Internal brain stimulation and detection system and method

Abstract

The present invention relates to a brain internal stimulation and detection system and method, which belongs to the technical field of electroencephalogram data processing, and solves the problems of poor accuracy and poor real-time performance of electroencephalogram signals collected in the prior art. The system includes a processor and electrodes; the electrodes are used to collect electroencephalogram signals at the target brain position and transmit them to the processor; the processor is used to obtain the β-band signal based on the electroencephalogram signal, and perform Hilbert transform and sliding window integration on the β-band signal in turn to obtain brain observation characteristic values, and judge whether the brain state is normal based on the brain observation characteristic values; if it is determined that the brain state is abnormal, the processor is also used to control the electrode to generate a corresponding stimulation signal to act on the target brain position. The system can accurately collect electroencephalogram signals and make judgments based on the frequency domain characteristics of the signals, with high accuracy, fast data processing speed, and excellent real-time performance.

Description

Technical Field

[0001] The present invention relates to the technical field of electroencephalogram (EEG) data processing, and in particular to a brain internal stimulation and detection system and method. Background Art

[0002] Parkinson's disease is a common neurodegenerative disease. According to a survey by the World Health Organization, there are about 4 million Parkinson's disease patients worldwide, of which more than 2 million are in China. Parkinson's disease is characterized by pathological movement symptoms, including tremor, rigidity, bradykinesia and postural instability. Among them, some pathological symptoms of Parkinson's disease are caused by dysfunction of the thalamus region of the brain. With the development of sensor technology, the performance of medical instruments has gradually become intelligent and applied to clinical surgery and treatment of neurological diseases. Deep brain stimulation technology is a technology that implants electrodes into the patient's brain through surgery, collects intracranial EEG data to predict symptoms and identify abnormal signals, and releases corresponding electrical pulse sequences to the deep structures of the brain to achieve the purpose of treatment.

[0003] Local field potential technology is a reliable EEG signal detection method used to display integrated analog morphological synaptic signals input to the cortex. Oscillations are classified according to the range of the main frequency bands: δ-(1-4Hz), θ-(4-10Hz), α-(10-14Hz), β-(14-36Hz) and γ-(36-100Hz). Different frequency bands play different roles in brain functions such as memory and synchronous neural activity. There is also coupling between local field potentials in different frequency bands, which affects the pathological state of Parkinson's patients. With the accumulation of local field potential data, many biomarkers for closed-loop neural regulation have been proposed. Recent studies have shown that beta oscillations in the basal ganglia are closely related to the severity of Parkinson's symptoms. When dopamine is depleted, energy increases; while when patients exercise or use deep brain stimulation, energy decreases. Deep brain stimulation technology has been proven to be an effective method for treating movement disorders such as Parkinson's disease. Deep brain stimulation of the subthalamic nucleus and the internal part of the globus pallidus is an effective intervention to improve the functional performance of patients with advanced Parkinson's disease. The stimulation device modulates the activity of certain cell nuclei through implanted electrodes, thereby disrupting pathological oscillatory ganglion loops of the corresponding band within the cortical base.

[0004] In the prior art, on the one hand, one electrode is usually used to collect EEG signals, and another electrode is used to generate corresponding stimulation signals to act on the target brain position; in addition, in the prior art, whether the brain state is normal is judged only based on the time domain characteristics of the collected EEG signals; on the other hand, the energy spectrum of the β-band of the EEG signal is usually used to analyze and judge the brain state.

[0005] The prior art has at least the following defects: first, the stimulation electrode is at the target point and the signal collection electrode is under the cortex, and the two electrodes cannot be located at the target point at the same time, resulting in large noise and errors in the signal collected by the signal collection electrode. Even if the two electrodes are located at the target point at the same time, the stimulation signal generated by one electrode will greatly interfere with the EEG signal collected by the other electrode, making it difficult to ensure the accuracy of the collected EEG signal. Second, whether the brain state is normal is judged only based on the time domain characteristics of the collected EEG signal, ignoring the frequency domain characteristics in the EEG signal, reducing the accuracy of brain state judgment. Third, although the energy spectrum of the β-band of the EEG signal is used for analysis, it can accurately judge the brain state, but it cannot meet the timeliness requirements in actual applications. Summary of the invention

[0006] In view of the above analysis, the embodiments of the present invention aim to provide a brain internal stimulation and detection system and method to solve the problem that the EEG signals collected in the prior art are poor in accuracy and real-time performance, thereby leading to inaccurate brain internal detection results.

[0007] In one aspect, the present invention provides an internal brain stimulation and detection system, comprising a processor and electrodes;

[0008] The electrodes are used to collect EEG signals at target brain locations and transmit them to the processor;

[0009] The processor is used to obtain a β-band signal based on the EEG signal, and perform Hilbert transform and sliding window integration on the β-band signal in sequence to obtain a brain observation characteristic value, and determine whether the brain state is normal based on the brain observation characteristic value;

[0010] If it is determined that the brain state is abnormal, the processor is also used to control the electrodes to generate corresponding stimulation signals to act on the target brain position.

[0011] Furthermore, the processor is connected to the electrode via a wire, and an analog switch is provided at the processor end for controlling the working mode of the electrode by selecting the output mode of the processor;

[0012] When the analog switch is switched on, the output mode of the processor is the analog-to-digital conversion mode, and the working mode of the electrode is the reading mode; when the analog switch is switched off, the output mode of the processor is the digital-to-analog conversion mode, and the working mode of the electrode is the stimulation mode.

[0013] Furthermore, the processor obtains the β-band signal based on the EEG signal in the following manner:

[0014] The first time-domain local field potential sequence is obtained by periodically sampling the EEG signal;

[0015] Performing fast Fourier transform on the first time-domain local field potential sequence using a time extraction method at preset time intervals to obtain a frequency-domain local field potential sequence;

[0016] The frequency domain local field potential sequence is filtered to obtain a β frequency domain local field potential sequence, namely, the β frequency band signal.

[0017] Furthermore, the steps of obtaining a β-band signal based on the EEG signal, and performing Hilbert transform and sliding window integration on the β-band signal in sequence to obtain a brain observation characteristic value include:

[0018] Performing an inverse fast Fourier transform on the β frequency domain local field potential sequence to obtain a second time domain local field potential sequence;

[0019] Performing Hilbert transform on the second time-domain local field potential sequence to obtain a third time-domain local field potential sequence after phase shifting by 90°;

[0020] constructing an analytical signal based on the second time-domain local field potential sequence and the third time-domain local field potential sequence, and taking a modulus of the analytical signal;

[0021] Using a sliding window of a preset size and a preset step length to perform an internal integration of the modulus value of the analytical signal to obtain a brain observation value sequence; the width of the sliding window is the same as the preset step length;

[0022] The brain observation characteristic value is obtained by performing exponential decay weighted averaging on the brain observation value sequence.

[0023] Furthermore, the method of performing fast Fourier transform on the first time-domain local field potential sequence using a time extraction method to obtain a frequency-domain local field potential sequence includes:

[0024] The frequency domain local field potential sequence is expressed by the first time domain local field potential sequence using the following formula using fast Fourier transform:

[0025]

[0026] Wherein, k represents the frequency domain component, Dt(t) represents the local field potential value with sequence number t in the first time domain local field potential sequence, represents the frequency domain local field potential sequence, N represents the number of local field potential values ​​in the first time domain local field potential sequence;

[0027] The frequency domain local field potential sequence is solved by the following formula using the time extraction method:

[0028]

[0029] Where Dt*(ω)2l represents the local field potential value with sequence number 2l in the frequency domain local field potential sequence, Dt*(ω) 2l+1 Represents the local field potential value with sequence number 2l+1 in the frequency domain local field potential sequence.

[0030] Furthermore, the β frequency domain local field potential sequence is subjected to an inverse fast Fourier transform by the following formula to obtain a second time domain local field potential sequence:

[0031]

[0032] Where k represents the frequency domain component, Dt(t)' represents the second time domain local field potential sequence, represents the frequency domain local field potential sequence, N represents the number of local field potential values ​​in the first time domain local field potential sequence;

[0033] The third time-domain local field potential sequence obtained by performing Hilbert transform on the second time-domain local field potential sequence after phase shift by 90° is:

[0034]

[0035] in, represents the third time-domain local field potential sequence, and Dt(τ)′ represents the local field potential value with sequence number τ in the second time-domain local field potential sequence.

[0036] Furthermore, an analytical signal is constructed based on the second time-domain local field potential sequence and the third time-domain local field potential sequence as follows:

[0037]

[0038] The modulo of the analytical signal is:

[0039]

[0040] Wherein, Ht represents the analytical signal.

[0041] Furthermore, a sliding window of a preset size is used with a preset step size to perform an internal integration of the modulus of the analytical signal using the following formula to obtain a brain observation value sequence, specifically:

[0042]

[0043] in, represents the integral value in the mth sliding window at time T, P is the preset step size, The integral values ​​in the sliding windows constitute the observation value sequence at the time T;

[0044] The observation value sequences corresponding to all moments within the preset time period are sequentially connected to form a brain observation value sequence.

[0045] Furthermore, if the brain observation characteristic value calculated by the processor is greater than or equal to a preset threshold, the processor controls the electrode to generate a corresponding stimulation signal to act on the target brain position;

[0046] If the calculated brain observation characteristic value is continuously smaller than the preset threshold value within the time period [2h, 3h], the processor controls the alarm device to issue a danger alarm.

[0047] In another aspect, the present invention provides a method for internal brain stimulation and detection, comprising:

[0048] Using electrodes to collect EEG signals at the target brain location;

[0049] Based on the EEG signal, a β-band signal is obtained, and the β-band signal is subjected to Hilbert transform and sliding window integration in sequence to obtain a brain observation characteristic value, and based on the brain observation characteristic value, whether the brain state is normal is determined;

[0050] If it is determined that the brain state is abnormal, the electrodes are controlled to generate corresponding stimulation signals to act on the target brain position.

[0051] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects:

[0052] 1. The brain internal stimulation and detection system and method proposed in the present invention first sets an electrode at the target brain position, i.e., the target point, and controls the working mode of the electrode by analog switch, so that the same electrode can be used for signal acquisition and stimulation in time-sharing multiplexing, which can improve the accuracy of signal acquisition and the accuracy of target stimulation, and avoid the defects of the prior art of using two electrodes for signal acquisition and target stimulation respectively, which leads to large acquisition signal noise, large error and inaccurate stimulation position.

[0053] 2. The internal brain stimulation and detection system and method proposed in the present invention converts the collected EEG signals into time domain and frequency domain, so that the brain observation characteristic values ​​used to judge the brain state include the time domain characteristics and frequency domain characteristics of the EEG signals, thereby more accurately judging whether the brain state is normal, avoiding the problem of poor accuracy of brain state judgment based only on the time domain characteristics of the EEG signals in the prior art.

[0054] 3. Compared with the discrete Fourier transform processing method in the prior art, the brain internal stimulation and detection system and method proposed by the present invention provides a finer time resolution and improves the accuracy and real-time performance of signal processing by using fast Fourier transform to process EEG signals. In addition, the Hilbert transform can prevent data overfitting, thereby providing a more accurate phase-shifted local field potential sequence, which is conducive to improving the accuracy of brain state judgment results.

[0055] In the present invention, the above-mentioned technical solutions can also be combined with each other to achieve more preferred combination solutions. Other features and advantages of the present invention will be described in the subsequent description, and some advantages can become obvious from the description, or can be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained through the contents particularly pointed out in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The drawings are only for the purpose of illustrating particular embodiments and are not to be considered limiting of the present invention. Like reference symbols denote like components throughout the drawings.

[0057] Figure 1 A schematic diagram of an internal brain stimulation and detection system according to an embodiment of the present invention;

[0058] Figure 2 This is a flow chart of the method for internal brain stimulation and detection according to an embodiment of the present invention.

[0059] Reference numerals:

[0060] 1-processor; 2-lead; 3-electrode. DETAILED DESCRIPTION

[0061] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0062] A specific embodiment of the present invention discloses a brain internal stimulation and detection system. Figure 1 As shown, the system includes a processor 1, an electrode 3 and a wire 2. Preferably, the processor can be a microchip, and the processor is connected to the electrode through a wire. In practical applications, the processor, the wire and the electrode are arranged inside the brain, and the electrode is arranged at the target brain position. For example, the electrode can be arranged at the position of the subthalamic nucleus or the position of the globus pallidus internal nucleus.

[0063] Preferably, the electrodes are used to collect EEG signals at the target brain location and transmit them to the processor via wires.

[0064] The processor is used to obtain a β-band signal based on the EEG signal, and to perform Hilbert transform and sliding window integration on the β-band signal in sequence to obtain a brain observation characteristic value, and to determine whether the brain state is normal based on the brain observation characteristic value.

[0065] Preferably, the system also includes an external display device and an alarm device, which are connected to the processor by wireless communication, and the external display device can display the brain signals collected by the electrodes and the brain observation characteristic values ​​obtained by the processor processing the EEG signals. In addition, if the brain observation characteristic value is greater than or equal to the preset threshold, it is determined that the brain state is abnormal. At this time, the processor is also used to control the electrode to generate a corresponding stimulation signal to act on the target brain position; if the brain observation characteristic value is continuously less than the preset threshold within the time period [2h, 3h], the processor controls the alarm device to issue a danger alarm to avoid the patient being in a false negative state and failing to discover the danger in time. Since the focus of medical staff or patients on EEG signals should mainly be on whether the signal itself is abnormal, it is easy to ignore the false negative situation after a long time. Although the performance of intracranial EEG signals is normal under this condition, there may actually be misjudgment. Therefore, this warning method can remind medical staff or patients that they may need to conduct a comprehensive analysis and judgment at the current moment in combination with other sensor information at the same time to reduce the false negative situation of the disease.

[0066] Preferably, the processor end is provided with an analog switch for controlling the working mode of the electrode by selecting the output mode of the processor. Specifically, when the brain observation characteristic value is less than the preset threshold value, the analog switch is switched to the on state, at which time the output mode of the processor is the analog-to-digital conversion mode, and the working mode of the electrode is the reading mode, that is, the EEG signal of the target brain position is collected; when the brain observation characteristic value is greater than or equal to the preset threshold value, the analog switch is switched to the off state, at which time the output mode of the processor is the digital-to-analog conversion mode, and the working mode of the electrode is the stimulation mode, that is, the corresponding stimulation signal is generated to act on the target brain position.

[0067] Preferably, the system also includes a memory. When the electrode mode is converted from a stimulation mode to a reading mode, the processor resets the local field potential sequence recently stored in the memory to zero and does not perform secondary stimulation within a certain period of time to avoid ineffective stimulation.

[0068] Preferably, the processor obtains the β-band signal based on the EEG signal in the following manner:

[0069] Step 1, periodically sample the EEG signal to obtain the first time-domain local field potential sequence. Preferably, the sampling period ranges from [4ms, 6ms], and preferably, the sampling period is 5ms. Considering that the detection of EEG abnormalities must be based on the abnormal EEG within a certain time period to determine the EEG abnormalities, therefore, the first time-domain local field potential sequence at the current moment is updated to obtain multiple first time-domain local field potential sequences corresponding to multiple moments within a preset time period, and the brain state is judged based on the multiple first time-domain local field potential sequences. Exemplarily, the first time-domain local field potential sequence corresponding to the time T-1 is {d0, d1, ..., d(N-1)}, and after the update, the first time-domain local field potential sequence corresponding to the time T is {d1, d2, ..., dN}, where N represents the number of local field potential values ​​in the first time-domain local field potential sequence, and N=2 n , n represents a positive integer.

[0070] Step 2. Considering that the time domain features alone cannot effectively characterize the EEG signals, which leads to low accuracy in judging the brain state, this solution also takes the frequency domain features of the EEG signals into account to better characterize the features of the EEG signals and improve the accuracy of the results of judging the brain state. Specifically, the first time domain local field potential sequence is subjected to a fast Fourier transform using a time extraction method at a preset time interval to obtain a frequency domain local field potential sequence. Preferably, the preset time interval is a times the sampling period, preferably, the value range of a is [4,6], and preferably, a is 6. This step specifically includes:

[0071] Step 21, using fast Fourier transform to express the frequency domain local field potential sequence using the first time domain local field potential sequence through the following formula:

[0072]

[0073] Wherein, k represents the frequency domain component, Dt(t) represents the local field potential value with sequence number t in the first time domain local field potential sequence, represents the frequency domain local field potential sequence, and N represents the number of local field potential values ​​in the first time domain local field potential sequence.

[0074] Step 22: Using the time extraction method and the butterfly algorithm to solve the frequency domain local field potential sequence through the following formula:

[0075]

[0076] Where Dt*(ω) 2l represents the local field potential value with sequence number 2l in the frequency domain local field potential sequence, Dt*(ω) 2l+1 Represents the local field potential value with sequence number 2l+1 in the frequency domain local field potential sequence.

[0077] Specifically, N = 2 n The local field potential sequence requires n levels of butterfly operations. In order to ensure the accuracy of the subsequent filtering frequency and prevent the complexity of the distribution from being too high, preferably, the value range of n is [6,9]. Preferably, in the butterfly operation, the operation results of each level are stored in the same group of memories, and the ordinal number of the operation is represented by binary. At this time, the order is exactly the order of the inversion of the frequency domain sequence code, so that the frequency domain local field potential sequence can be decoded.

[0078] Step 23, filtering the frequency domain local field potential sequence to obtain a β frequency domain local field potential sequence, that is, the β frequency band signal, which is specifically expressed as:

[0079]

[0080] Among them, ω β- ,ω β+ Represent the lower and upper limits of the β band respectively.

[0081] Preferably, the steps of obtaining a β-band signal based on an EEG signal, and performing Hilbert transform and sliding window integration on the β-band signal in sequence to obtain a brain observation characteristic value include:

[0082] S1. Performing an inverse fast Fourier transform on the β frequency domain local field potential sequence to obtain a second time domain local field potential sequence, specifically comprising:

[0083] The second time-domain local field potential sequence is obtained by performing an inverse fast Fourier transform on the β frequency-domain local field potential sequence using the following formula:

[0084]

[0085] Where k represents the frequency domain component, Dt(t)' represents the second time domain local field potential sequence, represents the frequency domain local field potential sequence, and N represents the number of local field potential values ​​in the first time domain local field potential sequence.

[0086] S2, performing Hilbert transform on the second time-domain local field potential sequence to obtain a third time-domain local field potential sequence after a 90° phase shift, specifically comprising:

[0087]

[0088] in, represents the third time-domain local field potential sequence, and Dt(τ)′ represents the local field potential value with sequence number τ in the second time-domain local field potential sequence.

[0089] S3, constructing an analytical signal based on the second time-domain local field potential sequence and the third time-domain local field potential sequence, and taking the modulus of the analytical signal, specifically including taking the second time-domain local field potential sequence as the real part of the analytical signal, taking the phase-shifted third time-domain local field potential sequence as the imaginary part of the analytical signal, and the constructed analytical signal is:

[0090]

[0091] The modulo of the analytical signal is:

[0092]

[0093] Wherein, Ht represents the analytical signal.

[0094] S4. Use a sliding window of a preset size and a preset step length to perform an internal integration of the modulus of the analytical signal to obtain a sequence of brain observation values. Preferably, the width of the sliding window is the same as the preset step length. Preferably, the sum of the width of the sliding window and the preset step length is This step specifically includes: in order to reduce the computational complexity, the modulus value of the analytical signal is divided into several windows of length a through a sliding window, and the envelope of each window is modulo summed to obtain the integral value of each window. The specific formula is as follows:

[0095]

[0096] in, represents the integral value in the mth sliding window at time T, P is the preset step size, The integrated values ​​within the sliding windows constitute the observation value sequence at the time T.

[0097] The observation value sequences corresponding to all moments within the preset time period are sequentially connected to form a brain observation value sequence So*(t).

[0098] Preferably, since the Hilbert transform will cause the signals at both ends of the time series to be distorted due to discontinuity, the present invention discards the windows at both ends and uses the windows corresponding to the time of the historical signal as reference values ​​to construct a brain observation value sequence So*(t) with windows as units. The time series S composed of each window at the current time T is T ={S1 T ,S2 T ,...,S N / P T}, and the time series S composed of each window at the previous moment (T-1) T-1 ={S1 T-1 ,S2 T-1 ,...,S N / P T-1}, where S (N-a) / PT (i.e. S2 T ) and S N / P T-1 The time of representation is the same, S1 T and S N / P-1 T-1 The time of representation is the same, but due to the marginal effect of the Hilbert transform at time T-1 and T, S N / P T-1 and S1 T The signal is distorted and needs to be discarded, so the sequence connection corresponding to the continuous time in So*(t) should be expressed as {...,S N / P-2 T-1 ,S N / P-1 T-1 ,S2 T-1 ,...,S N / P T-1}.

[0099] S5. Perform exponential decay weighted averaging on the brain observation value sequence to obtain the brain observation characteristic value. Specifically, the brain observation characteristic value is calculated by the following formula through the optimized neural network model:

[0100]

[0101] Among them, So represents the brain observation characteristic value, L represents the length of the brain observation value sequence So*(t), that is, the number of observations therein, and ξ represents the attenuation coefficient, ξ∈(0,1].

[0102] Another embodiment of the present invention discloses a method for stimulating and detecting the brain. Since the method embodiment and the above system embodiment are based on the same working principle, the repetitions can be referred to the above system embodiment and will not be repeated here.

[0103] Specifically, Figure 2 As shown, the method includes:

[0104] S110. Collecting EEG signals at the target brain location using electrodes.

[0105] S120, obtaining a β-band signal based on the EEG signal, and performing Hilbert transform and sliding window integration on the β-band signal in sequence to obtain a brain observation characteristic value, and judging whether the brain state is normal based on the brain observation characteristic value.

[0106] S130. If it is determined that the brain state is abnormal, the control electrodes generate corresponding stimulation signals to act on the target brain position.

[0107] Compared with the prior art, the internal brain stimulation and detection system and method disclosed in the embodiments of the present invention have the following advantages: first, in the present invention, an electrode is set at the target brain position, i.e., the target point, and the working mode of the electrode is controlled by an analog switch, so that the same electrode is time-division multiplexed for signal acquisition and stimulation, which can improve the accuracy of signal acquisition and the accuracy of the stimulation target point, and avoid the defects of the prior art that two electrodes are used for signal acquisition and target point stimulation respectively, resulting in large acquisition signal noise, large errors, and inaccurate stimulation position. Secondly, the internal brain stimulation and detection system and method proposed in the present invention converts the collected EEG signals into time domain and frequency domain, so that the brain observation characteristic values ​​used to judge the brain state include the time domain characteristics and frequency domain characteristics of the EEG signals, thereby being able to more accurately judge whether the brain state is normal, avoiding the problem of poor accuracy in judging the brain state based only on the time domain characteristics of the EEG signals in the prior art. Finally, the brain internal stimulation and detection system and method proposed by the present invention, compared with the discrete Fourier transform processing method in the prior art, provides a finer time resolution by using fast Fourier transform to process EEG signals, thereby improving the accuracy and real-time performance of signal processing. In addition, the Hilbert transform can prevent data overfitting, thereby providing a more accurate phase-shifted local field potential sequence, which is conducive to improving the accuracy of brain state judgment results.

[0108] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0109] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A brain internal stimulation and detection system, characterized in that: including a processor and an electrode; The electrodes are used to collect EEG signals at target brain locations and transmit them to the processor; The processor is used to obtain a β-band signal based on the EEG signal, and perform Hilbert transform and sliding window integration on the β-band signal in sequence to obtain a brain observation characteristic value, and judge whether the brain state is normal based on the brain observation characteristic value, wherein the processor obtains the β-band signal based on the EEG signal in the following manner: The first time-domain local field potential sequence is obtained by periodically sampling the EEG signal; Performing a fast Fourier transform on the first time-domain local field potential sequence using a time extraction method at a preset time interval to obtain a frequency-domain local field potential sequence specifically includes: The frequency domain local field potential sequence is expressed by the first time domain local field potential sequence using the following formula using fast Fourier transform: Wherein, k represents the frequency domain component, Dt(t) represents the local field potential value with sequence number t in the first time domain local field potential sequence, represents the frequency domain local field potential sequence, N represents the number of local field potential values ​​in the first time domain local field potential sequence; The frequency domain local field potential sequence is solved by the following formula using the time extraction method: Where Dt*(ω) 2l represents the local field potential value with sequence number 2l in the frequency domain local field potential sequence, Dt*(ω) 2l+1 represents the local field potential value with sequence number 2l+1 in the frequency domain local field potential sequence, N=2 n The local field potential sequence requires n levels of butterfly operations, and the value range of n is [6,9]; Performing filtering processing on the frequency domain local field potential sequence to obtain a β frequency domain local field potential sequence, namely, the β frequency band signal; The steps of obtaining a β-band signal based on the EEG signal, and performing Hilbert transform and sliding window integration on the β-band signal in sequence to obtain a brain observation characteristic value include: Performing an inverse fast Fourier transform on the β frequency domain local field potential sequence to obtain a second time domain local field potential sequence; Performing Hilbert transform on the second time-domain local field potential sequence to obtain a third time-domain local field potential sequence after phase shifting by 90°; constructing an analytical signal based on the second time-domain local field potential sequence and the third time-domain local field potential sequence, and taking a modulus of the analytical signal; Using a sliding window of a preset size and a preset step length to perform an internal integration of the modulus value of the analytical signal to obtain a brain observation value sequence; the width of the sliding window is the same as the preset step length; Performing exponential decay weighted averaging on the brain observation value sequence to obtain a brain observation characteristic value; If it is determined that the brain state is abnormal, the processor is also used to control the electrodes to generate corresponding stimulation signals to act on the target brain position.

2. The brain internal stimulation and detection system according to claim 1, characterized in that: The processor is connected to the electrode via a wire, and an analog switch is provided at the processor end for controlling the working mode of the electrode by selecting the output mode of the processor; When the analog switch is switched on, the output mode of the processor is the analog-to-digital conversion mode, and the working mode of the electrode is the reading mode; when the analog switch is switched off, the output mode of the processor is the digital-to-analog conversion mode, and the working mode of the electrode is the stimulation mode.

3. The brain internal stimulation and detection system according to claim 1, characterized in that: The β frequency domain local field potential sequence is subjected to inverse fast Fourier transform to obtain the second time domain local field potential sequence by the following formula: Where k represents the frequency domain component, Dt(t)' represents the second time domain local field potential sequence, represents the frequency domain local field potential sequence, N represents the number of local field potential values ​​in the first time domain local field potential sequence; The third time-domain local field potential sequence obtained by performing Hilbert transform on the second time-domain local field potential sequence after phase shift by 90° is: in, represents the third time-domain local field potential sequence, and Dt(τ)′ represents the local field potential value with sequence number τ in the second time-domain local field potential sequence.

4. The brain internal stimulation and detection system according to claim 3, characterized in that: The analytical signal constructed based on the second time-domain local field potential sequence and the third time-domain local field potential sequence is: The modulo of the analytical signal is: Wherein, Ht represents the analytical signal.

5. The brain internal stimulation and detection system according to claim 4, characterized in that: A sliding window of a preset size is used with a preset step size to integrate the modulus of the analytical signal within the sliding window using the following formula to obtain a brain observation value sequence, specifically: in, represents the integral value in the mth sliding window at time T, P is the preset step size, The integral values ​​in the sliding windows constitute the observation value sequence at the time T; The observation value sequences corresponding to all moments within the preset time period are sequentially connected to form a brain observation value sequence.

6. The brain internal stimulation and detection system according to claim 5, characterized in that: If the brain observation characteristic value calculated by the processor is greater than or equal to a preset threshold, the processor controls the electrode to generate a corresponding stimulation signal to act on the target brain position; If the calculated brain observation characteristic value is continuously smaller than the preset threshold value within the time period [2h, 3h], the processor controls the alarm device to issue a danger alarm.

Images