Neural regulation and control method based on multi-target point cooperation

Abstract

The invention relates to a nerve regulation and control method based on multi-target point cooperation, and the method comprises the following steps: building a dual-channel photocurrent model, building a quick response type photocurrent model for a PV channel, and building a continuous inhibition type photocurrent model for an SOM channel; establishing an electromagnetic coupling model, determining an electromagnetic induction stimulation voltage according to a memory equation of the memristor, and coupling the light current and the electromagnetic induction stimulation voltage determined based on the dual-channel light current model to obtain a total current; performing multi-target cooperative control effect analysis on the electromagnetic coupling model, and optimizing off-line parameters to obtain an off-line optimal parameter combination; and performing nerve regulation based on the electromagnetic coupling model after the offline parameter optimization, and performing online optimization of the operation parameters according to the result of the nerve regulation. Compared with the prior art, the method has the advantages of being capable of achieving accurate regulation and control of different target spots and the like.

Description

Technical Field

[0001] The present invention relates to the field of neuromodulation technology, and in particular to a neuromodulation method based on multi-target collaboration. Background Art

[0002] Optogenetics, a technology that combines optics and genetics, can precisely control the activity of specific neurons. In recent years, due to its high spatiotemporal precision and cell-specificity, optogenetics has been widely used to precisely control the activity of specific neurons. Neuromodulation, the study and application of neuromodulation methods to the nervous system, is of great significance for neuroscience research and the treatment of neurological diseases. For example, CN114042251A discloses a multi-target optical, magnetic, and electric coupled neuromodulation device and method. The device includes: an optical, magnetic and electric coupling module, including a transcranial magnetic stimulation module, a neuromuscular electrical stimulation module and a multi-channel optical signal acquisition module; a neuroregulatory analysis module, including a signal transmission module, a comparative analysis module and a data storage module for analysis results; a parameter control module, which can dynamically adjust the stimulation mode and stimulation parameters of the transcranial magnetic stimulation module and the neuromuscular electrical stimulation module according to the analysis and comparison results of the cerebral cortex optical signal by the neuroregulatory analysis module, so as to improve the effectiveness of transcranial magnetic stimulation and neuromuscular electrical stimulation. The transcranial magnetic stimulation module and the neuromuscular electrical stimulation module can stimulate multiple brain areas and limbs at the same time, and can also stimulate multiple target points in a time sequence. The comparative analysis module can retrieve the indicator data at any time in the data access module, and by comparing the indicator information before and after the two moments, it can determine the dose-effect of neuroregulation. However, this method has the following limitations:

[0003] 1. Lack of real-time feedback control of parameters makes it difficult to achieve accurate regulation of multiple targets;

[0004] 2. Energy allocation optimization is not considered, and the energy efficiency of optogenetic regulation is low. Summary of the Invention

[0005] The purpose of the present invention is to provide a neural regulation method based on multi-target collaboration, using optogenetic technology to achieve differentiated and precise regulation of PV and SOM neurons and improve energy efficiency.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] A multi-target synergistic neural regulation method comprises the following steps:

[0008] A dual-pathway photocurrent model was established, in which a fast-response photocurrent model was established for the PV path and a continuous suppression photocurrent model was established for the SOM path;

[0009] Establishing an electromagnetic coupling model, wherein the electromagnetic coupling model determines an electromagnetic induction stimulation voltage based on a memristor memory equation, and couples the photocurrent determined based on a dual-pathway photocurrent model and the electromagnetic induction stimulation voltage to obtain a total current, wherein the total current serves as a control signal for neural regulation;

[0010] Analyze the multi-target coordinated control effect of the electromagnetic coupling model, optimize the offline parameters, and obtain the offline optimal parameter combination;

[0011] Neuroregulation is performed based on the electromagnetic coupling model after offline parameter optimization, and the operating parameters are optimized online in real time according to the results of neuroregulation.

[0012] The fast-response photocurrent model of the PV path is expressed as:

[0013]

[0014] in, represent the photocurrent and control potential of the PV path, respectively, g pv represents PV conductivity, V represents the transmembrane voltage of the neuron, represents the probability of the photosensitive channel of the PV pathway being in the open state, γ pv Indicates that the attachment is in the open state The adjustment coefficient on ChR2 represents the reversal potential of ChR2, R m represents the average resistance of the stimulation group, N pv Indicates the channel density of the PV pathway.

[0015] The sustained suppression photocurrent model of the SOM pathway is expressed as:

[0016]

[0017] in, represent the photocurrent and control potential of the SOM pathway, g som represents the SOM conductivity, V represents the transmembrane voltage of the neuron, represents the probability of the photosensitive channel of the SOM pathway being in the open state, γ som Indicates that the attachment is in the open state The adjustment coefficient on ChR2 represents the reversal potential of ChR2, R m represents the average resistance of the stimulation group, N som represents the channel density of the SOM pathway.

[0018] The memristor memory equation is expressed as:

[0019] ρ(φ)=m1+n1cos(φ)

[0020] E=ρ(φ)PN

[0021]

[0022] Among them, m1 is the baseline conductivity, n1 is the memory strength, φ is the magnetic flux, PN is the related variable of the cortical principal neuron, E is the electromagnetic induction stimulation voltage, g0 is the magnetic flux feedback gain, g1 is the principal neuron activity gain, and g2 is the basic rate constant.

[0023] The electromagnetic coupling model is expressed as:

[0024] I total =I ChR2 +kE

[0025] Among them, I ChR2 is the photocurrent, E is the electromagnetic induction stimulation voltage, k is the electromagnetic coupling coefficient, I total is the total current.

[0026] The offline parameter optimization of the electromagnetic coupling model comprises the following steps:

[0027] Photocurrent calculation baseline based on dual-path photocurrent model;

[0028] Calculate electromagnetic interference effects based on total current;

[0029] Calculate the enhancement effect based on the electromagnetic interference effect and the baseline;

[0030] If the enhancement effect is greater than the preset threshold, the current parameters are recorded as the offline optimal parameter combination.

[0031] The optimization parameters of the offline parameter optimization include magnetic flux feedback gain, electromagnetic coupling coefficient and memory strength.

[0032] The operating parameters include light intensity, electromagnetic interference parameters and energy distribution parameters, wherein the electromagnetic interference parameters include magnetic flux feedback gain, main neuron activity gain, basic rate constant, electromagnetic coupling coefficient, and memory strength; the energy distribution parameters include pulse sequence parameters and energy distribution weight.

[0033] The real-time online optimization of operating parameters comprises the following steps:

[0034] Extract multimodal features based on the results of neural regulation;

[0035] Construct a state evaluation function based on the extracted multimodal features and energy allocation weights;

[0036] If the state evaluation function is greater than the preset threshold, the light intensity and electromagnetic interference parameters are adjusted; otherwise, the evaluation effect index is calculated;

[0037] If the evaluation effect index is less than the preset target value, the stimulation intensity is increased; otherwise, the stimulation intensity is maintained or reduced;

[0038] Energy allocation optimization is performed based on a multi-objective function to optimize light intensity and energy allocation parameters. The multi-objective function includes minimizing total energy consumption, maximizing abnormal discharge suppression efficiency, and maximizing stability.

[0039] The multimodal features include membrane potential features, peak discharge features and bifurcation point features.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] (1) This invention establishes different types of photocurrent models for the PV pathway and the SOM pathway, respectively. It also establishes an electromagnetic coupling model based on the memristor memory equation. The total current obtained by coupling the photocurrent and electromagnetic induction stimulation voltage is used as the control signal for neural regulation. This targeted modeling of different neural pathways and the optical-electromagnetic coupling method can more accurately simulate neural activity and provide a theoretical basis for the accurate regulation of multiple targets.

[0042] (2) The present invention combines the established model with offline parameter optimization and online parameter optimization processes to achieve accurate multi-target control. Among them, the offline parameter optimization analyzes the multi-target coordinated control effect of the electromagnetic coupling model and obtains the offline optimal parameter combination, which helps to set a suitable parameter basis for multi-target control in the initial stage and improve the accuracy of control; and in the process of neural control, the operating parameters are optimized online in real time according to the control results. This real-time feedback control mechanism can respond to changes in the neural control process in a timely manner, further ensuring accurate control of multiple targets.

[0043] (3) The present invention optimizes energy allocation based on a multi-objective function. The multi-objective function comprehensively considers multiple aspects such as energy consumption, regulation effect and stability. It can effectively reduce energy consumption while ensuring the neural regulation effect and improve the energy efficiency of optogenetic regulation.

[0044] (4) The present invention extracts multimodal features such as membrane potential characteristics, peak discharge characteristics, and bifurcation point characteristics during online optimization of operating parameters. These multimodal features can reflect the effects of neural regulation and the state of neural activity from different perspectives, providing more comprehensive and accurate information than single features. Constructing a state evaluation function based on these multimodal features and performing subsequent parameter adjustments can more accurately grasp the progress of neural regulation and improve the effectiveness and pertinence of regulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a flow chart of the method of the present invention;

[0046] Figure 2A state transition analysis diagram in one embodiment;

[0047] Figure 3 PN extreme value analysis diagram under different stimulation intensities in one embodiment;

[0048] Figure 4 This is a PN extreme value analysis diagram under dual stimulation synergy in an embodiment. DETAILED DESCRIPTION

[0049] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0050] Example 1

[0051] This embodiment provides a neural regulation method based on multi-target coordination, such as Figure 1 As shown, the following steps are included:

[0052] S1, establish a dual-path photocurrent model, wherein a fast response photocurrent model is established for the PV path, and a continuous suppression photocurrent model is established for the SOM path.

[0053] The fast response photocurrent model of the PV path is expressed as:

[0054]

[0055] in, represent the photocurrent and control potential of the PV path, respectively, g pv represents PV conductivity, V represents the transmembrane voltage of the neuron, represents the probability of the photosensitive channel of the PV pathway being in the open state, γ pv Indicates that the attachment is in the open state The adjustment coefficient on ChR2 represents the reversal potential of ChR2, R m represents the average resistance of the stimulation group, N pv Indicates the channel density of the PV pathway.

[0056] The sustained suppression photocurrent model of the SOM pathway is expressed as:

[0057]

[0058] in, represent the photocurrent and control potential of the SOM pathway, g som represents the SOM conductivity, represents the probability of the photosensitive channel of the SOM pathway being in the open state, γ somIndicates that the attachment is in the open state The adjustment coefficient on som represents the channel density of the SOM pathway.

[0059] In this embodiment, different parameters are set for the two photocurrent models. For the PV path, g pv =0.4mS / cm 2 , γ pv =0.1, N pv =1×10 7 cm -2 ; For the SOM pathway, g som =0.5mS / cm 2 , γ som =0.15, N som =8×10 6 cm -2 .

[0060] Based on the dual-path photocurrent model established above, this embodiment conducted experiments and obtained the following path characteristic comparison results:

[0061] PV path response time: <2ms;

[0062] SOM path response time: <3ms;

[0063] Channel crosstalk: <5%.

[0064] The control effects are as follows:

[0065] PV pathway inhibition efficiency: 75%;

[0066] SOM pathway inhibition efficiency: 65%;

[0067] System stability: 90%.

[0068] S2, considering the enhancing effect of electromagnetic induction on optogenetic regulation, an electromagnetic coupling model is established. The electromagnetic coupling model determines the electromagnetic induction stimulation voltage based on the memristor memory equation, and couples the photocurrent and electromagnetic induction stimulation voltage determined based on the dual-pathway photocurrent model to obtain the total current, which serves as the control signal for neural regulation.

[0069] In this embodiment, the memristor memory equation is expressed as:

[0070] ρ(φ)=m1+n1cos(φ)

[0071] E=ρ(φ)PN

[0072]

[0073] Among them, m1 is the baseline conductivity, n1 is the memory strength, φ is the magnetic flux, PN is the related variable of the cortical principal neuron, E is the electromagnetic induction stimulation voltage, g0 is the magnetic flux feedback gain, g1 is the principal neuron activity gain, and g2 is the basic rate constant.

[0074] Considering the coupling effect, the electromagnetic coupling model is expressed as:

[0075] I total =I ChR2 +kE

[0076] Among them, I ChR2 is the photocurrent, E is the electromagnetic induction stimulation voltage, k is the electromagnetic coupling coefficient, I total is the total current.

[0077] S3, analyze the multi-target collaborative control effect of the electromagnetic coupling model, optimize the offline parameters, and obtain the offline optimal parameter combination.

[0078] The optimization parameters for offline parameter optimization include the magnetic flux feedback gain g0, the electromagnetic coupling coefficient k, and the memory strength n1. By optimizing the memory strength n1, an adjustable memory effect can be achieved. Specifically, the offline optimization process includes the following steps:

[0079] S31, baseline for photocurrent calculation based on the dual-pathway photocurrent model;

[0080] S32, calculating electromagnetic interference effect based on the total current;

[0081] S33, calculation of enhancement effect based on electromagnetic interference effect and baseline;

[0082] S34: If the enhancement effect is greater than a preset threshold, the current parameters are recorded as an offline optimal parameter combination.

[0083] The corresponding code snippet is as follows:

[0084]

[0085] Taking into account the synergistic enhancement effect of electromagnetic coupling, this embodiment has been experimentally verified to achieve photocurrent amplification, shortened response time, and improved spatial accuracy. The experimental results are as follows:

[0086] 1. Electromagnetic enhancement effect: photocurrent increased by 45%, response time shortened by 35%, energy consumption reduced by 40%.

[0087] 2. The offline optimal parameter combination is: flux feedback gain g0 = 0.3, coupling coefficient k = 0.25, and memory strength n1 = 0.02.

[0088] Figure 2The results show that the SRN→PV connection strength (p 15 ) and PV→PN connection strength (p2), and the dynamic evolution of the system after the addition of electromagnetic induction. Color coding shows the state distribution of the system under different parameter combinations, where blue areas indicate normal states and red areas indicate pathological states.

[0089] S4, performs neural regulation based on the electromagnetic coupling model after offline parameter optimization, and optimizes the operating parameters in real time online according to the results of neural regulation.

[0090] By optimizing operating parameters online in real time, neuronal activity and abnormal discharge patterns can be analyzed in real time, thus achieving more precise regulation.

[0091] In this embodiment, the operating parameters include light intensity, electromagnetic interference parameters and energy distribution parameters, wherein the electromagnetic interference parameters include magnetic flux feedback gain g0, main neuron activity gain g1, basic rate constant g2, electromagnetic coupling coefficient k, memory strength n1, and the energy distribution parameters include pulse sequence parameters (including pulse frequency, duty cycle, duration, etc.) and energy distribution weights.

[0092] Real-time online optimization of operating parameters includes the following steps:

[0093] S41, extracting multimodal features based on the results of neural regulation, including membrane potential feature F1 = extract_membrane_potential (V), spike discharge feature F2 = extract_firing_rate (spikes), and bifurcation point feature F3 = extract_field_potential (LFP);

[0094] S42, construct a state evaluation function based on the extracted multimodal features and energy allocation weights:

[0095] State=w1F1+w2F2+w3F3

[0096] Among them, w1, w2, and w3 are the energy distribution weights of the corresponding modes, and the default initial values are w1 = 0.5, w2 = 0.3, and w3 = 0.2.

[0097] S43, if the state evaluation function State is greater than a preset threshold, then adjust the light intensity and electromagnetic interference parameters, otherwise, calculate the evaluation effect indicator current_effect;

[0098] S44, if the evaluation effect indicator current_effect is less than the preset target value target, then increase the stimulation intensity; otherwise, maintain or reduce the stimulation intensity;

[0099] S45, performs energy allocation optimization based on a multi-objective function to optimize light intensity and energy allocation parameters. The multi-objective function includes minimizing total energy consumption, maximizing abnormal discharge suppression efficiency, and maximizing stability.

[0100] The optimization and solution process of the multi-objective function will not be described in detail in this embodiment. The applicable solution methods include, but are not limited to, genetic algorithms, particle swarm optimization algorithms, ant colony algorithms, etc. In a preferred embodiment, the preset threshold can be dynamically updated and adjusted according to the results of the real-time membrane potential analysis.

[0101] This embodiment also conducts an experimental analysis on the online optimization process, and the experimental results are as follows:

[0102] 1. Path characteristics:

[0103] PV channel response time: 1.8ms, control accuracy reaches 90%;

[0104] SOM pathway response time: 2.5ms, spatial selectivity up to 92%;

[0105] Inter-channel interference <5%, system stability >90%.

[0106] 2. Regulation effect

[0107] PV channel: SWDs frequency is reduced by 75% and energy consumption is reduced by 45%;

[0108] SOM pathway: abnormal discharge inhibition rate 62%;

[0109] System as a whole: Treatment efficiency is 85%.

[0110] Figure 3 These extreme value analysis plots at different stimulation intensities demonstrate the effects of dual stimulation on PV and SOM neurons. The red and blue colors represent the maximum and minimum membrane potentials, visually demonstrating the modulatory effects of dual stimulation on neuronal activity.

[0111] Figure 4 This is an analysis diagram of the synergistic effect of dual stimulation, showing the regulatory effect of optogenetic stimulation on PV and SOM neurons under different magnetic flux feedback gains (g0), highlighting the synergistic characteristics of electromagnetic induction and optogenetic stimulation.

[0112] Example 2

[0113] The above is an introduction to the method embodiment. The following further illustrates the solution of the present invention through a system embodiment.

[0114] A neural regulation system based on multi-target collaboration includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the method described in Example 1 when executing the program.

[0115] In one embodiment, the system includes:

[0116] Dual-path photocurrent model establishment module: establishes a dual-path photocurrent model, wherein a fast-response photocurrent model is established for the PV path, and a continuous suppression photocurrent model is established for the SOM path;

[0117] Electromagnetic coupling model establishment module: establishes an electromagnetic coupling model, wherein the electromagnetic coupling model determines the electromagnetic induction stimulation voltage based on the memristor memory equation, and couples the photocurrent determined based on the dual-path photocurrent model and the electromagnetic induction stimulation voltage to obtain a total current, which serves as a control signal for neural regulation;

[0118] Offline optimization module: Analyze the multi-target coordinated control effect of the electromagnetic coupling model, optimize offline parameters, and obtain the offline optimal parameter combination;

[0119] Online optimization module: performs neural regulation based on the electromagnetic coupling model after offline parameter optimization, and optimizes operating parameters in real time online according to the results of neural regulation.

[0120] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0121] The above describes in detail the preferred embodiments of the present invention. It should be understood that those skilled in the art can make numerous modifications and variations based on the concepts of the present invention without inventive effort. Therefore, any technical solutions that can be derived by those skilled in the art through logical analysis, reasoning, or limited experimentation based on the concepts of the present invention and the prior art should be within the scope of protection defined by the claims.

Claims

1. A neural regulation method based on multi-target collaboration, characterized in that: The following steps are involved: A dual-pathway photocurrent model was established, in which a fast-response photocurrent model was established for the PV path and a continuous suppression photocurrent model was established for the SOM path; Establishing an electromagnetic coupling model, wherein the electromagnetic coupling model determines an electromagnetic induction stimulation voltage based on a memristor memory equation, and couples the photocurrent determined based on a dual-pathway photocurrent model and the electromagnetic induction stimulation voltage to obtain a total current, wherein the total current serves as a control signal for neural regulation; Analyze the multi-target coordinated control effect of the electromagnetic coupling model, optimize the offline parameters, and obtain the offline optimal parameter combination; Neuroregulation is performed based on the electromagnetic coupling model after offline parameter optimization, and the operating parameters are optimized online in real time according to the results of neuroregulation.

2. A multi-target synergistic neural regulation method according to claim 1, characterized in that: The fast-response photocurrent model of the PV path is expressed as: in, represent the photocurrent and control potential of the PV path, respectively, g pv represents PV conductivity, V represents the transmembrane voltage of the neuron, represents the probability of the photosensitive channel of the PV pathway being in the open state, γ pv Indicates that the attachment is in the open state The adjustment coefficient on ChR2 represents the reversal potential of ChR2, R m represents the average resistance of the stimulation group, N pv Indicates the channel density of the PV pathway.

3. The multi-target synergistic neural regulation method according to claim 1, characterized in that: The sustained suppression photocurrent model of the SOM pathway is expressed as: in, represent the photocurrent and control potential of the SOM pathway, g som represents the SOM conductivity, V represents the transmembrane voltage of the neuron, represents the probability of the photosensitive channel of the SOM pathway being in the open state, γ som Indicates that the attachment is in the open state The adjustment coefficient on ChR2 represents the reversal potential of ChR2, R m represents the average resistance of the stimulation group, N som represents the channel density of the SOM pathway.

4. The multi-target synergistic neural regulation method according to claim 1, characterized in that: The memristor memory equation is expressed as: ρ(φ)=m1+n1 cos(φ) E=ρ(φ)PN Among them, m1 is the baseline conductivity, n1 is the memory strength, φ is the magnetic flux, PN is the related variable of the cortical principal neuron, E is the electromagnetic induction stimulation voltage, g0 is the magnetic flux feedback gain, g1 is the principal neuron activity gain, and g2 is the basic rate constant.

5. The multi-target synergistic neural regulation method according to claim 1, characterized in that: The electromagnetic coupling model is expressed as: I total =I ChR2 +kE Among them, I ChR2 is the photocurrent, E is the electromagnetic induction stimulation voltage, k is the electromagnetic coupling coefficient, I total is the total current.

6. The multi-target synergistic neural regulation method according to claim 1, characterized in that: The offline parameter optimization of the electromagnetic coupling model comprises the following steps: Photocurrent calculation baseline based on dual-path photocurrent model; Calculate electromagnetic interference effects based on total current; Calculate the enhancement effect based on the electromagnetic interference effect and the baseline; If the enhancement effect is greater than the preset threshold, the current parameters are recorded as the offline optimal parameter combination.

7. The multi-target synergistic neural regulation method according to claim 1, characterized in that: The optimization parameters of the offline parameter optimization include magnetic flux feedback gain, electromagnetic coupling coefficient and memory strength.

8. The multi-target synergistic neural regulation method according to claim 1, characterized in that: The operating parameters include light intensity, electromagnetic interference parameters and energy distribution parameters, wherein the electromagnetic interference parameters include magnetic flux feedback gain, main neuron activity gain, basic rate constant, electromagnetic coupling coefficient, and memory strength; the energy distribution parameters include pulse sequence parameters and energy distribution weight.

9. The multi-target synergistic neural regulation method according to claim 8, characterized in that: The real-time online optimization of operating parameters comprises the following steps: Extract multimodal features based on the results of neural regulation; Construct a state evaluation function based on the extracted multimodal features and energy allocation weights; If the state evaluation function is greater than the preset threshold, the light intensity and electromagnetic interference parameters are adjusted; otherwise, the evaluation effect index is calculated; If the evaluation effect index is less than the preset target value, the stimulation intensity is increased; otherwise, the stimulation intensity is maintained or reduced; Energy allocation optimization is performed based on a multi-objective function to optimize light intensity and energy allocation parameters. The multi-objective function includes minimizing total energy consumption, maximizing abnormal discharge suppression efficiency, and maximizing stability.

10. The multi-target synergistic neural regulation method according to claim 9, characterized in that: The multimodal features include membrane potential features, peak discharge features and bifurcation point features.

Images