Brain-computer interface-driven system for reconstructing brain-controlled urination function in patients with spinal cord injury

Abstract

The present application relates to the technical field of electroencephalogram (EEG) and brain-wave signals, and specifically to a brain-computer interface-driven system for brain-controlled urination function reconstruction in patients with spinal cord injury. The system comprises: an EEG signal acquisition module for acquiring the patient's EEG signal; an EEG signal interference module for acquiring the artifact interference characteristic value of each signal cycle; an artifact recognition module for extracting the interference cycle and normal cycle of the EEG signal, and utilizing the difference and correlation between the EEG signals in the interference cycle and the normal cycle to obtain the artifact recognition degree of each component signal sequence in each interference cycle; an EEG signal processing module for reconstructing the screened component signal sequence to obtain the EEG signal after signal processing; a function reconstruction module for identifying the stimulation level of the patient's motor intention through the EEG signal after signal processing, so as to perform urination stimulation training and realize the reconstruction of the urination function after spinal cord injury. The present application can improve the accuracy of EEG signals and effectively reconstruct the urination function after spinal cord injury.

Description

Technical Field

[0001] The present application relates to the technical field of electroencephalogram (EEG) and brain electrical signal technology, and in particular to a brain-computer interface-driven system for reconstructing the brain-controlled urination function of spinal cord injury patients. Background Art

[0002] With the development of brain-computer interface (BCI) technology, the use of BCI to reconstruct urination function lost due to spinal cord injury can effectively restore urination function after spinal cord injury and improve the patient's autonomous interactive ability to urinate. In existing technologies, BCI technology collects, preprocesses, and extracts features from EEG signals to identify the stimulation level corresponding to the patient's motor intention. Based on the stimulation level, functional electrical stimulation equipment is controlled to stimulate the patient's pelvic floor muscles or bladder and other urination-related parts to train them, helping the patient gradually strengthen their pelvic floor muscles and bladder control, ultimately achieving urination function reconstruction after spinal cord injury.

[0003] Since EEG signals are easily affected by the complex interference of eye movement artifacts, existing technologies use wavelet decomposition algorithms to identify and remove eye movement artifact components in EEG signals. However, traditional wavelet decomposition algorithms cannot fully consider the interference characteristics of eye movement artifact components, resulting in insufficient ability to distinguish eye movement artifact components, which can easily cause distortion of the EEG signal after signal reconstruction, thereby affecting the accuracy of the patient's movement intention recognition and making it impossible to effectively reconstruct the urination function after spinal cord injury. Summary of the Invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a brain-computer interface driven system for reconstructing the brain-controlled urination function of spinal cord injury. The technical solutions adopted are as follows:

[0005] This application proposes a brain-computer interface-driven system for reconstructing brain-controlled urination function in spinal cord injury, which includes:

[0006] EEG signal acquisition module: used to obtain the patient's EEG signal;

[0007] EEG signal interference module: extracts the signal period of the EEG signal, obtains the EEG amplitude abnormality of each signal period by analyzing the amplitude deviation fluctuation degree of the EEG signal in the frequency domain, and analyzes the difference in EEG amplitude abnormality of different signal periods to obtain the artifact interference characteristic value of each signal period;

[0008] Artifact recognition module: derives each interference period and normal period of the EEG signal from the artifact interference feature value, performs wavelet decomposition on the EEG signal in the interference period to obtain each component signal sequence, and uses the difference and correlation between each component signal sequence in the interference period and the EEG signal in the normal period to obtain the artifact recognition degree of each component signal sequence in each interference period;

[0009] EEG signal processing module: using the artifact recognition degree to screen the component signal sequence, and reconstructing the screened component signal sequence through a wavelet reconstruction algorithm to obtain the EEG signal after signal processing;

[0010] Functional reconstruction module: Identify the stimulation level of the patient's motor intention through the processed EEG signals to conduct urination stimulation training and achieve reconstruction of urination function after spinal cord injury.

[0011] Preferably, the process of extracting the signal period of the EEG signal includes: arranging the EEG signal in time sequence to obtain the patient's EEG signal sequence, and after frequency domain transformation, taking the inverse of the frequency corresponding to the maximum amplitude in the frequency domain as the signal period of the EEG signal.

[0012] Preferably, the formula for obtaining the abnormality of the EEG amplitude of each signal cycle is:

[0013] ;

[0014] Where, is the abnormality of the EEG amplitude in the ith signal cycle, is the permutation entropy of the first-order difference sequence of the amplitude characteristic sequence of the i-th signal period, is the number of elements in the amplitude characteristic sequence of the i-th signal period, is the sth element in the amplitude characteristic sequence of the i-th signal cycle, is the instantaneous amplitude of the i-th signal cycle.

[0015] Preferably, the EEG signals of each signal cycle are arranged in time sequence to obtain an EEG signal subsequence of each signal cycle, and the EEG signal subsequence of each signal cycle is subjected to Hilbert-Huang transform to extract the instantaneous amplitude of each signal cycle; and the EEG signal subsequence of each signal cycle is subjected to frequency domain transform to arrange the amplitudes corresponding to all frequencies in the frequency domain of each signal cycle from small to large according to frequency to obtain an amplitude characteristic sequence of each signal cycle.

[0016] Preferably, the formula for obtaining the artifact interference characteristic value of each signal period is:

[0017] ;

[0018] Where, is the artifact interference characteristic value of the i-th signal period, is the sigmoid normalization function, is the abnormality of the EEG amplitude in the ith signal cycle, is the EEG amplitude abnormality of the kth adjacent signal cycle of the ith signal cycle, K is the number of adjacent signal cycles of the ith signal cycle, among which the K signal cycles with the closest time interval between each signal cycle are regarded as the K adjacent signal cycles of each signal cycle.

[0019] Preferably, obtaining each interference period and normal period of the EEG signal specifically includes:

[0020] Perform threshold segmentation on the artifact interference characteristic values ​​of all signal periods within the acquisition time period, obtain the segmentation threshold and record it as the first segmentation threshold, and take the signal period corresponding to the artifact interference characteristic value higher than the first segmentation threshold as the interference period of the EEG signal. Otherwise, take the signal period as the normal period of the EEG signal.

[0021] Preferably, the formula for obtaining the artifact recognition degree of each component signal sequence in each interference period is:

[0022] ;

[0023] Where, is the artifact recognition degree of the j-th component signal sequence in the a-th interference period, is the signal deviation significance of the j-th component signal sequence in the a-th interference period, is an exponential function with a natural constant as base, is the mean of the mutual information between the j-th component signal sequence of the a-th interference period and the EEG signal subsequences of its reference periods, and the normal periods with the closest P time intervals between each interference period are used as the P reference periods of each interference period.

[0024] Preferably, obtaining the signal deviation significance of each component signal sequence in each interference period further includes:

[0025] The ratio of the mean value to the variance of the difference distance between each component signal sequence of each interference period and the EEG signal subsequence of each reference period is calculated as the signal deviation significance of each component signal sequence of each interference period.

[0026] Preferably, the component signal sequence is screened using the artifact recognition degree, and the screened component signal sequence is reconstructed by a wavelet reconstruction algorithm, including: threshold segmentation of the artifact recognition degrees of all component signal sequences of each interference period, the obtained segmentation threshold is recorded as the second segmentation threshold, all component signal sequences corresponding to artifact recognition degrees less than the second segmentation threshold are screened out, and the wavelet reconstruction algorithm is used to obtain the reconstructed EEG signal of each interference period signal.

[0027] Preferably, the acquisition process of the signal-processed EEG signal is as follows: the EEG signal of the normal period within the acquisition time period and the EEG signal reconstructed from all the interference periodic signals are merged in time sequence, and smoothed to obtain the signal-processed EEG signal.

[0028] This application has the following beneficial effects:

[0029] This application analyzes the abnormal characteristics of EEG signal amplitudes in different signal periods, and analyzes the transient strong interference characteristics of eye movement artifacts on EEG signals based on the abnormal characteristics of EEG signal amplitudes in the signal period, thereby improving the ability to distinguish eye movement artifact components, which is conducive to the subsequent accurate identification of signal periods where artifact interference occurs, thereby more accurately eliminating artifact signals;

[0030] At the same time, by analyzing the artifact interference characteristic value, the interference period and the normal period are identified. The interference period is identified and the artifact component is removed, which can improve the accuracy of artifact removal in the EEG signal and avoid the problem of EEG signal distortion caused by incorrect processing of the EEG signal of the normal period.

[0031] This application takes into account the differences and interdependence between EEG signals and artifact signals, identifies the artifact signal features in all component signals after the EEG signal is decomposed, and removes the component signals with artifact signal features during signal reconstruction, thereby improving the accuracy of EEG signals and avoiding adverse effects on the patient's motor intention recognition, thereby more effectively reconstructing the urination function after spinal cord injury. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0033] Figure 1 A block diagram of a brain-computer interface-driven system for reconstructing brain-controlled urination function in spinal cord injury provided in one embodiment of the present application;

[0034] Figure 2 A schematic diagram of the EEG signal processing and acquisition process provided in one embodiment of the present application. DETAILED DESCRIPTION

[0035] In order to further illustrate the technical means and effects adopted by this application to achieve the predetermined invention purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features and effects of the brain-computer interface-driven spinal cord injury brain-controlled urination function reconstruction system proposed in this application. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics of one or more embodiments may be combined in any suitable form.

[0036] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0037] The specific scheme of the brain-computer interface driven spinal cord injury brain-controlled urination function reconstruction system provided by this application is described in detail below with reference to the accompanying drawings.

[0038] See also Figure 1 , which shows a block diagram of a brain-computer interface driven spinal cord injury brain-controlled urination function reconstruction system provided by an embodiment of the present application, the system includes:

[0039] EEG signal acquisition module: used to obtain the patient's EEG signals.

[0040] In order to improve the urination function of patients with spinal cord injury, it is necessary to eliminate the interference of eye movement artifacts on EEG signals, so as to more accurately identify the patient's movement intention and its corresponding stimulation level, and ultimately more effectively realize the reconstruction of urination function after spinal cord injury driven by brain-computer interface.

[0041] The patient's EEG signals are collected in real time through the EEG electrode cap worn by the patient. In this embodiment, the EEG electrode cap includes electrodes in the frontal lobe and central areas related to movement intention and movement decision-making. The collected EEG signals are transmitted to the signal acquisition terminal through brain-computer interface technology. In this embodiment, the signal acquisition terminal is a terminal notebook.

[0042] Furthermore, in this embodiment, the EEG signals received by the signal acquisition terminal of the brain-computer interface within the acquisition time period are arranged in chronological order to obtain the patient's EEG signal sequence. The sampling frequency of the EEG signal is 1024 Hz, and the acquisition time period is 10 seconds. The implementer can adaptively set the EEG signal sampling frequency and acquisition time period according to actual needs.

[0043] EEG signal interference module: extracts the signal period of the EEG signal, obtains the EEG amplitude abnormality of each signal period by analyzing the amplitude deviation fluctuation degree of the EEG signal in the frequency domain, and analyzes the difference in EEG amplitude abnormality of different signal periods to obtain the artifact interference characteristic value of each signal period.

[0044] Because EEG signals often contain numerous eye movement artifacts, they are subject to complex interference, and the presence of these artifacts can severely interfere with their interpretability. Therefore, in order to more accurately identify the patient's movement intentions and their corresponding stimulation levels, it is necessary to eliminate the interference caused by these artifacts on the EEG signals.

[0045] Therefore, in order to accurately eliminate the interference of artifact signals on EEG signals, the EEG signal sequence is used as the input of Fourier transform. The Fourier transform can be a fast Fourier transform or a discrete Fourier transform. This embodiment uses fast Fourier transform to extract the frequency domain data of the EEG signal sequence. In this embodiment, for the convenience of expression and understanding, the frequency corresponding to the maximum amplitude in the frequency domain data is used as the periodic frequency, and the inverse of the periodic frequency is used as the signal period of the EEG signal. The specific process of fast Fourier transform is a well-known technology and will not be elaborated on.

[0046] Furthermore, the EEG signals for each signal cycle within the acquisition period are arranged in chronological order to obtain an EEG signal subsequence for each signal cycle. Generally, EEG signals exhibit continuous variations, while eye movement artifacts are typically synchronized with eye movements. These artifacts are often short-lived and do not affect EEG signals in all signal cycles.

[0047] Therefore, in order to more accurately identify and remove artifact components in EEG signals, it is necessary to analyze the artifact interference characteristics of different signal cycles. The EEG signal subsequence of each signal cycle is used as the input of the Hilbert-Huang transform, and the Hilbert-Huang transform is used to extract the instantaneous amplitude of each signal cycle. At the same time, the EEG signal subsequence of each signal cycle is used as the input of the Fast Fourier Transform, and the Fast Fourier Transform is used to obtain the amplitude corresponding to all frequencies of each signal cycle. The Hilbert-Huang transform and the Fast Fourier Transform are both well-known technologies and will not be described in detail.

[0048] Generally speaking, if the amplitude changes corresponding to all frequencies in a certain signal period are more chaotic, and the difference level between the amplitudes corresponding to all frequencies and the instantaneous amplitude is higher, it can be said that the amplitude of the EEG signal in this signal period has higher abnormal characteristics, and the signal period is more likely to be affected by strong interference of eye movement artifacts.

[0049] Therefore, the amplitudes corresponding to all frequencies of each signal cycle are arranged in order from small to large frequency to obtain the amplitude characteristic sequence of each signal cycle. Based on the degree of difference between the amplitude within the amplitude characteristic sequence of each signal cycle and the instantaneous amplitude, the EEG amplitude abnormality of each signal cycle is calculated:

[0050] ;

[0051] Where, is the abnormality of the EEG amplitude in the ith signal cycle, is the permutation entropy of the first-order difference sequence of the amplitude characteristic sequence of the i-th signal period, is the number of elements in the amplitude characteristic sequence of the i-th signal period, is the sth element in the amplitude characteristic sequence of the i-th signal cycle, is the instantaneous amplitude of the i-th signal cycle.

[0052] Among them, it can be understood that the abnormality of EEG amplitude reflects the abnormal characteristics of the EEG signal amplitude within the signal period. The higher the abnormality of EEG amplitude, the more likely it is to be affected by strong interference from eye movement artifacts, causing the amplitude of the EEG signal in the frequency domain to show a higher degree of abnormal changes.

[0053] Furthermore, in order to more accurately identify the interference characteristics of the artifact component on the EEG signal, the K signal cycles with the closest time interval between each signal cycle are recorded as the K adjacent signal cycles of each signal cycle. In this embodiment, the value of K is 10, and the implementer can perform adaptive value selection.

[0054] Since eye movement artifacts have a high degree of interference with EEG signals, and the duration of the interference is often short, the greater the difference between the EEG amplitude abnormality in a certain signal cycle and its adjacent signal cycles, and the greater the EEG amplitude abnormality in the signal cycle, the more it can reflect the short-term strong interference characteristics of eye movement artifacts on EEG signals.

[0055] Therefore, the artifact interference characteristic value of each signal period is calculated:

[0056] ;

[0057] Where, is the artifact interference characteristic value of the i-th signal period, is the sigmoid normalization function, is the EEG amplitude abnormality of the kth adjacent signal cycle of the ith signal cycle, and K is the number of adjacent signal cycles of the ith signal cycle.

[0058] Among them, it can be understood that the artifact interference characteristic value reflects the short-term strong interference characteristics of eye movement artifacts on the EEG signal. The larger the artifact interference characteristic value, the more significant the interference effect of eye movement artifacts on the EEG signal in this signal period is, and the more important it is to identify and remove the artifact component in this signal period, so that the patient's movement intention and its corresponding stimulation level can be identified more accurately in the future.

[0059] Artifact recognition module: The interference periods and normal periods of the EEG signal are obtained through the artifact interference feature value, and the EEG signal in the interference period is subjected to wavelet decomposition to obtain the component signal sequences. The artifact recognition degree of each component signal sequence in each interference period is obtained by using the difference and correlation between the component signal sequences in the interference period and the EEG signal in the normal period.

[0060] Furthermore, in order to more accurately identify and remove the artifact components within the signal period and avoid the problem of distortion of the EEG signal after processing, the artifact interference characteristic values ​​of all signal periods within the acquisition time period are used as the input of the maximum inter-class variance algorithm, and the maximum inter-class variance algorithm is used to obtain the segmentation threshold, which is recorded as the first segmentation threshold. The maximum inter-class variance algorithm is a well-known technology and will not be elaborated on in detail.

[0061] Furthermore, each signal period corresponding to an artifact interference characteristic value higher than the first segmentation threshold is regarded as each interference period of the EEG signal, and each signal period corresponding to an artifact interference characteristic value less than or equal to the first segmentation threshold is regarded as each normal period of the EEG signal.

[0062] The maximum inter-class variance algorithm is used to identify the interference period and normal period within the acquisition time period through the artifact interference eigenvalue. Only the interference period within the acquisition time period is identified and the artifact component is removed, which can improve the accuracy of artifact removal in the EEG signal and avoid the problem of EEG signal distortion caused by incorrect processing of the EEG signal of the normal period.

[0063] In order to identify and remove the artifact components within the signal period, the EEG signal subsequence of each interference period in the acquisition time period is used as the input of the wavelet decomposition algorithm, wherein Haar wavelet is selected as the wavelet basis function, the decomposition layer number is set to 6, and the wavelet decomposition algorithm is used to decompose the EEG signal subsequence into each component signal sequence. The wavelet decomposition algorithm is a well-known technology and will not be described in detail.

[0064] Generally, artifact component signal sequences often differ significantly from normal EEG signal subsequences, and the two have a weak interdependence. Therefore, in this embodiment, the P normal periods closest to each interference period are used as the P reference periods for each interference period.

[0065] Furthermore, the difference distances between each component signal sequence of each interference cycle and its P reference cycle EEG signal subsequences are calculated, wherein the higher the average level of all difference distances and the smaller the discrete level of all difference distances, the more the difference characteristics between the artifact signal and multiple normal EEG signals can be highlighted. Therefore, the ratio of the average value to the variance of the difference distances between each component signal sequence of each interference cycle and its reference cycle EEG signal subsequences is recorded as the signal deviation significance of each component signal sequence of each interference cycle, reflecting the deviation significance between the artifact signal and the normal EEG signal. The greater the signal deviation significance, the more the component signal can reflect the artifact signal characteristics, and the more it should be removed during signal reconstruction to avoid interfering with the EEG signal.

[0066] At the same time, since there is a strong independence relationship between the EEG signal and the artifact signal, the mutual dependence between the two is poor. Therefore, by combining the mutual information characteristics between the two, the artifact recognition degree of each component signal sequence in each interference period is calculated:

[0067] ;

[0068] Where, is the artifact recognition degree of the j-th component signal sequence in the a-th interference period, is the signal deviation significance of the j-th component signal sequence in the a-th interference period, is an exponential function with a natural constant as base, is the mean of the mutual information between the j-th component signal sequence of the a-th interference period and the EEG signal subsequences of each reference period. The calculation of the mutual information is a well-known technology and will not be described in detail.

[0069] Artifact recognition reflects the artifact signal characteristics in all component signals after EEG signal decomposition. The greater the artifact recognition, the more likely the signal characteristics contained in the component signal sequence are artifact signal characteristics. The component signal should be removed during signal reconstruction, thereby improving the accuracy of EEG signals within the signal period.

[0070] The EEG signal processing module uses the artifact recognition degree to screen the component signal sequence, and reconstructs the screened component signal sequence through a wavelet reconstruction algorithm to obtain the EEG signal after signal processing.

[0071] Furthermore, the component signal sequences are screened using the artifact recognition degree of each component signal sequence of each interference period, and the EEG signals of the normal period and the reconstructed component signal sequences after filtering all interference period signals are merged in chronological order and smoothed to obtain the processed EEG signals.

[0072] Preferably, in this embodiment, the artifact recognition degree of all component signal sequences of each interference period is used as the input of the maximum inter-class variance algorithm, and the maximum inter-class variance algorithm is used to obtain the corresponding segmentation threshold, which is recorded as the second segmentation threshold, and all component signal sequences corresponding to the artifact recognition degree less than the second segmentation threshold are used as the input of the wavelet reconstruction algorithm, and the wavelet reconstruction algorithm is used to obtain the EEG signal after reconstructing each interference period signal, and obtain the EEG signal subsequence after reconstructing each interference period signal, wherein the maximum inter-class variance algorithm and the wavelet reconstruction algorithm are both well-known technologies and will not be elaborated on in detail.

[0073] In order to more accurately identify the patient's movement intention and its corresponding stimulation level, preferably, in this embodiment, all normal-cycle EEG signal subsequences and all reconstructed EEG signal subsequences of interference periodic signals within the acquisition time period are merged in chronological order, and the merged sequence is used as the input of the moving average method (MA). The merged sequence is smoothed using the moving average method to obtain the EEG signal after signal processing, which is recorded as the EEG signal sequence after signal processing, to eliminate the adverse interference caused by eye movement artifacts on the EEG signal and improve the accuracy of the EEG signal after signal reconstruction. The moving average method is a well-known technology and will not be described in detail. In actual application scenarios, the implementer may also use other data smoothing processing methods to process and analyze all normal-cycle EEG signal subsequences and all reconstructed EEG signal subsequences of interference periodic signals to obtain the final processed EEG signal to obtain the EEG signal sequence after signal processing.

[0074] Specifically, the EEG signal processing and acquisition flow diagram provided in this embodiment is as follows: Figure 2 shown.

[0075] Functional reconstruction module: Identify the stimulation level of the patient's motor intention through the processed EEG signals to conduct urination stimulation training and achieve reconstruction of urination function after spinal cord injury.

[0076] According to the above modules, interference analysis and processing of EEG signals can be achieved to obtain EEG signals after interference processing. The patient's movement intention is scored based on the processed EEG signals, and stimulation training of urination-related parts is performed based on this to complete the reconstruction of urination function after spinal cord injury. Preferably, in this embodiment, the time domain features and frequency domain features of the EEG signal sequence after signal processing are extracted by using brain-computer interface technology, wherein the time domain features include movement-related cortical potentials (MRCP) and movement-related myoelectricity (SMR), and the frequency domain features include EEG activities in the alpha and delta frequency bands related to movement;

[0077] Then, the entropy weight method was used to perform weighted fusion on the time-frequency features of EEG activity, and the sigmoid function was used to normalize the weighted fusion results to determine the patient's motor intention score. Based on the score interval of the patient's motor intention score, the stimulation level of the patient's motor intention was determined. The stimulation level was divided into low level, ordinary level and high level. The scoring interval of the low level was (0, 0.3], the scoring interval of the ordinary level was (0.3, 0.6], and the scoring interval of the high level was (0.6, 1).

[0078] Ultimately, the stimulation level based on the patient's motor intention assists the physician in selecting a stimulation plan suitable for the patient's condition, helping to stimulate the patient's pelvic floor muscles, bladder and other urination-related parts, so that the patient can gradually increase the strength of the pelvic floor muscles and control over the bladder, which is used to achieve the reconstruction of urination function after spinal cord injury.

[0079] Among them, the extraction of time-frequency features of EEG activity, the determination of movement intention scores, and the stimulation training of urination-related parts are all well-known technologies in the field of brain-computer interfaces and will not be elaborated on in detail.

[0080] It should be noted that the order of the embodiments of the present application is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0081] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0082] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A brain-computer interface-driven system for reconstructing the brain-controlled urination function in patients with spinal cord injury, characterized by: The system comprises: EEG signal acquisition module: used to obtain the patient's EEG signal; EEG signal interference module: extracts the signal period of the EEG signal, obtains the EEG amplitude abnormality of each signal period by analyzing the amplitude deviation fluctuation degree of the EEG signal in the frequency domain, and analyzes the difference in EEG amplitude abnormality of different signal periods to obtain the artifact interference characteristic value of each signal period; Artifact recognition module: derives each interference period and normal period of the EEG signal from the artifact interference feature value, performs wavelet decomposition on the EEG signal in the interference period to obtain each component signal sequence, and uses the difference and correlation between each component signal sequence in the interference period and the EEG signal in the normal period to obtain the artifact recognition degree of each component signal sequence in each interference period; EEG signal processing module: using the artifact recognition degree to screen the component signal sequence, and reconstructing the screened component signal sequence through a wavelet reconstruction algorithm to obtain the EEG signal after signal processing; Functional reconstruction module: This module uses processed EEG signals to identify the stimulation level of the patient's motor intention, allowing for urination stimulation training to restore urination function after spinal cord injury. The formula for obtaining the abnormality of the EEG amplitude of each signal cycle is: ； Where, is the abnormality of the EEG amplitude in the ith signal cycle, is the permutation entropy of the first-order difference sequence of the amplitude characteristic sequence of the i-th signal period, is the number of elements in the amplitude characteristic sequence of the i-th signal period, is the sth element in the amplitude characteristic sequence of the i-th signal cycle, is the instantaneous amplitude of the i-th signal cycle; The EEG signals of each signal cycle are arranged in time sequence to obtain an EEG signal subsequence of each signal cycle, and the Hilbert-Huang transform is performed on the EEG signal subsequence of each signal cycle to extract the instantaneous amplitude of each signal cycle; and the frequency domain transform is performed on the EEG signal subsequence of each signal cycle, and the amplitudes corresponding to all frequencies in the frequency domain of each signal cycle are arranged from small to large according to frequency to obtain the amplitude characteristic sequence of each signal cycle; The formula for obtaining the artifact interference characteristic value of each signal period is: ； Where, is the artifact interference characteristic value of the i-th signal period, is the sigmoid normalization function, is the abnormality of the EEG amplitude in the ith signal cycle, is the EEG amplitude abnormality of the kth adjacent signal cycle of the i-th signal cycle, K is the number of adjacent signal cycles of the i-th signal cycle, among which the K signal cycles with the closest time interval between each signal cycle are regarded as the K adjacent signal cycles of each signal cycle; The formula for obtaining the artifact recognition degree of each component signal sequence in each interference period is: ； Where, is the artifact recognition degree of the jth component signal sequence in the ath interference period, is the signal deviation significance of the j-th component signal sequence in the a-th interference period, is an exponential function with a natural constant as base, is the mean of the mutual information between the j-th component signal sequence of the a-th interference period and the EEG signal subsequences of its reference periods, and the normal periods with the closest P time intervals between each interference period are used as the P reference periods of each interference period.

2. The brain-computer interface driven spinal cord injury brain-controlled urination function reconstruction system according to claim 1, characterized in that: The process of extracting the signal period of the EEG signal includes: arranging the EEG signal in time sequence to obtain the patient's EEG signal sequence, and taking the reciprocal of the frequency corresponding to the maximum amplitude in the frequency domain as the signal period of the EEG signal after frequency domain transformation.

3. The brain-computer interface driven spinal cord injury brain-controlled urination function reconstruction system according to claim 1, characterized in that: The acquisition of each interference period and normal period of the EEG signal specifically includes: Perform threshold segmentation on the artifact interference feature values ​​of all signal periods within the acquisition time period, obtain the segmentation threshold and record it as the first segmentation threshold, and take the signal period corresponding to the artifact interference feature value higher than the first segmentation threshold as the interference period of the EEG signal. Otherwise, take the signal period as the normal period of the EEG signal.

4. The brain-computer interface driven spinal cord injury brain-controlled urination function reconstruction system according to claim 1, characterized in that: The acquisition of the signal deviation significance of each component signal sequence in each interference period further includes: The ratio of the mean to the variance of the difference distance between each component signal sequence of each interference period and the EEG signal subsequence of each reference period is calculated as the signal deviation significance of each component signal sequence of each interference period.

5. The brain-computer interface driven spinal cord injury brain-controlled urination function reconstruction system according to claim 1, characterized in that: The component signal sequences are screened using the artifact recognition degree, and the screened component signal sequences are reconstructed using a wavelet reconstruction algorithm, including: performing threshold segmentation on the artifact recognition degrees of all component signal sequences of each interference period, recording the obtained segmentation threshold as a second segmentation threshold, screening out all component signal sequences corresponding to artifact recognition degrees less than the second segmentation threshold, and using a wavelet reconstruction algorithm to obtain the reconstructed EEG signal of each interference period signal.

6. The brain-computer interface driven spinal cord injury brain-controlled urination function reconstruction system according to claim 5, characterized in that: The acquisition process of the signal-processed EEG signal is as follows: the EEG signal of the normal period within the acquisition time period and the EEG signal reconstructed from all the interference periodic signals are merged in time sequence, and smoothed to obtain the signal-processed EEG signal.

Images