Method for aligning neural signal in unhealthy state to neural signal in healthy state
Through the health evaluation and analysis of the target object, the response delay and waveform of neural signals are corrected and noise is eliminated, and the problem of inaccurate neural signal acquisition in non-healthy states is solved, signal alignment and noise cleaning are realized, and acquisition quality is improved.
Patent Information
- Application Number
- CN202411003948.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-25
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-25
AI Technical Summary
The prior art does not consider the target health status when collecting neural signals, resulting in inaccurate neural signal acquisition in non-healthy states, problems such as delay in response time, waveform changes and noise signals.
Through the evaluation of the behavioral health of the target object, stimulus response recognition, fever analysis and cross-functional cell analysis, the response delay coefficient, amplitude and width correction coefficient are calculated, the neural signal is corrected, and noisy cells are eliminated to achieve signal alignment.
Adjust the quality of neural signal in non-healthy states to a healthy state, clean the noise signal, and improve the accuracy and quality of signal acquisition.
Smart Images

Figure CN118924246A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of neural signal alignment, and in particular to a method for aligning an unhealthy neural signal to a healthy neural signal. Background Art
[0002] At present, invasive and non-invasive technologies are usually used to collect neural signals. However, the target's health status is not taken into account when collecting neural signals. When the body is in an unhealthy state, the central nervous system will respond more slowly, which can directly lead to delayed actions and movements; fever will cause brain blood vessels to dilate and increase cerebral blood flow. A large number of studies have shown that the brain can directly regulate neural activity through heartbeats; when the heart supplies blood, the brain's blood vessels dilate, causing the distance and resistance from the electrode to the neuron to change, thereby reducing the measured signal amplitude and lengthening the width. Therefore, the health status of the target will adjust its neural signal response time, waveform amplitude / width, etc., and the collected signal may contain noise signals. Therefore, there is an urgent need to solve this problem by aligning unhealthy neural signals to healthy neural signals. Summary of the invention
[0003] In view of the above problems, the present invention is proposed to provide a method for aligning a neural signal in an unhealthy state to a neural signal in a healthy state, which overcomes the above problems or at least partially solves the above problems.
[0004] In order to solve the above technical problems, the embodiments of the present application disclose the following technical solutions:
[0005] In a first aspect, an embodiment of the present invention discloses a method for aligning an unhealthy neural signal to a healthy neural signal, characterized by comprising:
[0006] S100. Evaluate the behavioral health of the target object to obtain a behavioral health coefficient of the target object;
[0007] S200. Identify the stimulus response of the target object, and obtain the response delay coefficient of the neural signal in combination with the behavioral health coefficient; obtain the neural signal after the response time correction based on the response delay coefficient;
[0008] S300. Performing fever analysis on the target object to obtain characteristic information before and after fever of the target object;
[0009] S400. Obtaining amplitude and width correction coefficients of the neural signal according to the characteristic information before and after the target object is heated, and performing width and amplitude correction on the neural signal corrected at the response time according to the amplitude and width correction coefficients to obtain a neural signal after amplitude and width correction;
[0010] S500. Analyze the target object across functional areas, remove the noise cells, and obtain the final quality-enhanced neural signal.
[0011] Furthermore, the target object's behavioral health assessment at least includes an analysis of the target object's eating, defecation, movement and sleep; wherein the specific method of the target object's eating analysis includes:
[0012] S101. Obtaining a video of the target object eating within a preset period before and after electrode implantation, respectively, and analyzing the video to obtain a video clip of the target object eating;
[0013] S102. Identifying the target object's eating segment, and when there are continuous preset frame images in the eating segment that are identified as abnormal eating, determining the continuous preset frame images as abnormal eating images;
[0014] S103. According to the abnormal eating image, obtain the model confidence corresponding to the image;
[0015] S104. According to the model confidence corresponding to the image, the eating status results of the target object before electrode implantation and the eating status results after electrode implantation are obtained respectively.
[0016] Furthermore, the defecation analysis, action analysis and sleep analysis of the target object are performed to obtain the behavioral health coefficient of the target object. The specific method includes:
[0017] S105. Obtain the defecation, movement and sleep videos of the target object within a preset time before and after electrode implantation, repeat S101-S104, and obtain the defecation status results, electrical defecation status results and sleep status results before and after electrode implantation respectively;
[0018] S106. Obtain the health coefficient of the target object based on the target object's eating results, defecation results, electrical defecation results and sleeping results before and after electrode implantation.
[0019] Further, in S200, multiple stimulations are performed on the target object to obtain the stimulation reaction time of the target object before and after the electrode is implanted, and the response delay coefficient of the neural signal is obtained by combining the behavioral health coefficient. The neural signal collected by the electrode is corrected based on the response delay coefficient to obtain the neural signal after the response time correction; the specific method includes:
[0020] S201. Stimulate the target object before and after the electrode implantation, wherein the stimulation includes at least motion stimulation, visual stimulation, auditory stimulation and tactile stimulation, obtain a stimulation video of the target object, load a stimulation recognition model, and when the continuous preset frame image is recognized as having stimulation, determine the continuous preset frame image as a stimulation image, and determine a stimulation implementation frame according to the stimulation image;
[0021] S202. Training the target object stimulus response recognition model, when the continuous preset frame image is recognized as a normal response, determining the continuous preset frame image as a normal response image, and determining the stimulus response normal frame according to the normal response image;
[0022] S203. Calculating the stimulation response time of the target object before and after electrode implantation according to the stimulation implementation frame and the stimulation response normal frame;
[0023] S204. Calculate the response delay coefficient according to the target object stimulation reaction time before and after the electrode implantation and the behavioral health coefficient;
[0024] S205. Obtain a corrected response time according to the neural signal acquisition time and the response delay coefficient, and obtain a neural signal after the response time is corrected according to the corrected response time.
[0025] Furthermore, in S300, the characteristic information of the target object before fever includes the average head temperature before fever, the blood vessel diameter before fever, the equivalent center of mass position of the electrode before fever, and the spatial position of each neuron before fever; the characteristic information of the target object after fever includes the average head temperature after fever, the blood vessel diameter after fever, the equivalent center of mass position of the electrode after fever, and the spatial position of each neuron after fever.
[0026] Furthermore, the average head temperature of the target object before and after fever, and the blood vessel diameter before and after fever are obtained. The specific method includes: based on an infrared imaging device, obtaining the average head temperature of the target object before and after fever; fusing the acquired CT tomography image and MRI tomography blood vessel diameter scanning image to obtain a fused image, segmenting the blood vessels in the fused image, and marking the blood vessel area to obtain a marked image; performing three-dimensional reconstruction on the marked image to obtain a three-dimensional reconstructed image, and obtaining the blood vessel diameters before and after fever according to the three-dimensional reconstructed image.
[0027] Furthermore, the equivalent centroid position of the electrode before and after fever, and the spatial position of each neuron before and after fever are obtained. The specific method includes: obtaining the equivalent centroid position of the electrode before and after fever according to the three-dimensional reconstructed image; extracting the spike signal by the threshold method for each acquired neural signal, calculating the three-dimensional coordinate value of each spike segment based on a preset formula and using an optimization algorithm, clustering the three-dimensional coordinate values of the spike segments, and obtaining the spatial position of each neuron before and after fever.
[0028] Further, in S400, after obtaining the amplitude and width correction coefficients, the corrected width and amplitude are obtained according to the positional relationship between the cells and electrodes and the blood vessels. The specific method includes: when the cells and electrodes are on both sides of the blood vessel, the time span between the peak and the trough of the spike signal collected after fever is defined as the waveform width, and the time width after fever is obtained, then the corrected time width is the product of the time width after fever and the amplitude and width correction coefficient; the maximum amplitude after fever is obtained, and the corrected maximum amplitude is the quotient of the maximum amplitude after fever and the amplitude and width correction coefficient; when the cells and electrodes are on the same side of the blood vessel, the corrected time width is the quotient of the time width after fever and the amplitude and width correction coefficient, and the corrected maximum amplitude is the product of the maximum amplitude after fever and the amplitude and width correction coefficient; according to the corrected time width and the corrected maximum amplitude, the neural signal with amplitude and width correction is obtained.
[0029] Further, in S500, the target object is analyzed across functional areas of cells, specifically including stimulation pattern recognition, functional area segmentation and neuron distribution, and the specific method includes: when a visual stimulation experiment with or without light is conducted and a high-decibel noise is generated, the stimulation pattern is firstly recognized and an abnormal sound recognition model is trained. When the abnormal sound reaches the threshold of the target object's auditory functional area neurons to generate a stress response, the abnormal sound recognition model automatically runs and gives a prompt; then the functional area is segmented, the functional area segmentation model is trained, each functional area in the fused image is segmented, and a segmented image with functional area boundary lines is obtained, and then the segmented image is three-dimensionally reconstructed; finally, the neuron spatial coordinates are matched and divided with each functional area in the three-dimensional reconstructed image, the neurons of the auditory functional area are obtained, and a neural signal set is formed according to the neurons; based on the RNN neural network model, a visual or auditory response recognition model is trained, and the neural signals that form the neural signal set are classified to obtain an auditory response pattern neural signal set; the auditory response pattern neural signal set is eliminated from the amplitude and width corrected neural signals to obtain a final quality-enhanced neural signal set.
[0030] In a second aspect, an embodiment of the present invention discloses an electronic device, including:
[0031] one or more processors;
[0032] A memory for storing one or more programs;
[0033] When the one or more programs are executed by the one or more processors, the one or more processors implement the signal alignment method.
[0034] The beneficial effects of the above technical solution provided by the embodiment of the present invention include at least:
[0035] The present invention discloses a method for aligning an unhealthy neural signal with a healthy neural signal, comprising: evaluating the behavioral health of a target object to obtain a behavioral health coefficient of the target object; identifying the stimulus response of the target object, and obtaining a response delay coefficient of the neural signal in combination with the behavioral health coefficient; obtaining a neural signal corrected at the response time based on the response delay coefficient; performing a fever analysis on the target object to obtain characteristic information before and after fever of the target object; obtaining an amplitude and width correction coefficient of the neural signal based on the characteristic information before and after fever of the target object, and performing width and amplitude correction on the neural signal corrected at the response time based on the amplitude and width correction coefficient to obtain an amplitude and width corrected neural signal; performing a cross-functional cell analysis on the target object, eliminating noise cells, and obtaining a neural signal with final quality enhancement.
[0036] The present invention adjusts the response time of the neural signal, corrects the waveform amplitude and width, and removes noise based on factors such as the target's behavioral health status, stimulus response, body temperature, and cross-functional noise cells, thereby aligning the quality of neural signals in unhealthy conditions to neural signals in healthy conditions and cleaning noise signals. This solves the problem that the current collection of neural signals does not take into account the target's health status, and the collection of neural signals is inaccurate when the organism is in an unhealthy state.
[0037] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:
[0039] Figure 1 This is a flow chart of a method for aligning an unhealthy neural signal with a healthy neural signal in Embodiment 1 of the present invention;
[0040] Figure 2 This is a schematic diagram of neural signal response delay correction in Example 1 of the present invention;
[0041] Figure 3 This is a schematic diagram of correction of the width and amplitude height of a neural signal waveform in Example 1 of the present invention;
[0042] Figure 4 This is a schematic diagram of the structure of an electronic device in Embodiment 2 of the present invention. DETAILED DESCRIPTION
[0043] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.
[0044] In order to solve the problems existing in the prior art, an embodiment of the present invention provides a method for aligning an unhealthy neural signal with a healthy neural signal.
[0045] Example 1
[0046] The present invention discloses a method for aligning an unhealthy neural signal to a healthy neural signal. Figure 1 ,include:
[0047] S100. Evaluate the behavioral health of the target object to obtain the behavioral health coefficient of the target object; specifically, evaluate the behavioral health of the target object; obtain the behavioral health coefficient of the target object by analyzing the target object's eating, defecation, movement and sleep.
[0048] Specifically, in this embodiment S100, by analyzing the target object's eating, the specific method includes:
[0049] S101. Obtaining a video of the target object eating within a preset period before and after electrode implantation, respectively, and analyzing the video to obtain a video clip of the target object eating;
[0050] Specifically, a video of the target object eating within a preset time before and after electrode implantation is obtained, and the video is analyzed to obtain a total of N 1 target object eating video clips, the i-th video clip gets M i Frame continuous frame image;
[0051] S102. Identify the target object's eating segment, and when there are continuous preset frame images in the eating segment that are identified as abnormal eating, determine that the continuous preset frame images are abnormal eating images; specifically, use a deep learning classification model (such as VGG / Resnet, etc.) to classify N 1 target object eating segments to identify, when the i-th segment M i When there are consecutive preset frame images in the frame image that are identified as eating abnormality, this segment is identified as eating abnormality judgment. At this time, the number of segments belonging to eating abnormality is N 11 ;
[0052] S103. According to the abnormal eating image, obtain the model confidence corresponding to the image; when the i-th video segment M i If the jth frame in the frame image is identified as abnormal eating, the model confidence corresponding to the image is score ij ;
[0053] S104. Obtain the eating status result of the target object before electrode implantation and the eating status result after electrode implantation according to the model confidence corresponding to the image. Calculate the eating status result score before electrode implantation 1q , the score 1q The calculation formula is:
[0054]
[0055] Among them, N 1 is the number of video clips of the target object eating, N 11 is the number of episodes with abnormal eating, M i1 is the i-th video segment M i The number of abnormal eating patterns in the frame image; similarly, the score of eating status after electrode implantation can be obtained. 1h .
[0056] After analyzing the target's eating habits, we then analyze its defecation, movement, and sleep to obtain its behavioral health coefficient. Specific methods include:
[0057] S106. Obtain the defecation, movement and sleep videos of the target object within the preset time before electrode implantation, repeat S101-S105, and obtain the defecation status scores before and after electrode implantation respectively. 2q and score 2h , score of defecation status before and after electrode implantation 3q and score 3h , sleep status score before and after electrode implantation 4q and score 4h ;
[0058] S107. Score based on eating status before and after electrode implantation 1q and score 1h , score of defecation status before and after electrode implantation 2q and score 2h , score of defecation status before and after electrode implantation 3q and score 3h , sleep status score before and after electrode implantation 4q and score 4h, calculate the health factor of the target object
[0059] S200. Identify the stimulus response of the target object; perform multiple stimulations on the target object, obtain the stimulus response time of the target object before and after electrode implantation, and obtain the response delay coefficient of the neural signal by combining the behavioral health coefficient; based on the response delay coefficient, correct the neural signal collected by the electrode to obtain the neural signal U after the response time correction 1 ;
[0060] In S200 of this embodiment, multiple stimulations are performed on the target object, the stimulation reaction time of the target object before and after the electrode implantation is obtained, and the response delay coefficient of the neural signal is obtained by combining the behavioral health coefficient; the specific method includes:
[0061] S201. Stimulate the target object before and after the electrode implantation, wherein the stimulation includes at least motion stimulation, visual stimulation, auditory stimulation and tactile stimulation, obtain a stimulation video of the target object, load a stimulation recognition model, and when the continuous preset frame image is recognized as having stimulation, determine the continuous preset frame image as a stimulation image, and determine a stimulation implementation frame according to the stimulation image;
[0062] S202. Training the target object stimulus response recognition model, when the continuous preset frame image is recognized as a normal response, determining the continuous preset frame image as a normal response image, and determining the stimulus response normal frame according to the normal response image;
[0063] S203. Calculating the stimulation response time of the target object before and after electrode implantation according to the stimulation implementation frame and the stimulation response normal frame;
[0064] S204. Calculate the response delay coefficient according to the target object stimulation reaction time before and after the electrode implantation and the behavioral health coefficient;
[0065] S205. Obtain a corrected response time according to the neural signal acquisition time and the response delay coefficient, and obtain a neural signal after the response time is corrected according to the corrected response time.
[0066] Specifically, before the electrode is implanted, the target is stimulated, the stimulation includes at least motion stimulation, visual stimulation, auditory stimulation and tactile stimulation, the target stimulation video is obtained, the stimulation recognition model is loaded, and when the continuous τ frames are identified as having stimulation, the τ frame picture is obtained [P c0 , P c1 …P ci …P cτ ], among which, the P c0The frame is a behavioral stimulus implementation frame; for example, in roller motion stimulation, a rolling rotation recognition model (which can be Resnet, etc.) is trained, and the model recognizes that the roller starts to work; in light / no light visual stimulation, a light source recognition model (which can be VGG, etc.) is trained, and the model recognizes that a stimulating light source appears in the shooting picture; in auditory stimulation, a speech recognition model (which can be LSTM, etc.) is trained, and the model recognizes that a stimulating speech appears.
[0067] Train the target stimulus response recognition model (for example, in the case of roller motion stimulation, the target response recognition model can be a classification network such as Resnet, which recognizes that the target starts to walk / run; in the case of light / no light visual stimulation, the target response recognition model can be a classification network such as Vgg, which recognizes that the target has closed eyes). When n consecutive frames are recognized as normal action responses, obtain n frames of images Among them, P d0 Frame is the stimulus-response normal frame;
[0068] Calculate the target stimulation reaction time before electrode implantation Where FPS is the frame rate of the shooting device, t dq =[t dq1 ,t dq2 ,t dq3 ,t dq4 ], t dq1 For sports stimulation, dq2 For visual stimulation, dq3 For auditory stimulation, t dq4 For tactile stimulation;
[0069] The electrodes are implanted, and S201-S203 are repeated to calculate the stimulation reaction time t of the target object after the electrodes are implanted. dh ;
[0070] According to the target stimulation reaction time t before and after electrode implantation dq ,t dh and behavioral health coefficient Calculate the response delay coefficient t υ , I is the number of delayed response experiment types;
[0071] In order to obtain the response delay coefficient t υ Then, the neural signal collected by the electrode is corrected based on the response delay coefficient to obtain the neural signal U after the response time correction 1 Specific methods include:
[0072] like Figure 2 , according to the neural signal acquisition time t 采集 and response delay coefficient t υ , and obtain the corrected response time t修正 =t 采集 -t υ , according to the corrected response time t 修正 Get the corrected neural signal U at the response time 1 .
[0073] S300. Perform fever analysis on the target object to obtain characteristic information before and after fever of the target object, wherein the characteristic information before fever of the target object includes the average head temperature before fever, the blood vessel diameter before fever, the equivalent centroid position of the electrode before fever, and the spatial position of each neuron before fever; the characteristic information after fever of the target object includes the average head temperature after fever, the blood vessel diameter after fever, the equivalent centroid position of the electrode after fever, and the spatial position of each neuron after fever.
[0074] Get the average head temperature T before and after fever ZC and T FR , blood vessel diameter d before and after fever q and d h , the equivalent center of mass position of the electrode before and after heating p q and p h , the spatial position of each neuron before and after fever q and h ;
[0075] Specifically, firstly, the average head temperature before fever T is obtained based on the infrared imaging device. ZC Similarly, the average head temperature after fever is obtained FR Then the CT scan image and the MRI vascular diameter scan image are fused to obtain the image img ctmri-j , segment the blood vessels in the image, and mark the blood vessel area to obtain the marked image b_img ctmri-j ; Perform three-dimensional reconstruction on the multi-layer images to obtain the blood vessel diameter d in the three-dimensional images before and after heating q and d h Finally, according to the three-dimensional reconstructed image, the equivalent centroid position p of the electrode before and after heating is obtained in the three-dimensional image. q and ph; for each acquired neural signal, the spike signal is extracted by the threshold method, based on the formula The optimization algorithm is used to solve the minimum problem and calculate the x, y, and z values of each spike segment, where ptp c is the peak-to-peak value of each spike signal, α is the maximum amplitude, and x c ,y c ,z c is the channel position on the electrode, and the spikes obtained are clustered to obtain the spatial position of each neuron before and after fever. q and h.
[0076] S400. Based on the average head temperature T before and after heating ZC and T FR , blood vessel diameter d before and after fever q and d h , the equivalent center of mass position of the electrode before and after heating p q and p h , the spatial position of each neuron before and after fever q and h , get the amplitude and width correction coefficient of the neural signal Correction factor based on the amplitude and width For neural signals 1 Perform width and amplitude correction to obtain the amplitude and width corrected neural signal U 2 ;
[0077] The average head temperature T before and after fever ZC and T FR , blood vessel diameter d before and after fever q and d h , the equivalent center of mass position of the electrode before and after heating p q and p h , the spatial position of each neuron before and after fever q and h After that, the amplitude and width correction factors Among them, d q and d h is the diameter of blood vessels before and after fever, p q and p h is the equivalent center of mass position of the electrode before and after heating, o q and h is the spatial position of each neuron before and after fever, T ZC and T FR is the average head temperature before and after fever; obtain the amplitude and width correction coefficient Then, according to the positional relationship between the cells, electrodes and blood vessels, the corrected width and amplitude are obtained. The specific method includes: Figure 3 When the cells and electrodes are on different sides of the blood vessel, the thickening of the blood vessel squeezes the surrounding cells and electrodes to both sides of the blood vessel (when the distance is sufficient and the blood vessel diameter is appropriate, the electrode can still collect cell signals). The time span between the peak and trough of the spike signal collected after heating is defined as the waveform width, and the time width Δt after heating is obtained h , then the corrected width is Get the maximum amplitude including the negative maximum amplitude and the maximum positive amplitude Amplitude Correction When the cell and the electrode are on the same side of the blood vessel, when the blood vessel thickens, the distance between the cell and the electrode will become closer (the electrode is fixed on the skull, and its moving distance is relatively small compared to the cell moving distance, so the cell tends to get closer to the electrode). When the distance between the cell and the electrode decreases, the amplitude of the signal collected after the distance is shortened increases and the waveform width becomes narrower than the signal collected before the distance changes. , then the corrected width is Amplitude correction Get the amplitude and width corrected neural signal U 2 .
[0078] S500. Analyze the target's cells across functional areas; remove noise cells by identifying stimulation patterns, segmenting functional areas, and distributing neurons, and obtain the final quality-enhanced neural signal set U 3 .
[0079] In S500 of this embodiment, by identifying the stimulation pattern, segmenting the functional area, distributing the neurons, and removing the noise cells, the final quality-enhanced neural signal set U is obtained. 3 The specific method includes: when conducting a light / no light visual stimulation experiment and generating a high-decibel noise, first identify the stimulation pattern and train an abnormal sound recognition model. When the abnormal sound reaches the stress response threshold of the neurons in the auditory functional area of the test target, the model automatically runs and gives a prompt; then segment the functional area, train the functional area segmentation model, and perform the fusion of the tomographic image img ctmri-j Each functional area in the image is segmented to obtain b_img with functional area boundary lines ctmri-j , then for the multi-layer image b_img ctmri-j Perform three-dimensional reconstruction; finally, the neuron spatial coordinates o h Match and divide each functional area in the 3D reconstructed 3D image, and find the neurons in the auditory functional area. hT , and form a neural signal set U 2T Based on the RNN neural network model, a visual or auditory response recognition model is trained to identify the U 2T The neural signals contained are classified to obtain the auditory response pattern neural signal set U 2T_T ;Change U 2T_T From U 2 The final quality-enhanced neural signal set U is obtained 3 , where U 3 ∈U 2 .
[0080] The present invention discloses a method for aligning an unhealthy neural signal to a healthy neural signal, comprising: evaluating the behavioral health of a target object; obtaining a behavioral health coefficient of the target object by analyzing the eating, defecation, movement and sleep of the target object. Identify the stimulus response of the target object; perform multiple stimulations on the target object, obtain the stimulus response time of the target object before and after electrode implantation, and obtain the response delay coefficient of the neural signal by combining the behavioral health coefficient; based on the response delay coefficient, correct the neural signal collected by the electrode to obtain the neural signal U after the response time correction 1 ; Analyze the heat of the target object; obtain the average head temperature T before and after the heat ZC and T FR , blood vessel diameter d before and after fever q and d h , the equivalent center of mass position of the electrode before and after heating p q and p h , the spatial position of each neuron before and after fever q and h ; According to the average head temperature T before and after the fever ZC and T FR , blood vessel diameter d before and after fever q and d h , the equivalent center of mass position of the electrode before and after heating p q and p h , the spatial position of each neuron before and after fever q and h , get the amplitude and width correction coefficient of the neural signal Correction factor based on the amplitude and width For neural signals 1 Perform width and amplitude correction to obtain the amplitude and width corrected neural signal U 2 ; Analyze the target's cells across functional areas; By identifying the stimulation pattern, segmenting the functional areas and distributing the neurons, the noise cells are eliminated to obtain the final quality-enhanced neural signal set U 3 .
[0081] This embodiment adjusts the response time of the neural signal, corrects the waveform amplitude and width, and removes noise based on factors such as the target's behavioral health status, stimulus response, body temperature, and cross-functional noise cells, thereby aligning the quality of neural signals in unhealthy conditions to neural signals in healthy conditions and cleaning noise signals. This solves the problem that the current collection of neural signals does not take into account the target's health status, and the collection of neural signals is inaccurate when the organism is in an unhealthy state.
[0082] Example 2
[0083] Based on the same inventive concept, an embodiment of the present disclosure also provides an electronic device. Figure 4 FIG. 1 is a schematic diagram of the structure of an electronic device according to an embodiment of the present disclosure. Figure 2 As shown, an embodiment of the present disclosure provides an electronic device including: one or more processors 101, a memory 102, and one or more I / O interfaces 103. The memory 102 stores one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement any optimization method in the above embodiments; the one or more I / O interfaces 103 are connected between the processor and the memory, and are configured to implement information interaction between the processor and the memory.
[0084] Among them, the processor 101 is a device with data processing capabilities, including but not limited to a central processing unit (CPU), etc.; the memory 102 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory (FLASH); the I / O interface (read-write interface) 103 is connected between the processor 101 and the memory 102, and can realize information interaction between the processor 101 and the memory 102, including but not limited to a data bus (Bus), etc.
[0085] In some embodiments, the processor 101 , the memory 102 , and the I / O interface 103 are connected to each other via a bus 104 , and further connected to other components of the computing device.
[0086] In some embodiments, the one or more processors 101 include a field programmable gate array.
[0087] According to an embodiment of the present disclosure, a computer-readable medium is further provided, wherein a computer program is stored on the computer-readable medium, wherein when the program is executed by a processor, the steps in any optimization method in the above-mentioned embodiment are implemented.
[0088] It should be understood that the specific order or hierarchy of steps in the disclosed process is an example of an exemplary method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process can be rearranged without departing from the scope of protection of the present disclosure. The attached method claims present the elements of the various steps in an exemplary order and are not intended to be limited to the specific order or hierarchy described.
[0089] In the above detailed description, various features are grouped together in a single embodiment to simplify the disclosure. This method of disclosure should not be interpreted as reflecting an intention that the embodiments of the claimed subject matter require more features than are clearly stated in each claim. On the contrary, as reflected in the appended claims, the invention is in a state of having less than all the features of the disclosed individual embodiments. Therefore, the appended claims are hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate preferred embodiment of the invention.
[0090] Those skilled in the art will also appreciate that the various illustrative logic blocks, modules, circuits, and algorithmic steps described in conjunction with the embodiments herein can all be implemented as electronic hardware, computer software, or a combination thereof. In order to clearly illustrate the interchangeability between hardware and software, various illustrative components, blocks, modules, circuits, and steps are generally described above around their functions. Whether such functions are implemented as hardware or software depends on specific applications and the design constraints imposed on the entire system. A skilled person can implement the described functions in an alternative manner for each specific application, but such implementation decisions should not be interpreted as departing from the scope of protection of the present disclosure.
[0091] The steps of the method or algorithm described in conjunction with the embodiments herein may be directly embodied as hardware, a software module executed by a processor, or a combination thereof. The software module may be located in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a mobile disk, a CD-ROM, or any other form of storage medium well known in the art. An exemplary storage medium is connected to the processor so that the processor can read information from the storage medium and can write information to the storage medium. Of course, the storage medium may also be an integral part of the processor. The processor and the storage medium may be located in an ASIC. The ASIC may be located in a user terminal. Of course, the processor and the storage medium may also be present in a user terminal as discrete components.
[0092] For software implementation, the techniques described in this application can be implemented with modules (e.g., procedures, functions, etc.) that perform the functions described in this application. These software codes can be stored in a memory unit and executed by a processor. The memory unit can be implemented within the processor or outside the processor. In the latter case, it is coupled to the processor in a communication manner via various means, which are well known in the art.
[0093] The above description includes examples of one or more embodiments. Of course, it is impossible to describe all possible combinations of components or methods for the purpose of describing the above embodiments, but it should be recognized by those skilled in the art that the various embodiments may be further combined and arranged. Therefore, the embodiments described herein are intended to cover all such changes, modifications and variations that fall within the scope of protection of the appended claims. In addition, with respect to the term "comprising" used in the specification or claims, the word is covered in a manner similar to the term "including", just as "including," is explained as a transitional word in the claims. In addition, any term "or" used in the specification of the claims is intended to mean "non-exclusive or".
Claims
1. A method for aligning an unhealthy neural signal to a healthy neural signal, characterized in that: include: S100. Evaluate the behavioral health of the target object to obtain a behavioral health coefficient of the target object; S200. Identify the stimulus response of the target object, and obtain the response delay coefficient of the neural signal in combination with the behavioral health coefficient; obtain the neural signal after the response time correction based on the response delay coefficient; S300. Performing fever analysis on the target object to obtain characteristic information before and after fever of the target object; S400. Obtaining amplitude and width correction coefficients of the neural signal according to the characteristic information before and after the target object is heated, and performing width and amplitude correction on the neural signal corrected at the response time according to the amplitude and width correction coefficients to obtain a neural signal after amplitude and width correction; S500. Analyze the target object across functional areas, remove the noise cells, and obtain the final quality-enhanced neural signal.
2. A method for aligning an unhealthy neural signal to a healthy neural signal as claimed in claim 1, characterized in that: The target's behavioral health assessment at least includes eating analysis, defecation analysis, movement analysis and sleep analysis; wherein the eating analysis of the target includes: S101. Obtaining a video of the target object eating within a preset period before and after electrode implantation, respectively, and analyzing the video to obtain a video clip of the target object eating; S102. Identifying the target object's eating segment, and when there are continuous preset frame images in the eating segment that are identified as abnormal eating, determining the continuous preset frame images as abnormal eating images; S103. According to the abnormal eating image, obtain the model confidence corresponding to the image; S104. According to the model confidence corresponding to the image, the eating status results of the target object before electrode implantation and the eating status results after electrode implantation are obtained respectively.
3. A method for aligning an unhealthy neural signal to a healthy neural signal as claimed in claim 2, characterized in that: Analyze the defecation, action and sleep of the target to obtain the behavioral health coefficient of the target. The specific methods include: S105. Obtain the defecation, movement and sleep videos of the target object within a preset time before and after electrode implantation, repeat S101-S104, and obtain the defecation status results, electrical defecation status results and sleep status results before and after electrode implantation respectively; S106. Obtain the health coefficient of the target object based on the target object's eating results, defecation results, electrical defecation results and sleeping results before and after electrode implantation.
4. The method for aligning an unhealthy neural signal to a healthy neural signal according to claim 1, characterized in that: In S200, a target object is subjected to multiple stimulations to obtain the stimulation reaction time of the target object before and after electrode implantation, and a response delay coefficient of the neural signal is obtained by combining the behavioral health coefficient. The neural signal collected by the electrode is corrected based on the response delay coefficient to obtain a neural signal with a corrected response time. The specific method includes: S201. Stimulate the target object before and after the electrode implantation, wherein the stimulation includes at least motion stimulation, visual stimulation, auditory stimulation and tactile stimulation, obtain a stimulation video of the target object, load a stimulation recognition model, and when the continuous preset frame image is recognized as having stimulation, determine the continuous preset frame image as a stimulation image, and determine a stimulation implementation frame according to the stimulation image; S202. Training the target object stimulus response recognition model, when the continuous preset frame image is recognized as a normal response, determining the continuous preset frame image as a normal response image, and determining the stimulus response normal frame according to the normal response image; S203. Calculating the stimulation response time of the target object before and after electrode implantation according to the stimulation implementation frame and the normal stimulation response frame; S204. Calculate the response delay coefficient according to the target object stimulation reaction time before and after the electrode implantation and the behavioral health coefficient; S205. Obtain a corrected response time according to the neural signal acquisition time and the response delay coefficient, and obtain a neural signal after the response time is corrected according to the corrected response time.
5. The method for aligning an unhealthy neural signal to a healthy neural signal according to claim 1, characterized in that: In S300, the characteristic information of the target object before fever includes the average head temperature before fever, the blood vessel diameter before fever, the equivalent centroid position of the electrode before fever, and the spatial position of each neuron before fever; the characteristic information of the target object after fever includes the average head temperature after fever, the blood vessel diameter after fever, the equivalent centroid position of the electrode after fever, and the spatial position of each neuron after fever.
6. A method for aligning an unhealthy neural signal to a healthy neural signal as claimed in claim 5, characterized in that: The average head temperature of the target object before and after fever, and the blood vessel diameter before and after fever are obtained. The specific method includes: based on an infrared imaging device, obtaining the average head temperature of the target object before and after fever; fusing the obtained CT tomography image and the MRI tomography blood vessel diameter scanning image to obtain a fused image, segmenting the blood vessels in the fused image, and marking the blood vessel area to obtain a marked image; performing three-dimensional reconstruction on the marked image to obtain a three-dimensional reconstructed image, and obtaining the blood vessel diameters before and after fever according to the three-dimensional reconstructed image.
7. A method for aligning an unhealthy neural signal to a healthy neural signal as claimed in claim 6, characterized in that: The equivalent centroid position of the electrode before and after fever, and the spatial position of each neuron before and after fever are obtained. The specific method includes: obtaining the equivalent centroid position of the electrode before and after fever according to the three-dimensional reconstructed image; extracting the spike signal by the threshold method for each acquired neural signal, calculating the three-dimensional coordinate value of each spike segment based on a preset formula and using an optimization algorithm, clustering the three-dimensional coordinate values of the spike segments, and obtaining the spatial position of each neuron before and after fever.
8. A method for aligning an unhealthy neural signal to a healthy neural signal as claimed in claim 7, characterized in that: In S400, after obtaining the amplitude and width correction coefficients, the corrected width and amplitude are obtained according to the positional relationship between the cells and electrodes and the blood vessels. The specific method includes: when the cells and electrodes are on both sides of the blood vessel, the time span between the peak and the trough of the spike signal collected after fever is defined as the waveform width, and the time width after fever is obtained, and the corrected time width is the product of the time width after fever and the amplitude and width correction coefficients; the maximum amplitude after fever is obtained, and the corrected maximum amplitude is the quotient of the maximum amplitude after fever and the amplitude and width correction coefficients; when the cells and electrodes are on the same side of the blood vessel, the corrected time width is the quotient of the time width after fever and the amplitude and width correction coefficients, and the corrected maximum amplitude is the product of the maximum amplitude after fever and the amplitude and width correction coefficients; according to the corrected time width and the corrected maximum amplitude, the neural signal with amplitude and width correction is obtained.
9. The method for aligning an unhealthy neural signal to a healthy neural signal according to claim 7, characterized in that: In S500, the target object is analyzed across functional areas, specifically including stimulation pattern recognition, functional area segmentation and neuron distribution. The specific method includes: when a visual stimulation experiment with or without light is conducted and a high-decibel noise is generated, the stimulation pattern is firstly recognized and an abnormal sound recognition model is trained. When the abnormal sound reaches the threshold of the target object's auditory functional area neurons to generate a stress response, the abnormal sound recognition model automatically runs and gives a prompt; then the functional area is segmented, the functional area segmentation model is trained, each functional area in the fused image is segmented, and a segmented image with functional area boundary lines is obtained, and then the segmented image is three-dimensionally reconstructed; finally, the neuron spatial coordinates are matched and divided with each functional area in the three-dimensional reconstructed image, the neurons of the auditory functional area are obtained, and a neural signal set is formed according to the neurons; based on the RNN neural network model, a visual or auditory response recognition model is trained, and the neural signals forming the neural signal set are classified to obtain an auditory response pattern neural signal set; the auditory response pattern neural signal set is eliminated from the amplitude and width corrected neural signals to obtain a final quality-enhanced neural signal set.
10. An electronic device, comprising: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the signal alignment method in claims 1-9.
Citation Information
Patent Citations
Bio information measurement device and bio information measurement method
CN104936514A
Neural stimulation signal adjusting method and device
CN114534096A
Electroencephalogram signal correction method and device based on human factor intelligence
CN117982151A
Systems and methods for real-time monitoring of physiological biomarkers through nerve signals and uses thereof
US20200179600A1
System and method of eliminating systematic noise in stimulus-response systems
US5640124A
Cited By
System and method for evaluating health status of poultry bred under forest
CN120226999A
