CN103425983A - Brain network topology difference fast extracting method based on network synchronicity - Google Patents

Abstract

Translated fromChinese

本发明公开了一种基于网络同步性的脑网络拓扑差异的快速提取方法，利用本发明提出的基于网络同步性的脑网络拓扑差异的快速提取技术，可以显著地减少工作量和时间，同时很快地就能找出脑网络拓扑结构上存在的差异，整个过程能够快速准确的构建出差异连接图，64导的脑功能网络几秒钟就能完成。该方法通过基本计算脑电各导联时间序列间的同步性系数构建同步系数矩阵，然后利用脑网络自身特性阈值化为二值网络，然后利用卡方检验和交叉验证构建一幅拓扑结构差异连接图。这样可以减少繁琐的复杂的网络分析，缩短实验时间和成本，同时与常规的网络差异连接图进行对比发现差异连接更加的明显，更便于观察。

The invention discloses a method for quickly extracting topological differences of brain networks based on network synchronization. Using the technology for quickly extracting topological differences of brain networks based on network synchronization, the workload and time can be significantly reduced, and at the same time, the The difference in the topological structure of the brain network can be quickly found out, and the whole process can quickly and accurately build a difference connection map, and the 64-channel brain function network can be completed in a few seconds. This method builds a synchronization coefficient matrix by basically calculating the synchronization coefficient between the time series of EEG leads, and then uses the characteristics of the brain network to threshold it into a binary network, and then uses chi-square test and cross-validation to construct a topological difference connection picture. This can reduce cumbersome and complicated network analysis, shorten experiment time and cost, and compare with the conventional network differential connection diagram to find that the differential connection is more obvious and easier to observe.

Description

A kind of rapid extracting method of brain network topology difference of synchronism Network Based

Technical field

The present invention is a kind of rapid extraction technology of brain network topology difference, belongs to the nerve information technical field, is specifically related to a kind of rapid extraction technology of brain network topology difference of synchronism Network Based.

Background technology

Eeg collection system be a kind of utilize sophisticated electronics in the scalp position without wound the brain cell group's that records spontaneity, rhythmicity electrical activity; It is human body brain tissue bioelectric to be amplified to a special kind of skill of record, be mainly used in the inspection of the nervous system disease, due to its reflection be the brain function state of " work ", so play an important role for the diagnosis of the nervous system disease in the past few decades always.

In the past ten years, researchers utilize the technology such as magnetic resonance and brain electricity to carry out the research of brain network, and wherein the brain function network is exactly an important component part in the brain network research.A lot of early-stage Study has shown that the 26S Proteasome Structure and Function network of human brain all has the worldlet topological property.Yet on the brain network statistics except calculating the difference that they exist on network parameter, can not find out well very soon the brain network topology structure upper difference existed that distributes.Here use the rapid extraction technique construction of the brain network topology difference of synchronism Network Based to go out the difference connection layout, thereby observed out the difference existed on topological structure between the difference in functionality network intuitively.

Yet, for any two brain networks, due to suitable intensive of the connection of brain network itself, traditional method can not be found out the difference existed between two kinds of brain function networks well.Routine only has the graph theory parameter by calculating two brain function networks to compare whole network.For the problems referred to above, although having proposed various parameters, different researcher goes to weigh the difference existed between two kinds of networks.Yet be substantially all for for whole network when calculating these parameters, can not reflect well some features of localized network.And more loaded down with trivial details consuming time on these parameters of calculating, be not easy to understand, can not show intuitively the difference existed between the brain network.

In order intuitively to find out quickly the difference existed between any two brain function networks, the inconvenience of avoiding the network Connection Density to bring greatly, we utilize the synchronism between time series to construct the synchronization factor matrix; Then by the characteristic of network self, come the importance on limit in supervising network (node) to find out the limit (node) of conspicuousness; Secondly construct in the mode with cross validation again the difference connection layout existed between two networks.There are differences thereby found out effectively rapidly on the brain network topology structure, observe more intuitively some variations of network.

Summary of the invention

The objective of the invention is the difference zone existed on topological structure between two kinds of brain networks in order quick and precisely to orient, reduce the loaded down with trivial details of time of Analysis of Complex network and calculating, make it to be widely used in clinical, a kind of rapid extracting method of brain network topology difference of synchronism Network Based.

To achieve these goals, technical scheme of the present invention is:

The rapid extracting method of the brain network topology difference of synchronism Network Based comprises the following steps:

A. adopt the two dimension of under EEG measuring equipment records tranquillization state, opening eyes and closing one's eyes to lead measuring-signal more, this test data is carried out to pre-service (comprising the pre-service such as average reference, bandpass filtering); Extract the time series of after frequency division, respectively leading the brain electricity;

B. calculate to need analyze under different condition and respectively lead the synchronism coefficient between time series, construct two synchronism matrix of coefficients (what adopt is 64 to lead brain wave acquisition equipment, therefore the synchronism matrix of coefficients is 64*64) here;

C. the thresholding method of Adoption Network expense (cost) (or other threshold methods) carries out binaryzation by two kinds of synchronism matrix of coefficients to be analyzed, thereby obtains two kinds of need brain function network relatively;

D. utilize Chi-square Test to carry out statistical test to the importance on every limit (point) in two functional networks, importance limit (point) the record found out in two functional networks mark;

E. the importance on the limit (point) in two function brain networks is carried out to cross validation, find out the limit of significant difference between two functional networks, thereby construct the difference connection layout (significant limit (point) marks by different colors respectively) of brain network topology structure.

F: the difference connection layout constructed is carried out to analysis-by-synthesis, thereby find out the zone there are differences on the brain topological structure between two kinds of functional networks;

Beneficial effect of the present invention: the rapid extraction technology of the brain network topology difference of the synchronism Network Based of utilizing the present invention to propose, workload and time can be reduced significantly, just can find out the difference existed on the brain network topology structure soon simultaneously, whole process can construct the difference connection layout fast and accurately, and the 64 brain function networks of leading just can complete several seconds.The synchronism coefficient that the method is respectively led between time series by basic calculating brain electricity builds the synchronization factor matrix, then utilize brain network self-characteristic threshold value to turn to the two-value network, then utilize Chi-square Test and cross validation to build a width topological structure difference connection layout.Can reduce the network analysis of loaded down with trivial details complexity like this, shorten experimental period and cost, simultaneously with conventional network discrepancy connection layout contrast find differences be connected more obvious, the observation of being more convenient for.The method is utilized the difference existed on topological structure between network, for clinical treatment and diagnosis provide certain foundation.

The accompanying drawing explanation

Fig. 1 is main flow chart of the present invention.

Fig. 2 is the comparison that the present invention opens eyes under the tranquillization state with conventional method and is connected with the difference of each frequency range of closing one's eyes.

Embodiment

Below in conjunction with specific embodiment, the present invention is described in detail.

As shown in Figure 1, a kind of rapid extraction technology of brain network topology difference of synchronism Network Based comprises the following steps:

A. adopt 64 lead eeg collection system recorded under the tranquillization state within 3 minutes, closing one's eyes and 3 minutes eeg datas of opening eyes, sampling rate is 500Hz; The observation raw data of opening eyes and close one's eyes, intercept data preferably and carry out pre-service, obtains the corresponding eeg data of opening eyes and close one's eyes of each frequency range.

B. the eeg data of opening eyes and close one's eyes of pretreated each frequency range is used likelihood with footwork calculate respectively open eyes and closed-eye state under respectively lead the synchronism coefficient between time series, build synchronization factor matrix (what adopt is 64 to lead brain wave acquisition equipment, therefore the synchronism matrix of coefficients is 64*64) here;

C. the thresholding method of Adoption Network expense (cost) carries out binaryzation by the synchronism matrix of coefficients of each frequency range of under the tranquillization state, opening eyes and closing one's eyes, and obtains two states brain function network;

D. utilize Chi-square Test to carry out statistical test to the importance on every limit (point) in two functional networks, importance limit (point) the record found out in two functional networks mark;

E. the importance on the limit (point) in two function brain networks is carried out to cross validation, find out the limit of significant difference between two functional networks, thereby construct the difference connection layout (significant limit (point) marks by different colors respectively) of brain network topology structure.

F. analyzed at difference node diagram and the difference edge graph of Alpha frequency range opening eyes and closing one's eyes respectively, found out the zone that under the tranquillization state, the normal person opens eyes and closes one's eyes and there are differences between two states on the Alpha frequency range on the brain topological structure.

Effect for the brain function network discrepancy under the observation and comparison two states, Fig. 2 is the comparison at each frequency range difference connection layout of this method and conventional difference connection layout construction method, and the 1st row is the difference connection layout between opening eyes and closing one's eyes of obtaining by traditional analytical approach; Wherein A, B, C mean respectively Theta, the Alpha in EEG research, the difference connection layout of beta frequency range; The 2nd row is to adopt the inventive method to build the difference connection layout of corresponding Theta, Alpha, beta frequency range, know that from figure the inventive method more significant effective extracted two internetwork differences and connected, better reflect the difference existed between network on topological structure.Can find that method of the present invention is quicker, accurately, effectively, and the inventive method is convenient to not possess people's observation of this respect knowledge more.Simultaneously, by the result after analyzing and the document contrast of report before, find that normal person's closed-eye state is more consistent in the middle diversity ratio of brain with the state of opening eyes under the tranquillization state.Illustrate that the inventive method can fast find out the difference existed between two kinds of functional networks really exactly, and effect is apparent in view.

Those of ordinary skill in the art will appreciate that, embodiment described here is in order to help reader understanding's principle of the present invention, should be understood to that protection scope of the present invention is not limited to such special statement and embodiment.Those of ordinary skill in the art can make various other various concrete distortion and combinations that do not break away from essence of the present invention according to these technology enlightenments disclosed by the invention, and these distortion and combination are still in protection scope of the present invention.

Claims (4)

1. the rapid extracting method of the brain network topology difference of a synchronism Network Based, is characterized in that, comprises the following steps:

A. adopt EEG measuring equipment records two dimension to lead measuring-signal more, this test data is carried out to pre-service, described pre-service comprises average reference, bandpass filtering; Extract the time series of after frequency division, respectively leading the brain electricity; Treat following analysis;

B. calculate open eyes and closed-eye state under respectively lead the synchronism coefficient between time series, obtain the synchronism matrix of coefficients;

C. the thresholding method of Adoption Network expense carries out binaryzation by two kinds of synchronism matrix of coefficients to be analyzed, thereby obtains two states brain function network;

D. utilize Chi-square Test to carry out statistical test to the importance of every limit in two functional networks or point, find out importance limit in two functional networks or point record and mark;

E. cross validation is carried out in the limit in two kinds of function brain networks or the importance of point, find out the limit of significant difference between two functional networks, thereby construct the difference connection layout of brain network topology structure, respectively significant limit or point are marked by different colors;

F: the difference connection layout constructed is carried out to analysis-by-synthesis, thereby find out the zone there are differences on the brain topological structure between two kinds of functional networks.

2. method according to claim 1, is characterized in that, the equipment of EEG measuring described in steps A is one of the EEG signals register system of 32 Dao, 64 Dao, 128 Dao, 256 road electrodes of standard.

3. method according to claim 1, is characterized in that, step B adopts is 64 to lead brain wave acquisition equipment, and the synchronism matrix of coefficients is 64*64.

4. method according to claim 1, is characterized in that, what described in step C, process employing is the Cost threshold method.

Images