阅读说明：本技术 一种脑电波信号的滤波装置及方法 (Filtering device and method for brain wave signals ) 是由 廖志贤 石佳怡 谭承恒 莫胜胜 刘前虔 李振盛 于 2019-12-10 设计创作，主要内容包括：本发明公开一种脑电波信号的滤波装置及方法,由电极头阵列、混沌信号电路、信号采样保持与预处理模块、DSP信号处理器和数字信号可移动窗口组成。本发明解决了由于穿戴式检测传感器在人的运动过程中导致的传感器短时间移位、接触不良导致的脑电波信号突变问题,提高信号的有效性、可靠性和准确性,具有很高的实用价值和创新性。(The invention discloses a brain wave signal filtering device and a brain wave signal filtering method. The invention solves the problem of brain wave signal mutation caused by short-time displacement and poor contact of the wearable detection sensor in the human motion process, improves the effectiveness, reliability and accuracy of the signal, and has high practical value and innovation.)

1. A filtering device for brain wave signals comprises a filtering device body and is characterized by consisting of an electrode tip array, a chaotic signal circuit, a signal sampling, holding and preprocessing module, a DSP signal processor and a digital signal movable window;

the electrode tip array forms the input end of the acquisition filtering device body and is connected with the human brain; the output ends of the electrode tip array and the chaotic signal circuit are simultaneously connected with the input end of the signal sampling holding and preprocessing module; the output end of the signal sampling, holding and preprocessing module is connected with the input end of the DSP signal processor; the output end of the DSP signal processor forms the output end of the filtering device body;

the storage input end of the DSP signal processor is connected with the input end of the digital signal movable window, the output end of the digital signal movable window is connected with the storage output end of the DSP signal processor, and the control end of the DSP signal processor is connected with the control end of the digital signal movable window.

2. The apparatus and method for filtering brain wave signals according to claim 1, wherein the digital signal movable window is composed of a FIFO chip array, an analog switch array and a high-speed memory chip; the input end of the FIFO chip array forms the input end of the digital signal movable window, the output end of the FIFO chip is connected with the input end of the high-speed storage chip through the analog switch array, and the output end of the high-speed storage chip forms the output end of the digital signal movable window.

3. The apparatus and method for filtering brain wave signals according to claim 2, wherein the number of electrode heads included in the electrode head array, the number of FIFO chips included in the FIFO chip array, and the number of analog switches included in the analog switch array are the same.

4. The apparatus and method for filtering brain wave signals according to claim 1, wherein the chaotic signal circuit is composed of resistors R2-R3, a filter inductor L2, and filter capacitors C11-C12; after the filter capacitor C12 and the filter inductor L2 are connected in parallel, one end of the filter capacitor C12 is connected with one end of the filter capacitor C11 through a resistor R3, and the other end of the filter capacitor C11 is directly connected with the other end of the filter capacitor C11; two ends of the resistor R2 are connected in parallel with two ends of the filter capacitor C11; and the output end of the chaotic signal circuit is formed by the connection ends of the resistor R2, the resistor R3 and the filter capacitor C11.

5. The filtering method of brain wave signals implemented by the filtering device of claim 1, comprising the steps of:

step 1, collecting multi-dimensional original brain wave signals by an electrode tip array, and sending the signals to a signal sampling, holding and preprocessing module;

step 2, the chaotic signal circuit generates a random chaotic signal, adds the chaotic signal into the multidimensional original brain wave signal and sends the chaotic signal into a signal sampling, holding and preprocessing module;

step 3, a signal sampling, holding and preprocessing module samples and preprocesses the multi-dimensional original brain wave signals with the chaotic signals, filters out the chaotic signals to obtain multi-dimensional enhanced brain wave signals, and inputs the signals into a DSP signal processor;

step 4, the DSP signal processor sends the multi-dimensional enhanced brain wave signals into a digital signal movable window; under the control of the DSP signal processor, the digital signal movable window performs window movement extraction on the multi-dimensional enhanced brain wave signals to obtain enhanced brain wave signals of all electrode tips of the electrode tip array in the current window;

and 5, calculating the probability p of the enhanced brain wave signal of the nth section of the mth electrode tip in the current window appearing in the enhanced brain wave signals of the nth section of all the M electrode tips in the current window by the DSP signal processor according to the enhanced brain wave signals of all the electrode tips of the electrode tip array extracted by the digital signal movable window in the current window_mn(ii) a Wherein:

step 6, the DSP signal processor according to the probability p_mnCalculating the comprehensive entropy value of the enhanced brain wave signals of all M electrode heads of the electrode head array in the current window

step 7, the DSP signal processor carries out processing according to the comprehensive entropy valueCalculating an effective integrated entropy value

the nth section of the enhanced brain wave signal of the mth FIFO chip of the W [ m ] [ n ] digital signal movable window; m is 1, 2, 3.. M, wherein M represents the number of FIFO chips in the FIFO chip array, namely the number of electrode terminals of the electrode terminal array; n1, 2, 3.. N, N denotes a movable window length of a FIFO chip in the FIFO chip array; τ denotes a gain coefficient.

Technical Field

The invention relates to the technical field of brain wave processing, in particular to a brain wave signal filtering device and method.

Background

An Electroencephalogram (EEG) is a method for recording brain activity by using electrophysiological indexes, wherein the electrophysiological activity of brain nerve cells is totally reflected on the surface of a cerebral cortex or a scalp, the frequency variation range is 1-30 times per second, and the frequency variation range can be divided into wave bands such as delta (1-3 Hz), theta (4-7 Hz), α (8-13 Hz), β (14-30 Hz) along with the frequency variation, and the EEG is characterized by low signal frequency, small intensity and unstable amplitude range.

Currently, the acquisition of brain wave signals is often achieved by attaching an electrode tip to the brain of a human. However, in the process of collecting the brain wave signals, the body of a tester moves or the electrode head is not firmly contacted with the skin, so that the electrode head moves suddenly, and the collected brain wave signals are in a sudden frequency range, which brings great interference to the subsequent research and analysis of the brain wave signals. However, the filtering methods commonly used at present, such as wiener filtering, kalman filtering, FIR filtering, etc., do not meet ideal requirements for the filtering effect of brain waves, which is a special signal, and are low in efficiency, and are not suitable for processing brain wave signals.

Disclosure of Invention

The invention aims to solve the problem that the existing filtering method is not suitable for processing a special signal such as brain waves, and provides a filtering device and a filtering method for brain wave signals.

In order to solve the problems, the invention is realized by the following technical scheme:

a filtering device for brain wave signals comprises a filtering device body, a signal processing module and a digital signal processing module, wherein the filtering device body consists of an electrode tip array, a chaotic signal circuit, a signal sampling holding and preprocessing module, a DSP signal processor and a digital signal movable window;

the electrode tip array forms the input end of the acquisition filtering device body and is connected with the human brain; the output ends of the electrode tip array and the chaotic signal circuit are simultaneously connected with the input end of the signal sampling holding and preprocessing module; the output end of the signal sampling, holding and preprocessing module is connected with the input end of the DSP signal processor; the output end of the DSP signal processor forms the output end of the filtering device body;

the storage input end of the DSP signal processor is connected with the input end of the digital signal movable window, the output end of the digital signal movable window is connected with the storage output end of the DSP signal processor, and the control end of the DSP signal processor is connected with the control end of the digital signal movable window.

In the scheme, the digital signal movable window consists of an FIFO chip array, an analog switch array and a high-speed storage chip; the input end of the FIFO chip array forms the input end of the digital signal movable window, the output end of the FIFO chip is connected with the input end of the high-speed storage chip through the analog switch array, and the output end of the high-speed storage chip forms the output end of the digital signal movable window.

In the above scheme, the number of the electrode heads included in the electrode head array, the number of the FIFO chips included in the FIFO chip array, and the number of the analog switches included in the analog switch array are the same.

In the scheme, the chaotic signal circuit consists of resistors R2-R3, a filter inductor L2 and filter capacitors C11-C12; after the filter capacitor C12 and the filter inductor L2 are connected in parallel, one end of the filter capacitor C12 is connected with one end of the filter capacitor C11 through a resistor R3, and the other end of the filter capacitor C11 is directly connected with the other end of the filter capacitor C11; two ends of the resistor R2 are connected in parallel with two ends of the filter capacitor C11; and the output end of the chaotic signal circuit is formed by the connection ends of the resistor R2, the resistor R3 and the filter capacitor C11.

The method for filtering the brain wave signal, which is realized by the square filtering device, comprises the following steps:

step 1, collecting multi-dimensional original brain wave signals by an electrode tip array, and sending the signals to a signal sampling, holding and preprocessing module;

step 2, the chaotic signal circuit generates a random chaotic signal, adds the chaotic signal into the multidimensional original brain wave signal and sends the chaotic signal into a signal sampling, holding and preprocessing module;

step 3, a signal sampling, holding and preprocessing module samples and preprocesses the multi-dimensional original brain wave signals with the chaotic signals, filters out the chaotic signals to obtain multi-dimensional enhanced brain wave signals, and inputs the signals into a DSP signal processor;

step 4, the DSP signal processor sends the multi-dimensional enhanced brain wave signals into a digital signal movable window; under the control of the DSP signal processor, the digital signal movable window performs window movement extraction on the multi-dimensional enhanced brain wave signals to obtain enhanced brain wave signals of all electrode tips of the electrode tip array in the current window;

and 5, calculating the probability p of the enhanced brain wave signal of the nth section of the mth electrode tip in the current window appearing in the enhanced brain wave signals of the nth section of all the M electrode tips in the current window by the DSP signal processor according to the enhanced brain wave signals of all the electrode tips of the electrode tip array extracted by the digital signal movable window in the current window_mn(ii) a Wherein:

step 6, the DSP signal processor according to the probability p_mnCalculating the comprehensive entropy value of the enhanced brain wave signals of all M electrode heads of the electrode head array in the current window

Wherein:

step 7, the DSP signal processor carries out processing according to the comprehensive entropy value

Calculating an effective integrated entropy value

Then the effective comprehensive entropy value is obtainedAnd a judgment threshold value

And (3) comparison: when effective integrated entropy value

Greater than the judgment threshold

When the current window is not used, the enhanced brain wave signals of all the electrode heads of the electrode head array in the current window are discarded; otherwise, the enhanced brain wave signals of all the electrode heads of the electrode head array in the current window are reserved and filtered to be output as final brain wave signals; wherein:

the nth section of the enhanced brain wave signal of the mth FIFO chip of the W [ m ] [ n ] digital signal movable window; m is 1, 2, 3.. M, wherein M represents the number of FIFO chips in the FIFO chip array, namely the number of electrode terminals of the electrode terminal array; n1, 2, 3.. N, N denotes a movable window length of a FIFO chip in the FIFO chip array; τ denotes a gain coefficient.

Compared with the prior art, the invention better solves the problem of EEG (brain wave) signal mutation caused by short-time displacement and poor contact of the wearable detection sensor in the motion process of a person, improves the validity, reliability and accuracy of signals, and has very high practical value and innovation.

Drawings

Fig. 1 is a schematic diagram of a filtering apparatus for brain wave signals.

Fig. 2 is a flowchart of a method of filtering brain wave signals.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to specific examples.

Referring to fig. 1, the collecting and filtering device for collecting brain waves comprises an electrode tip array, a chaotic signal circuit, a signal sampling, holding and preprocessing module, a DSP signal processor, an FIFO chip array, an analog switch array and a high-speed memory chip.

The number of the electrode heads contained in the electrode head array, the number of the FIFO chips contained in the FIFO chip array and the number of the analog switches contained in the analog switch array are the same. In the preferred embodiment of the present invention, the number of FIFO chips included in the FIFO chip array and the number of analog switches included in the analog switch array are both 9.

The chaotic signal circuit is a Chua's circuit structure consisting of resistors R2-R3, a filter inductor L2 and filter capacitors C11-C12. After the filter capacitor C12 and the filter inductor L2 are connected in parallel, one end of the filter capacitor C11 is connected to one end of the filter capacitor C3526 via a resistor R3, and the other end of the filter capacitor C11 is directly connected to the other end of the filter capacitor C11. Two ends of the resistor R2 are connected in parallel with two ends of the filter capacitor C11. And the output end of the chaotic signal circuit is formed by the connection ends of the resistor R2, the resistor R3 and the filter capacitor C11.

The FIFO chip array, the analog switch array and the high-speed storage chip form a digital signal movable window. Under the control of the DSP signal processor, each dimensional brain wave signal is stored in one FIFO chip, namely, the multi-dimensional brain wave signals are split into multiple paths of signals which are respectively transmitted into the FIFO chips with corresponding number, the DSP signal processor can control the on-off of the analog switch array according to the entropy value of the signal sequence in each FIFO chip, and because the analog switches in the analog switch array are in one-to-one correspondence with the FIFO chips of the FIFO chip array, effective data can be selected to be transmitted into the high-speed storage chip by controlling the on-off of the analog switch array for being taken by the DSP signal processor.

The electrode tip array forms the input end of the collecting and filtering device and is connected with the human brain; the output ends of the electrode tip array and the chaotic signal circuit are simultaneously connected with the input end of the signal sampling holding and preprocessing module, and the output end of the signal sampling holding and preprocessing module is connected with the input end of the DSP signal processor; the output end of the DSP signal processor forms the output end of the acquisition filtering device. The storage input end of the DSP signal processor is connected with the input end of the FIFO chip, the output end of the FIFO chip is connected with the input end of the high-speed storage chip through the analog switch array, and the output end of the high-speed storage chip is connected with the storage output end of the DSP signal processor. And the control end of the DSP signal processor is connected with the control end of the analog switch array.

The electrode head array is attached to the human brain and acquires multi-dimensional original brain wave signals. The chaotic signal circuit is utilized to generate random chaotic signals, the chaotic signals are added into the multidimensional original brain wave signals, according to the stochastic resonance theory, the weak signals are strengthened after the weak multidimensional original brain wave signals are added into proper random signals, and the strengthened multidimensional original brain wave signals carrying the random chaotic signals are sent to the signal sampling, holding and preprocessing module. In the signal sampling, holding and preprocessing module, a band-pass filter is adopted to filter the added random signals, an enhanced brain wave signal is left, and the enhanced brain wave signal is sent to the DSP signal processor through a high-speed serial communication bus. The DSP signal processor transmits brain wave signals into the digital signal movable window through a high-speed serial communication bus; the digital signal movable window carries out moving processing on brain wave signals and then transmits the brain wave signals into the DSP signal processor. And the DSP signal processor calculates a signal entropy value, and outputs the multidimensional brain wave signal after cleaning and filtering the brain wave signal according to the calculated signal entropy value.

Referring to fig. 2, the method for filtering brain wave signals implemented by the above-mentioned apparatus specifically includes the following steps:

step 1, an electrode tip array collects original multi-dimensional brain wave signals and sends the signals to a signal sampling, holding and preprocessing module;

step 2, the chaotic signal circuit generates a random chaotic signal, adds the chaotic signal into the original multidimensional brain wave signal and sends the chaotic signal into a signal sampling, holding and preprocessing module;

step 3, a signal sampling, holding and preprocessing module samples original multidimensional brain wave signals with chaotic signals and preprocesses hardware layers, removes the chaotic signals in the original multidimensional brain wave signals to obtain enhanced multidimensional brain wave signals, and inputs the enhanced multidimensional brain wave signals into a DSP (digital signal processor) by utilizing a high-speed serial communication bus;

step 4, the DSP signal processor sends the multi-dimensional enhanced brain wave signals into a digital signal movable window; under the control of the DSP signal processor, the digital signal movable window performs window movement extraction on the multi-dimensional enhanced brain wave signals to obtain enhanced brain wave signals of all electrode tips of the electrode tip array in a window with the length of N.

The DSP signal processor controls the on-off of the analog switch in the analog switch array in real time so as to store the enhanced brain wave signals in the corresponding FIFO chip in the FIFO chip array into the high-speed storage chip. Two-dimensional array W [ M ] of M × N][N]For digital signal sequence, it is first-in first-out, i.e. movable, and we refer to it as digital signal movable window W M][N]. Using FIFO chip arrays, a two-dimensional array of M N W [ M ] is created][N]The array is a first-in-first-out (FIFO) structured buffer. Two-dimensional array W [ M ]][N]In, W [ m ]][n]Are elements thereof. M represents the quantity of FIFO chips in the FIFO chip array, and every FIFO chip of FIFO chip array is used for depositing the real-time reinforcing brain wave signal of an electrode tip of electrode tip array, and M also represents the total number of electrode tips, and M is the row sequence number of array, represents the mth FIFO chip, and M1, 2, 3. N denotes the length of the movable window of the FIFO chip, N is the column number of the array, and represents the sampling point number of the enhanced brain wave signal in the movable window of the FIFO chip, where N is 1, 2, 3. Let the real-time digital signal of the mth electrode tip at the z-th time (initial time, in this embodiment, let z equal to 0) be x_mzAnd loading the digital signal sequence after the z time into a two-dimensional array W [ M [ ]][N]I.e. x_m(z+n)Load into two-dimensional array W [ M ]][N]W [ m ] of][n]In the elements.

And 5, utilizing the enhanced brain wave signals of each electrode head of the electrode head array in the window with the length of N extracted by the DSP signal processor through the digital signal movable window, wherein the movable window comprises a plurality of sections of enhanced brain wave signals, and calculating the enhanced brain wave signals of the nth section of the M electrode heads in the window appear in the enhanced brain wave signals of the nth section of all the M electrode heads in the windowProbability p of_mn：

At this time, the uncertainty probability f (P) of the nth segment of brain wave signal of the mth electrode tip in the window_mn)：

Step 6, comprehensively calculating the digital signals of all the electrode tips in the window to obtain the comprehensive entropy values of all the electrode tips at the current moment;

step 6.1, calculating the entropy value E of the nth segment of brain wave signals of all the M electrode heads_n：

Step 6.2, calculating the information entropy redundancy D of the nth section brain wave signals of all the M electrode heads_n：

D_n＝1-E_n

Step 6.3, calculating the weight G of the nth section signals of all the M electrode heads in all the N sections of brain wave signals_n：

In particular 0. ltoreq. W_nLess than or equal to 1 and

step 6.4, calculating the comprehensive entropy value of the N-th section of brain wave signals of all the M electrode heads

Step 7, the above steps are performed

After multiplying by a coefficient tau, calculating the effective comprehensive entropy value by an activating function Sigmoid function

Step 5.8, output result of the signal entropy value algorithm module

Sending the signals into a signal cleaning module to perform the following treatment: setting a judgment threshold

When in use

A value of greater than ζ

The signal contained in the window is contaminated and is discarded; otherwise, the signal contained in the window is judged to be effective and can be sent to the MIMO digital filter module and output.

And (5) repeating the steps 1 to 5 along with the time, processing the collected brain wave signals, and improving the collection and analysis accuracy of the brain wave signals.

It should be noted that, although the above-mentioned embodiments of the present invention are illustrative, the present invention is not limited thereto, and thus the present invention is not limited to the above-mentioned embodiments. Other embodiments, which can be made by those skilled in the art in light of the teachings of the present invention, are considered to be within the scope of the present invention without departing from its principles.