP study was stripped of the EEG. 研究显示,大脑无时无刻不在运转,即使在不给任何外界刺激的情况下,中枢神经系统亦存在着节律性、自发性放电现象,而外部事件诱发的ERP信号幅值远小于自发脑电,且通常被淹没在自发脑电中。 Studies have shown that the brain is not running all the time, even without any external stimulation to the central nervous system, there is also a rhythmic, spontaneous discharge phenomenon, and ERP signal amplitude is much smaller than external events induced EEG, and often They lost in the spontaneous EEG. 由于自发脑电具有很大的个体差异性和随机性,因此不能形成一个固定的自发脑电模板,使得ERP信号便于剥离。 Since EEG great individual differences and randomness, and therefore can not form a fixed template EEG, ERP signals to facilitate such release. 实际过程中,通常采用多次重复施加外部刺激、再求平均的方式提高ERP信号的幅值和纯度,进而将其与自发脑电剥离。 The actual process, usually repeated application of an external stimulus, and then averaging the magnitude of ways to enhance the purity and ERP signal, which in turn release the EEG. 为了得到信噪比较好的ERP信号,通常需要几十甚至几百次的重复外界刺激,一方面,多次重复刺激必然会引起感官系统的疲劳,且难以保持完全一致的重复性;另一方面,大量刺激材料的准备并不容易,尤其是对于特定含义的图片、声音等较复杂的刺激。 ERP in order to obtain a better signal to noise ratio, typically require tens or even hundreds of times repeated external stimulation, on the one hand, will inevitably lead to fatigue repeated stimulation of the sensory system, and it is difficult to maintain consistent repeatability; other respect, a lot of preparation to stimulate the material is not easy, especially for more complex stimuli specific meaning of pictures, sounds and so on.

[0004] 另外,以往的ERP分析多集中于某个或者某几个ERP成分(如P1,N1,P3等)的分析, 然而对于重复刺激次数较少、信噪比欠佳的ERP信号,具有明确物理意义的ERP成分往往难以识别,也造成不同刺激下ERP特征提取的困难。 Analysis [0004] Further, conventional ERP analysis concentrated on one or a few ERP components (e.g., P1, N1, P3, etc.), whereas for less frequently repeated stimulation, ERP poor signal to noise ratio, having clear physical meaning of ERP components are often difficult to identify, but also cause difficulties ERP features extracted under different stimuli.

[0005] 基于配对样本T检验的事件相关电位分析方法从显著差异的角度分析ERP信号,能够提取出有显著意义的差异特征,避开了单个ERP成分提取的困难,是ERP对比分析的新思路。 [0005] Analysis of paired samples T test event-related potential analysis methods based on the perspective of significant differences in ERP signal can be extracted difference characteristics was significant, and to avoid the difficulty of a single ERP components extracted Ownenship ERP comparative analysis . 另外,由于两种不同外界刺激下的自发脑电虽不完全一致,但也不具有显著的差异性, 若对两种刺激下的ERP信号进行配对样本T检验,得到的显著差异时段必为两真实ERP信号具有显著差异的时段,可以实现ERP信号与自发脑电的间接剥离。 Further, since the EEG at two different external stimuli, though not exactly the same, but not a significant difference, when the two signals on the ERP stimulation paired samples T-test, a significant difference will be obtained for the two periods ERP signal period having real significant difference can be achieved indirectly ERP signal the spontaneous release of EEG.

发明内容 SUMMARY

[0006] 本发明所要解决的技术问题是,提供一种可用于信噪比较差、且单个成分并不明显的ERP研究的基于配对样本T检验的事件相关电位分析方法。 [0006] The present invention solves the technical problem, to provide a poor signal to noise ratio, based on an event-related potential analysis methods paired sample T test and the individual components are not significant ERP research.

[0007] 本发明所采用的技术方案是:一种基于配对样本T检验的事件相关电位分析方法, 包括如下步骤: [0007] The technical proposal of the present invention is: an event-related potential analysis methods based on paired samples T test, comprising the steps of:

[0008] 1)设计含两种刺激的脑电诱发实验,并利用脑电采集设备记录多个导联的头皮脑电信号,进行初步的预处理; [0008] 1) Experiment design containing both stimuli induced EEG, and EEG acquisition device using a plurality of recording scalp EEG leads, initial pretreatment;

[0009] 2)提取两种刺激下的ERP信号; [0009] 2) extracting ERP signal at two stimuli;

[0010] 3)对两种刺激下的ERP信号进行配对样本T检验,确定具有显著差异的时段; [0010] 3) of the ERP signal under two stimulus paired samples T test to determine significant differences in period;

[0011] 4)计算两种刺激下的ERP信号在显著差异时段内的差异面积,并绘制脑地形图,确定差异脑区。 [0011] 4) Calculate the difference in the area under the two ERP signal significant differences in the stimulation period, and draw brain mapping, determine the difference brain regions.

[0012] 步骤1)所述的初步的预处理,是为去除头皮脑电信号记录过程中的低频漂移、高频干扰以及眼动电生理信号的干扰,对原始脑电信号进行变平均参考、0.5-10HZ带通滤波以及独立成分分析去眼电的预处理操作。 Preliminary pretreatment of the [0012] Step 1), for the removal of the scalp EEG recording low frequency drift, the high-frequency interference and interference eye electrophysiological signals, the original EEG mean reference variable, 0.5-10HZ band-pass filtering and pre-processing operations independent component analysis to the eye power.

[0013] 步骤2)所述的提取两种以上刺激下的ERP信号,包括如下过程: [0013] Step 2) Extraction of the ERP in the stimulation of two or more signals, the process comprising:

[0014] (1)对初步预处理之后的整段头皮脑电信号进行分割,得到20个时长为4s的静息脑电片段和20个时长为Is的诱发脑电片段,其中,两种刺激对应的诱发脑电片段各10个; [0014] (1) the whole of the scalp EEG after initial pretreatment is divided, to obtain 20 when resting EEG fragment length 4s and 20 when the length Is induced EEG segments, wherein the two stimuli corresponding to each evoked potentials segment 10;

[0015] (2)选取刺激呈现之前的200ms,即静息期的后200ms为基准脑电,并计算基准脑电的平均幅值,将每个诱发脑电片段减去基准脑电平均幅值,实现去基线操作; [0015] (2) 200ms before selecting stimulus presentation, i.e., after the period of 200ms resting EEG as a reference, and calculates the average amplitude of the reference EEG, average EEG amplitude for each reference segment is subtracted Evoked Potentials , to achieve a baseline operation;

[0016] (3)分别对去基线后的两种诱发脑电片段进行叠加平均,得到每位被试者的每个导联在两种刺激下的ERP信号,分别表示为X= {Xijk}和Y= {Yijk},其中,i = l,2,……,他,见=15,Ni是被试者数目;j = 1,2,......,N2,N2 = 32,N2是导联数目,k = 1,2,......,N3,N3 = 1024, N3是数据点数。 [0016] (3) respectively of the two fragments were induced to baseline EEG overlapped and averaged to obtain each lead in each subject's ERP both stimulation signals, denoted as X = {Xijk} and Y = {Yijk}, where, i = l, 2, ......, he see = 15, Ni is the number of subjects; j = 1,2, ......, N2, N2 = 32, N2 is the number of leads, k = 1,2, ......, N3, N3 = 1024, N3 is the number of data points.

[0017] 步骤3)中所述的对两种刺激下的ERP信号进行配对样本T检验,是对每个导联中的每个数据点对应的ERP序列进行配对样本T检验,对于第j个导联第k个数据点,首先建立一个新变量Z= {Zijk},Zijk = Xijk-Yijk,i = l,2,……,仏,计算新变量的均 In [0017] Step 3) the ERP signals at the two stimuli paired samples T-test, paired samples T-test is performed for each data point in each ERP sequence corresponding lead, for the j-th lead k-th data point, first create a new variable Z = {Zijk}, Zijk = Xijk-Yijk, i = l, 2, ......, Fo, new variables were calculated

Figure CN104545900BD00051

和方% And Fang%

Figure CN104545900BD00052

,构造统计量; , Construction statistics;

Figure CN104545900BD00053

检验tjk是否服从自由度为Ni-1的T分布,并计算出对应的显者程度?』^右?」以〇.〇5,则序列.沐1,,.12#,.'___,工^/«:丨和序列〖&。 Verify tjk obey the degrees of freedom of the T distribution of Ni-1, and calculate the corresponding significant degree who? "^ Right?" To 〇.〇5, then the sequence. Mu 1 ,,. # 12, .'___, workers ^ / «: Shu and sequence & 〖. ,.61/*:.:,"',.匕1,} 具有显著性差异,即两种刺激下的第j个导联第k个数据点具有显著性差异,否则,两种刺激下的第j个导联第k个数据点不具有显著性差异。 , .61 / *:.:, '' ,. dagger 1,} having a significant difference, i.e., the j-th k-th data leads in the two stimulation points having significant difference, otherwise, the next two stimuli leads the j-th k-th data point does not have a significant difference.

[0018] 步骤3)中所述的确定具有显著差异的时段,是在已确定两种刺激下的第j个导联第k个数据点是否具有显著性差异的基础上进行的,包含如下过程: [0018] Step 3) in said determined to have a period of significant difference, is performed in the determined j th lead of the k-th data point in the two stimulation whether significant differences on the basis of, comprising the following process :

[0019] (1)对于第j个导联的N3个数据点,若存在10个以上的连续数据点k,使得序列^^,12^__,1^}和序列丨}^:^,一,}^丨具有显著性差异,那么这些连续的数据点1^对应的时段就是第j个导联在两种刺激下的ERP信号具有显著差异的时段,若不存在10个以上的连续数据点k,使得序列和序列…,u具有显著性差异,则第j个导联在两种刺激下的ERP信号不具有显著差异时段; [0019] (1) N3 for the j-th data points leads, if more than 10 consecutive data points k exists, such that the sequence ^^, 12 __ ^, ^ 1}} and a sequence Shu ^: ^, a } ^ Shu significant difference, then the 1 ^ corresponding to the period of a consecutive data points is the j-th lead having a period of a significant difference in the ERP signals at two stimuli, if the presence of more than 10 consecutive data points k, such that the sequence and the sequence ..., u has a significant difference, the j-th ERP signal leads in both the stimulation period no significant difference;

[0020] (2)根据所有导联的显著差异时段分布,选择一个以上相对大的时段,使得尽可能包含多数导联的显著差异时段,所选的显著差异时段对应的数据点集标记为其中, r=l,2,…,m,m为所选的显著差异时段数,可选范围为{1,2,…,100},nr分别是每个显著差异时段所对应的数据点数,取值均大于10。 [0020] (2) The significant difference period all leads distribution, selecting one or more relatively large period, so that significant differences period possible comprises a plurality of leads, significant differences in the chosen slot corresponding to the set of data points labeled wherein , r = l, 2, ..., m, m is the number of significant differences in the time period selected, the selectable range of {1,2, ..., 100}, nr are significantly different for each time period corresponding to the number of data points taken values are greater than 10.

[0021] 步骤4)中所述的计算两种刺激下的ERP信号在显著差异时段内的差异面积,是对每位被试者的每个导联均计算两种刺激下的ERP信号在所选显著差异时段内的差异面积, 其中,在显著差异时段内的两种刺激下的£1^信号的差异面积关 [0021] Step. 4) computed for two ERP signal according to difference in the area of stimulation in a significant difference period, is in each ERP signal leads are calculated for each test subject in the two stimuli selected from the significant differences in the area of the difference period, £ 1 ^ difference signal at which a significant difference in the two stimulation period off area

Figure CN104545900BD00061

并对所有被试者的数据进行叠加平均,得到第j个导联在所选显著差异时段内两种刺激下的ERP信号差异面积 All subjects and data are overlapped and averaged, to obtain the difference signal area ERP j th lead in the selected significant difference between the two stimuli period

Figure CN104545900BD00062

[0022] 步骤4)中所述的绘制脑地形图、确定差异脑区,是对于每一个显著差异时段,均根据所有导联的两种刺激下的ERP信号差异面积绘制脑地形图,进而分析两种刺激所诱发ERP 信号差异的空间分布状况,得到在显著差异时段内的主要激活脑区分布。 [0022] Step 4) drawn in the brain mapping, determine a difference brain, it is a significant difference for each period, according to both brain mapping drawing ERP signal in the two stimulation leads all the difference area, further analysis ERP signal differences both stimuli induced spatial distribution, to obtain the distribution of the major brain regions in the activation significant difference period.

[0023] 本发明的一种基于配对样本T检验的事件相关电位分析方法,从配对样本T检验出发,确定了两种刺激下的ERP显著差异时段,并绘制了基于ERP波形差异面积的脑地形图,进而得到在显著差异时段内的差异脑区分布。 [0023] A present invention is based on paired samples T-test analysis of the event-related potential, starting from paired samples T test, ERP is determined under two significantly different stimulation period, and brain mapping plotted based on the difference waveform area ERP FIG thus obtained difference significant difference in cerebral distribution period. 本发明主要针对包含多种外界刺激、且单种刺激重复次数较少的诱发脑电研究,对于信噪比较差、且单个成分并不明显的ERP研究具有重要意义,并且为ERP信号和自发脑电的剥离提供了新的思路。 The present invention is primarily directed comprising a plurality of external stimuli, and a smaller number of studies EEG evoked single repeated stimulation, for poor signal to noise ratio, and the individual components are not significant ERP important to study, and for the ERP signals and spontaneous EEG peel provides a new way of thinking.

附图说明 BRIEF DESCRIPTION

[0024] 图1是一种基于配对样本T检验的事件相关电位分析方法的流程图。 [0024] FIG. 1 is a flowchart illustrating a method of analyzing the paired sample T test based on event-related potential.

具体实施方式 detailed description

[0025] 下面结合实施例和附图对本发明的一种基于配对样本T检验的事件相关电位分析方法做出详细说明。 [0025] The following Examples and detailed description of the drawings made according to the present invention is based on one kind of event-related potential analysis of paired samples T test.

[0026] 本发明的一种基于配对样本T检验的事件相关电位(Event-related Potentials, ERP)分析方法,首先利用脑电采集设备记录两种刺激下的多导头皮脑电信号,并进行初步的预处理;其次提取两种刺激下的ERP信号;再次分别对每一导联在两种外界刺激下的ERP 信号进行配对样本T检验,得到每一导联ERP信号在两种刺激下的显著差异时段;最后通过计算显著差异时段内两ERP信号包围的面积,可以得到两种刺激下全脑所有导联在显著差异时段内的差异面积分布,进而通过脑地形图显示两种刺激所诱发的ERP信号差异的空间分布。 [0026] A present invention is based on paired samples T test event-related potential (Event-related Potentials, ERP) analysis method, firstly the multichannel EEG scalp EEG recording both stimuli acquisition device, and a preliminary pretreatment; ERP signal extracting secondly at two stimuli; respectively for each lead again ERP signal in both external stimulation paired samples T test, ERP signals obtained for each lead in two significant stimulation differences period; and finally by calculating the area of the significant difference in period two ERP signal surrounded can be obtained cerebral both stimulation differences for all leads in a significant difference in period area distribution, and further two stimuli elicited displayed by the brain mapping of ERP differences in the spatial distribution of the signal.

[0027] 如图1所示,本发明的一种基于配对样本T检验的事件相关电位分析方法,具体包括如下步骤: [0027] 1 A of the present invention based on paired samples T-test analysis of the event-related potential, includes the following steps:

[0028] 1)设计含两种刺激的脑电诱发实验,并利用脑电采集设备记录多个导联的头皮脑电信号,进行初步的预处理; [0028] 1) Experiment design containing both stimuli induced EEG, and EEG acquisition device using a plurality of recording scalp EEG leads, initial pretreatment;

[0029] 所述设计含两种刺激的脑电诱发实验,是设计两种视觉、听觉或体感刺激的脑电诱发实验,以两种图片刺激为例,从国际情绪图片库(International Affective Picture System,IAPS)中选取积极场景情境图片和消极场景情境图片各10张,图片采用随机出现的方式进行呈现,每张图片呈现时间为ls,图片呈现之前有4s的静息期,用于平复上一张图片所引起的脑电变化。 [0029] The experimental design comprising two stimulation evoked EEG is to design two kinds of visual, auditory or experimental EEG evoked somatosensory stimulation, stimulation of two pictures, for example, from the international mood gallery (International Affective Picture System , IAPS) selected positive and negative scenes situational picture situational pictures of each scene 10, by way of random pictures were presented, each picture presentation time is ls, 4s have a resting period before a picture that looks for calm on a EEG changes caused by the pictures.

[0030] 所述的利用脑电采集设备记录多个导联的头皮脑电信号,是采用Biosemi ActiveTwo脑电采集系统记录15位被试在实验过程中的32导头皮脑电信号,采样率为1024Hz,记录信号总时长为100s。 Scalp EEG recording using a plurality of EEG leads capture device [0030] the process, the use Biosemi ActiveTwo EEG acquisition systems 15 are again recorded during the experiment in the scalp EEG guide 32, the sampling rate is 1024Hz, the total length of the recording signal 100s.

[0031] 所述的初步的预处理,是为去除头皮脑电信号记录过程中的低频漂移、高频干扰以及眼动等其他电生理信号的干扰,对原始脑电信号进行变平均参考、0.5-lOHz带通滤波以及独立成分分析去眼电等的预处理操作。 [0031] The preliminary preprocessing is interference from other electrophysiological signals scalp EEG signal to remove low frequency drift during recording, high-frequency interference and the like eye movements, the original EEG mean reference variable, 0.5 -lOHz band-pass filtering and pre-processing operations to the independent component analysis or the like EOG.

[0032] 2)提取两种刺激下的ERP信号; [0032] 2) extracting ERP signal at two stimuli;

[0033] 所述的提取两种以上刺激下的ERP信号,包括如下过程: [0033] The extracts ERP stimulation signals at two or more, the process comprising:

[0034] (1)对初步预处理之后的整段头皮脑电信号进行分割,得到20个时长为4s的静息脑电片段和20个时长为Is的诱发脑电片段,其中,两种刺激对应的诱发脑电片段各10个; [0035] (2)选取刺激呈现之前的200ms,即静息期的后200ms为基准脑电,并计算基准脑电的平均幅值,将每个诱发脑电片段减去基准脑电平均幅值,实现去基线操作; [0034] (1) the whole of the scalp EEG after initial pretreatment is divided, to obtain 20 when resting EEG fragment length 4s and 20 when the length Is induced EEG segments, wherein the two stimuli corresponding to each evoked potentials segment 10; prior [0035] (2) select the stimulus presentation 200ms, i.e., after the period of 200ms resting EEG as a reference, and calculates the average amplitude of EEG reference, each evoked electrical segment by subtracting the reference average EEG amplitude, baseline operation to achieve;

[0036] (3)分别对去基线后的两种诱发脑电片段进行叠加平均,得到每位被试者的每个导联在两种刺激下的ERP信号,分别表示为X= {Xijk}和Y= {Yijk},其中,i = l,2,……,他,见=15,Ni是被试者数目;j = 1,2,......,N2,N2 = 32,N2是导联数目,k = 1,2,......,N3,N3 = 1024, N3是数据点数。 [0036] (3) respectively of the two fragments were induced to baseline EEG overlapped and averaged to obtain each lead in each subject's ERP both stimulation signals, denoted as X = {Xijk} and Y = {Yijk}, where, i = l, 2, ......, he see = 15, Ni is the number of subjects; j = 1,2, ......, N2, N2 = 32, N2 is the number of leads, k = 1,2, ......, N3, N3 = 1024, N3 is the number of data points.

[0037]由多次重复刺激下的头皮脑电信号叠加平均得到ERP信号的依据是:自发脑电是一种类随机信号,多次叠加可使自发脑电幅值降低;而ERP信号具有明显的锁时特性,多次叠加可使ERP信号幅值增加。 [0037] EEG from the scalp under repeated stimulation averaged superposition signal is based ERP: EEG signal is a random type, multiple stacking can EEG amplitude reduction; and ERP signal having significant when the characteristics of the lock, multiple stacking can increase ERP signal amplitude.

[0038] 3)对两种刺激下的ERP信号进行配对样本T检验,确定具有显著差异的时段; [0038] 3) of the ERP signal under two stimulus paired samples T test to determine significant differences in period;

[0039] 由10次重复刺激下的头皮脑电信号求平均得到的ERP信号中,自发脑电的幅值仍然很大,因此,各个ERP成分并不突出,无法进行ERP成分幅值和潜伏期的提取,也无法在不同种类刺激之间进行对比。 [0039] ERP scalp EEG signal from the demand stimulation in 10 replicates averaged, the amplitude of the spontaneous EEG is still large, therefore, the individual components are not projecting ERP, ERP components can not be amplitude and latency extraction, can not be compared between different types of stimuli. 由于不同种类刺激下的自发脑电虽然不是完全一致,但也不具有显著的差异性,因此,通过对不同种类刺激下的ERP信号进行T检验,得到的显著差异时段必为两真实ERP信号具有显著差异的时段,进一步实现了ERP信号与自发脑电的剥离。 Since EEG under different kinds of stimuli, although not exactly the same, but do not have significant differences, and therefore, by T-test on the ERP signals at different kinds of stimuli, significant differences period obtained will have two real ERP signal significant time difference, to achieve a further release of ERP EEG signal.

[0040] 所述的对两种刺激下的ERP信号进行配对样本T检验,是对每个导联中的每个数据点对应的ERP序列进行配对样本T检验,由于是同一批被试在同一个实验中接受的两种刺激,所以选用配对样本T检验,显著性水平设置为0.05。 ERP signals to the stimulation of the two [0040] paired samples T test, ERP is a sequence for each data point in each lead corresponding paired samples T test, because it is the same subjects in the same receiving a two stimulation experiments, so the selection of paired samples T test, significance level was set to 0.05. 对于第j个导联第k个数据点,首先建立一个新变量Z= {Zijk},Zijk = Xijk-Yijk,i = l,2,……,仏,计算新变量的均值 For the k-th data points leads the j-th, first create a new variable Z = {Zijk}, Zijk = Xijk-Yijk, i = l, 2, ......, Fo, mean calculation of new variables

Figure CN104545900BD00071

和方差 And variance

Figure CN104545900BD00072

、构造统计1 , Construction statistics 1

Figure CN104545900BD00073

&验tjk是否服从自由度为Ni_l 的T分布,并计算出对应的显著程度Pjk,若Pjk〈0.05,则序列,…,'沁和序列…,4,#}具有显著性差异,即两种刺激下的第j个导联第k个数据点具有显著性差异,否则,两种刺激下的第j个导联第k个数据点不具有显著性差异。 & Inspection tjk whether to obey degree of freedom Ni_l T of the distribution, and calculate significant degree corresponding PJK, if Pjk <0.05, then the sequence, ..., 'Qin and sequences ..., 4, #} having significant difference, i.e., two kinds of leads the j-th k-th data point in the stimulation of significant differences, otherwise, lead j-th k-th data in the two stimulation points have no significant difference.

[0041] 所述的确定具有显著差异的时段,是在已确定两种刺激下的第j个导联第k个数据点是否具有显著性差异的基础上进行的,包含如下过程: [0041] The period of time determined to have a significant difference, is performed in the determined lead j-th k-th data point in the two stimulation whether significant differences on the basis of the following process comprising:

[0042] (1)对于第j个导联的N3个数据点,若存在10个以上的连续数据点k,使得序列夂…,和序列伏;*..,K#.,…,心#丨具有显者性差异,那么这些连续的数据点k对应的时段就是第j个导联在两种刺激下的ERP信号具有显著差异的时段,若不存在10个以上的连续数据点k,使得序列^^和序列心,•••+#}具有显著性差异,则第j个导联在两种刺激下的ERP信号不具有显著差异时段; [0042] (1) N3 for the j-th data points leads, if more than 10 consecutive data points k exists, such that the sequence ... Fan, sequence and V; * .., K #, ..., # heart. Shu period with significant differences were, then k successive data points corresponding to the j-th period is significantly different leads in the ERP under two stimulus signal, if there are more than 10 consecutive data points k, such that ^^ sequences and sequences heart, ••• + #} significant difference, the j-th ERP signal leads in both the stimulation period no significant difference;

[0043] (2)分别对每个导联的N3个数据点进行配对样本T检验,每个导联可能含有多个显著差异时段,也可能不含有显著差异时段。 [0043] (2) were test for paired samples T N3 data points each lead, each lead may contain a plurality of significantly different time, or may not contain a significant difference period. 根据所有导联的显著差异时段分布,选择一个以上相对大的时段,使得尽可能包含多数导联的显著差异时段,所选的显著差异时段对应的数据点集标记为V™,丨,其中,r=l,2,…,m,m为所选的显著差异时段数,可选范围为{1, 2,…,100},nr分别是每个显著差异时段所对应的数据点数,取值均大于10。 The significant difference period all leads distribution, selecting one or more relatively large period, so that significant differences period possible comprises a plurality of leads, the selected set of data points significantly different period corresponding to the mark V ™, Shu, wherein r = l, 2, ..., m, m is the number of significant differences in the time period selected, the selectable range of {1, 2, ..., 100}, nr are significantly different for each time period corresponding to the number of data points, the value were more than 10.

[0044] 4)计算两种刺激下的ERP信号在显著差异时段内的差异面积,并绘制脑地形图,确定差异脑区。 [0044] 4) Calculate the difference in the area under the two ERP signal significant differences in the stimulation period, and draw brain mapping, determine the difference brain regions.

[0045] 所述的计算两种刺激下的ERP信号在显著差异时段内的差异面积,是对每位被试者的每个导联均计算两种刺激下的ERP信号在所选显著差异时段内的差异面积,其中,在显著差异时段试^}内的两种刺激下的ERP信号的差异面积3 ERP signal computed for two stimulation [0045] The difference in the areas of significant difference in time, in the ERP signal are calculated for each lead of each of the two stimuli subjects significant difference in the selected time period differences in the area, wherein a significant difference in the period of the test area ERP ^} difference signal in both stimulating 3

Figure CN104545900BD00081

并对所有被试者的数据进行叠加平均,得到第j个导联在所选显著差异时段内两种刺激下的ERP fg号差异面积t All subjects and data are overlapped and averaged, to obtain a j-th difference leads ERP fg number t at the selected area of significant difference between the two stimuli period

Figure CN104545900BD00082

[0046] 所述的绘制脑地形图、确定差异脑区,是对于每一个显著差异时段,均根据所有导联的两种刺激下的ERP信号差异面积绘制脑地形图,进而分析两种刺激所诱发ERP信号差异的空间分布状况,得到在显著差异时段内的主要激活脑区分布。 Brain mapping of the drawing [0046], a difference brain, is a significant difference for each period, according to both brain mapping drawing ERP signal in the two stimulation leads all the difference area, further analysis of the two stimuli ERP signal induced spatial distribution of differences to obtain the distribution of the major brain regions in the activation significant difference period.

[0047] 脑电地形图是一种集中表达大脑电生理信息的图形技术,通常将多个导联的单个特征用不同颜色映射其值大小而得到的头部平面彩色图形(或灰度差图像),能比较直观地反应大脑神经活动的波幅和分布。 [0047] BEAM is a graphical techniques to express the brain electrical physiological information, wherein the plurality of individual generally leads mapped magnitude values obtained by color graphics plane head (or grayscale image with a different color difference ), the reaction can be more intuitive and amplitude distribution of brain activity.

[0048]尽管上面结合附图对本发明的优选实施例进行了描述,但是本发明并不局限于上述的具体实施方式,上述的具体实施方式仅仅是示意性的,并不是限制性的。 [0048] Although the drawings of the embodiments have been described preferred embodiments of the present invention, but the present invention is not limited to the above specific embodiments in conjunction with the specific embodiments are merely illustrative, and not restrictive.

[0049] 本领域的普通技术人员在本发明的启示下,在不脱离本发明宗旨和权利要求所保护的范围情况下,还可以作出很多形式,这些均属于本发明的保护范围之内。 [0049] Those of ordinary skill in the art in light of the present invention, the scope of the present invention without departing from the spirit and the protection of the claims, can be made in many forms, which fall within the scope of the present invention.

Citations de brevets
Brevet cité Date de dépôt Date de publication Déposant Titre
CN102778949A *14 juin 201214 nov. 2012天津大学Brain-computer interface method based on SSVEP (Steady State Visual Evoked Potential) blocking and P300 bicharacteristics
CN103092971A *24 janv. 20138 mai 2013电子科技大学Classification method used in brain-computer interfaces
*US2002/0103428 Titre non disponible
*US2006/0036153 Titre non disponible
US5331969 *2 août 199026 juil. 1994Swinburne LimitedEquipment for testing or measuring brain activity
US5709214 *2 mai 199620 janv. 1998Enhanced Cardiology, Inc.PD2i electrophysiological analyzer
Classifications
Classification internationaleA61B5/0476
Événements juridiques
DateCodeÉvénementDescription
29 avr. 2015C06Publication
27 mai 2015C10Entry into substantive examination
22 févr. 2017C14Grant of patent or utility model