CN107024987A - A kind of real-time human brain Test of attention and training system based on EEG - Google Patents

Abstract

The invention discloses a real-time human brain attention test and training system based on EEG. The system includes five parts: attention experiment, signal collection, data analysis, real-time transmission and test feedback. The attention experiment part is divided into system internal and external In the experiment, the signal acquisition part uses the EEG acquisition equipment to collect the EEG data of the user; the data analysis part uses the data analysis program to denoise, filter and analyze the relevant rhythm waves of the collected signal; the real-time transmission part analyzes the quantified value Save it for extraction at any time, and transmit it through the corresponding interface. The test feedback part uses the corresponding program to read the data of the real-time transmission part, and realizes the feedback through a visual interface. The present invention effectively combines EEG signals with attention levels and presents them in the form of diversified experiments, which improves the fun of treatment and the sustainable time of treatment, and can effectively help people with attention deficits improve their attention levels .

Description

A real-time human brain attention test and training system based on EEG

technical field

The invention belongs to the comprehensive application of the fields of cognitive neuroscience, information technology and automatic control, and relates to a brain-computer interface (BCI) technology that uses the interaction between the human brain and a computer to test the user's attention level in real time and perform training for improvement.

Background technique

Electroencephalogram (Electroencephalograph, EEG) accompanies us all the time in our lives. It is the overall response of the spontaneous and rhythmic electrical activity of brain cell groups in the cerebral cortex and scalp, which can be detected by electrodes placed on the scalp. EEG can be divided into four rhythmic waves of δ, θ, α, and β according to different frequencies. Many foreign scholars and experts have found through a large number of experimental analyzes that the alpha band in the human brain wave is the main frequency of activity in the quiet and awake state. Children with ADHD show increased theta EEG activity and theta/beta power ratio, and decreased alpha and beta activity. Therefore, it is generally believed that increased theta slow wave activity, increased theta/beta power ratio, and decreased alpha and beta activity are the main features of attention decline, but other wave bands often have some effects.

BCI: Brain Computer Interface technology (Brain Computer Interface, BCI) is to collect the EEG signals generated by the activity of the cerebral cortex nervous system, and convert them into signals that can be recognized by computers through amplification, filtering and other methods, so as to distinguish the true nature of people. intention.

EEGLAB: This is a toolbox based on Matlab. It is mainly used to process continuously recorded electroencephalogram (EEG), magnetoencephalon (MEG) and other electrophysiological data. The methods it uses mainly include independent component analysis (ICA), time-frequency analysis, drawing ERP diagrams, removing artifacts, and several useful visualization modes (for averaging and single-shot data extraction), etc.

Existing brain-computer interface patent technology is rarely applied to the human brain attention test, and the existing patent technology only involves the training of attention (such as the patent application number CN201020206845), the evaluation of attention under the driving environment ( For example, the application number is the invention of CN201410381256), the real-time test and training system for human brain attention has not yet been disclosed by relevant patents.

Contents of the invention

The technical problem solved by the present invention is to provide a brain-computer interface system that realizes real-time testing and training of attention and has high speed and precision. The present invention analyzes the attention of the human brain with a comprehensive multi-feature method, feeds back the attention level in real time through a terminal such as a computer or a mobile phone, and performs attention training according to the feedback result. The system is highly accurate and has a certain degree of fun.

For this reason, the solution that the present invention proposes is a kind of real-time human brain attention test and training system based on EEG, and system comprises five parts of attention experiment, signal collection, data analysis, real-time transmission and test feedback, attention experiment part It is divided into internal and external experiments of the system. The signal acquisition part uses the EEG acquisition equipment to collect the EEG data of the user; the data analysis part uses the data analysis program to denoise, filter and analyze the relevant rhythm waves of the collected signals; the real-time transmission part The quantified value obtained from the analysis is saved for extraction at any time, and transmitted through the corresponding interface. The test feedback part uses the corresponding program to read the data of the real-time transmission part, and realizes the feedback through a visual interface.

Furthermore, the above-mentioned system external experiment is any experiment that can distinguish the degree of concentration of attention, and the user can decide by himself to play the role of analysis and detection.

There are various forms of attention experiments in the above-mentioned system, which are used to improve the interest of users. The attention experiments in the system can be fed back in real time. The state of each moment is affected by the attention level and can be clearly reflected to the user, so that Users make psychological hints to achieve the effect of improving attention.

As preferably, in the signal acquisition part, the EEG signal acquisition frequency may be 800-1200 Hz, the selected leads are Fp1, Fp2, F7, F3, Fz, F4 and F8, and the EEG signal acquisition equipment and data processing program are realized by programming An interface for real-time EEG data transmission.

The real-time transmission part is divided into two parts. The first part is to transmit the collected data to the data analysis part in real time, and the other part is to transmit the analysis results to the test feedback and attention experiment part.

As a preference, the real-time transmission of the collected data to the data analysis part is realized by BCI2000 software.

The part of transmitting the analysis results to the test feedback and attention experiment is to read the required data through the corresponding reading program inside the system, and the frequency of transmission is determined by the frequency of the collected signal.

The above-mentioned data analysis part processes the collected EEG data, and judges the degree of concentration of attention. The sequential processing is ICA denoising and artifact removal, filtering, and EEG signal attention-related feature extraction. ICA mainly completes the analysis of ECG, For the removal of oculoelectricity and random noise, the filter mainly realizes the removal of low-frequency, high-frequency and 50Hz power frequency interference noise, and separates the rhythmic waves of each frequency band to prepare for feature extraction. Feature extraction uses BP neural network multi-parameter Analytical method.

The data of the attention experiment is transmitted to the test feedback part, and is fed back to the user in real time through a visual interface.

Compared with prior art, the beneficial effect of the present invention:

1. The present invention can effectively help people with attention deficits improve their attention levels. The previous drug therapy for attention-deficit people has extremely side effects, but the present invention effectively combines EEG signals with attention levels and presents them in the form of diversified experiments, which improves the fun of treatment and further improves the efficacy of treatment. The duration is the time of concentration.

2. In the present invention, through real-time feedback, users can know their attention level in real time, so as to give themselves psychological hints to improve their attention.

3. The present invention plays a certain role in promoting the realization of low-cost and high-efficiency attention deficit treatment in the future, and also indicates the great potential of brain-computer interface in life and medical treatment.

Description of drawings

Fig. 1 is a flowchart of the present invention.

Figure 2 is a schematic diagram of the attention experiment part and the test feedback part.

Figure 3 shows the raw data reading results.

Figure 4 is the result after ICA treatment.

Figure 5 shows the results of the filtering process.

detailed description

The present invention will be described in detail below in conjunction with the accompanying drawings and specific examples.

The basic principle of the present invention is that when the user performs the attention experiment, the degree of concentration can be reflected by the rhythm wave in the EEG at this time. When the attention is concentrated, the θ brain electrical activity and the θ/β power ratio decrease, and the α and β activities increase. , so use multi-parameters to measure the EEG level as a whole, assign corresponding weights to each parameter, and finally quantify the attention level and give feedback. The specific analysis process will be described in detail in the data analysis section below.

The EEG-based real-time human brain attention test and training system proposed by the present invention includes the following parts: attention experiment part, EEG collection part, data analysis part, real-time transmission and test feedback part. The attention experiment is divided into system internal and system external experiments. The system external experiment is any experiment that can distinguish the degree of concentration. Helps increase the user's attention level. The signal acquisition part uses EEG acquisition equipment to collect the EEG data of the user, and transmits the data to the data analysis program in real time. The data analysis part uses the data analysis program to denoise, filter and analyze the relevant rhythm wave on the collected signal, so as to reflect the user's attention level. The real-time transmission part saves the quantified value obtained by analysis for extraction at any time, and transmits it through the corresponding interface. The test feedback part reads the data of the real-time transmission part with the corresponding program, and makes a corresponding response.

Attention experiments inside the system come in various forms and are used to increase user interest. There are: blooming flowers, growing leaves, sinking, etc. The common point is that the state of each moment of the game is affected by the attention level, and it can be clearly fed back to the user, so that the user can make psychological hints and achieve the effect of improving attention.

In the signal acquisition part, use Scan4.5 acquisition software, the EEG signal acquisition frequency can be 800 ~ 1200Hz, the selected electrodes are Fp1, Fp2, F7, F3, Fz, F4 and F8, programming to realize EEG signal acquisition equipment and data processing An interface for real-time EEG data transmission between programs.

The data analysis part processes the collected EEG data and judges the degree of concentration of attention. The sequential processing is ICA denoising and artifact removal, filtering, and attention-related feature extraction of EEG signals. ICA mainly completes the removal of ECG, EEG and random noise, etc. The filter mainly realizes the removal of low frequency, high frequency and 50Hz power frequency interference noise, and separates the rhythm wave of each frequency band to prepare for feature extraction.

The real-time transmission part is divided into two parts. The first part is to transmit the collected data to the analysis part in real time, and the other part is to transmit the analysis results to the test feedback and attention experiment part. The former is transmitted through the BCI2000 software, and the latter reads the required data through the corresponding reading program inside the system. The frequency of transmission is determined by the frequency of the acquisition signal.

The test feedback part is related to the attention experiment inside the system. The feedback of the attention test results is realized through a visual interface, mainly to visualize the data and feedback in the form of graphics.

A specific example is provided below to describe in detail the implementation of the present invention.

In this example, the EEG signal acquisition equipment adopts NeuroScan equipment, and the Scan4.5 software transmits the EEG signal data collected by the BCI2000 platform to the MATLAB software in real time for data processing.

Referring to Figure 1, the whole system includes five parts: attention experiment, EEG acquisition, data analysis, real-time transmission and test feedback

The attention experiment part and the test feedback part are shown in Figure 2. The attention experiment is divided into system internal and system external experiments. The system external experiment is any experiment that can distinguish the degree of attention concentration. Users can decide by themselves. If the system external experiment is selected, the system will directly open to the visual feedback part for real-time and quantitative feedback. level of concentration. If you choose the attention experiment inside the system, there will be game programs such as flower opening, leaf growth, and submersion. The state of each moment of the game is affected by the level of attention, and there is a visual feedback section. The flower opening experiment is selected as an example demonstration here.

Use the NeuroScan device to collect the user's EEG data in real time. The EEG signal acquisition frequency can be 1000Hz. Among them, since the characteristic potential of attention is mainly generated in the frontal area of the brain, the electrode cap is selected according to the "10-20 international standard leads". For the seven leads labeled Fp1, Fp2, F7, F3, Fz, F4 and F8, select the default positions on the electrode cap equipped with NeuroScan for the reference electrode and ground electrode. The EEG acquisition results of each channel are shown in Figure 3.

The data analysis part is mainly realized by MATLAB. After receiving the EEG signal data, MATLAB processes the EEG signal data collected in the previous 5 seconds every 5 seconds, and saves the data to a text file for real-time extraction by the transmission part.

The sequential processing of the collected EEG data is ICA denoising and artifact removal, filtering, and EEG signal attention feature extraction. The EEG data were collected for 5 seconds for each analysis.

(1) ICA denoising and de-artifacting

EEG signal is a highly random electrophysiological signal, and various emotions and mentalities will affect its changes. Therefore, the EEG signal has high time-varying sensitivity and is easily polluted by irrelevant noise, thus forming various EEG artifacts, among which the most influential ones are ECG and EEG artifacts. ICA mainly completes the removal of ECG, electrooculogram, and random noise. The advantage is that the components obtained by ICA processing not only remove the correlation, but also are statistically independent of each other. The theoretical knowledge is:

Step 1: Assume that the N-dimensional observation signal is Y(t), Y(t)=[y ₁ (t),y ₂ (t)......y _N (t)] ^T , including the acquired Various artifacts and noise components, S(t) are M mutually statistically independent source signals that generate the observed signal, S(t)=[s ₁ (t),s ₂ (t)...s _M (t)] ^T .

Step 2: The observation signal is generated by the source signal after being mixed linearly by the system, that is, Y(t)=BS(t), and B is the system matrix.

Step 3: When the hybrid system matrix B and the source signal S(t) are unknown, only using the assumption that the observed signal Y(t) and the source signal are statistically independent, find a linear transformation separation matrix D such that L(t) =DY(t)=DBS(t) is equal to the source signal S(t) as much as possible. At this time, the original S(t) signal can be approximately replaced by the finally obtained L(t) signal, and each component is equivalently replaced and separated.

In the example, it is considered that various potential differences and EEG signals are generated by independent sources, and are mixed instantaneously and linearly. The analysis results are shown in FIG. 4 . The abscissa represents time, and the ordinate represents EEG amplitude. (2) Filtering

The main purpose of the filter is to remove low-frequency, high-frequency and 50Hz power frequency interference noise, and to separate the rhythmic waves of each frequency band to prepare for feature extraction.

Low-frequency interference is mainly baseline drift, which is caused by poor contact between electrodes and human body, amplifier temperature drift or breathing during measurement. High-frequency interference is mainly radio frequency interference and myoelectric interference in acquisition. The Butterworth filter can be used for bandpass filtering. In MATLAB, the butter function and the filtfilt function can be called directly.

The method of removing 50Hz power frequency interference is to use a digital notch filter. In matlab, a self-designed Butterworth-type 50Hz notch filter function function[Num,Den]=ZB_50_filter (f0,B1,N), where f0, B1, N are the center frequency of the notch filter, the unilateral bandwidth and the order of the filter, respectively. This function has passed the verification of the fdatool toolbox.

The FIR digital filter is used to separate various rhythm waves, in which the frequency of delta wave is 1-4Hz, the frequency of theta wave is 4-7Hz, the frequency of alpha wave is 8-13Hz, and the frequency of beta wave is 13-20Hz. The separation results are shown in Figure 5.

(3) Feature extraction

In order to assess the level of attention accurately, the present invention adopts multi-feature parameters as a standard, specifically as follows:

W ₁ : the percentage of delta wave energy to the total energy of the EEG signal;

W ₂ : The percentage of theta wave energy to the total energy of the EEG signal;

W _α : the percentage of α wave energy in the total energy of EEG signal;

Rel: The ratio of θ wave energy to β wave energy;

P _β : absolute value of β energy;

f _max : the frequency point of maximum energy in the β wave.

A three-layer BP neural network is used for nonlinear fitting, the number of neurons in the input layer is N=6, the number of neurons in the output layer is K=2, and the number of neurons in the hidden layer M is based on the empirical formula:

Desirable M=5, P≈32, and the excitation function is a non-linear monotonically rising Sigmoid function. Set the EEG data of the concentration of attention collected in the early stage of the learning sample, and determine the weight of each parameter (between 0 and 1) through sample learning. The initial weight is set between 0.1 and 0.3, and an approximate The formula for calculating the concentration of attention, and calculate the range of values at the time of concentration. Afterwards, through a variety of traditional attention testing methods (such as the Schulte grid method), the values of different attention situations are determined, and then the data obtained by real-time collection and analysis is compared with the above data to reflect the attention level. The details are reflected in the visual feedback section.

The real-time transmission part is divided into two parts, the first part is to transmit the collected data to the analysis part in real time through the BCI2000 software, and the other part is to read the required data to the visual feedback and attention through the corresponding reading program inside the system Experimental part. The frequency of transmission is determined by the frequency of the acquisition signal, which should be 1000Hz in this example.

The test feedback part is related to the attention experiment inside the system. The feedback of the attention test results is realized through a visual interface, mainly to visualize the data feedback. The visualization adopts the feedback in the form of graphics, which is specifically shown in the attention experiment part explained.

It should be noted here that, in order to make the implementation examples more detailed, the above embodiments are preferred embodiments, and those skilled in the art can also adopt other alternative implementations for some known technologies; and the accompanying drawings are only for more specific description Examples are not intended to specifically limit the present invention.

The present invention is not limited to the specific technical solutions described in the above embodiments, and all technical solutions formed by equivalent replacement are the protection required by the present invention.

Claims (9)

1. A real-time human brain attention test and training system based on EEG is characterized in that it includes five parts: attention experiment, signal acquisition, data analysis, real-time transmission and test feedback, and the attention experiment part is divided into system interior and exterior In the experiment, the signal acquisition part uses the EEG acquisition equipment to collect the EEG data of the user; the data analysis part uses the data analysis program to denoise, filter and analyze the relevant rhythm waves of the collected signal; the real-time transmission part analyzes the quantified value Save it for extraction at any time, and transmit it through the corresponding interface. The test feedback part uses the corresponding program to read the data of the real-time transmission part, and realizes the feedback through a visual interface. 2. the real-time human brain attention test and training system based on EEG according to claim 1, it is characterized in that described system external experiment is the experiment that can distinguish the degree of concentration of attention arbitrarily, the user can decide by oneself, plays analysis and The role of detection. 3. the real-time human brain attention test and training system based on EEG according to claim 1 is characterized in that the attention experiment form inside described system is various, for improving user's interest, the attention experiment inside system can be Real-time feedback, the state of each moment is affected by the level of attention, and can be clearly reflected to the user, so that the user can give psychological hints to achieve the effect of improving attention. 4. the method for the real-time human brain attention test and training system based on EEG according to claim 1 is characterized in that in the signal acquisition part, the EEG signal acquisition frequency can be 800～1200Hz, and the lead of choosing is Fp1, Fp2, F7, F3, Fz, F4 and F8 realize the interface of real-time EEG data transmission between EEG signal acquisition equipment and data processing program through programming. 5. the real-time human brain attention test and training system based on EEG according to claim 1 is characterized in that being divided into two pieces in the real-time transmission part, the first piece is the real-time transmission of the data collected to the data analysis part, and the other One is to transmit the results of the analysis to the test feedback and attention experiment part. 6. the real-time human brain attention test and training system based on EEG according to claim 5, is characterized in that the real-time transmission of the data collected to the data analysis part is realized by BCI2000 software. 7. the real-time human brain attention test and training system based on EEG according to claim 5, it is characterized in that the result of the analysis is transmitted to the test feedback and the attention experiment part is read by the corresponding reading program inside the system Take the required data, the frequency of transmission is determined by the frequency of the acquisition signal. 8. the real-time human brain attention test and training system based on EEG according to claim 1, is characterized in that described data analysis part processes the EEG data of collection, judges the processing that the degree of concentration of attention carries out successively For ICA denoising and artifact removal, filtering, and EEG signal attention-related feature extraction, ICA mainly completes the removal of ECG, eye electricity and random noise, etc. The filter mainly realizes the removal of low frequency, high frequency and 50Hz power frequency Interference noise, and separate the rhythmic waves of each frequency band, in preparation for feature extraction, feature extraction uses BP neural network multi-parameter analysis method. 9. the real-time human brain attention test and training system based on EEG according to claim 1, is characterized in that the data of described attention experiment is transmitted to the test feedback part, and feeds back to the user in real time by the visual interface.