Intracranial brain electrical, intracranial pressure, brain temperature monitoring closed-loop neuroregulation system

Abstract

The application discloses a kind of based on intracranial electroencephalogram, intracranial pressure, brain temperature monitoring closed-loop nerve control system, including multimodal integration sEEG probe, multi-source signal synchronous acquisition module, peripheral steady-state monitoring module, central processing and control module and dynamic spatial conformal pulse generation module.Probe synchronous acquisition local field potential, intracranial pressure and brain temperature data reconstruct three-dimensional conductivity tensor matrix.System extracts abnormal electroencephalogram feature to trigger spatial conformal, and solves multi-channel current distribution in combination with Poisson equation.Meanwhile, extract peripheral electrocardiogram features to carry out inner ring parameter slow calibration, and execute phase-locked loop gate control output based on cardiovascular dynamics phase.The application effectively offsets the physical impedance deviation caused by microenvironment dramatic change, realizes high-precision space-time coupling dynamic closed-loop regulation to white matter target.

Description

Technical Field

[0001] This invention relates to the field of neurocritical care and neuromodulation technology, specifically to a closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring. Background Technology

[0002] As a routine intervention method, closed-loop neuromodulation systems mainly monitor the physiological or electrophysiological signals of brain tissue in real time, triggering neural stimulation when specific pathological features appear, thereby inhibiting abnormal neural activity and promoting functional recovery.

[0003] In existing technologies, to obtain multi-dimensional physiological information, it is often necessary to insert different probes into the skull. For example, a dedicated pressure probe is used to measure intracranial pressure, a temperature probe is used to monitor local brain temperature, and independent EEG electrodes are used to collect local field potentials. Existing modulation devices mainly use the collected single EEG signal as a feedback trigger source. When abnormal discharge is detected, an electrical stimulation pulse with fixed parameters is output to a set target area.

[0004] However, existing closed-loop neuromodulation systems suffer from limitations in achieving high-precision spatiotemporal dynamic control when addressing acute brain injury. The use of discrete multi-probe monitoring in existing systems not only increases the risk of puncture-related damage to brain tissue, but also prevents the synchronous acquisition of discrete physical parameters with electrophysiological signals within a unified spatial domain. Furthermore, acute cerebral edema and temperature changes cause dynamic drift in local physical impedance, and existing systems cannot correct for the dielectric properties of the tissue in real time. This makes it easy for the calculated electric field distribution of the preset stimulation parameters to deviate in the abnormal tissue microenvironment. Additionally, conventional single-central feedback mechanisms lack coordinated assessment of the peripheral autonomic nervous system's homeostasis, easily leading to overstimulation and autonomic nervous system disorders. Moreover, the pulse output fails to consider the cerebral blood flow pulsation cycle, and applying stimulation during the brain tissue micro-expansion phase affects the accuracy of local modulation. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring, comprising a multimodal integrated sEEG probe, a multi-source signal synchronous acquisition module, a peripheral steady-state monitoring module, a central processing and control module, and a dynamic spatial shaping pulse generation module.

[0006] The multimodal integrated sEEG probe includes an insulated flexible probe body. The outer surface of the insulated flexible probe body is equipped with a microelectrode array for acquiring local field potential signals and outputting electrical stimulation pulses, a pressure sensing unit for acquiring intracranial pressure data, and a temperature sensing unit for acquiring brain tissue temperature data. The microelectrode array contains multiple independent microelectrode contacts, the pressure sensing unit contains multiple miniature Fabry-Perot pressure sensor targets, and the temperature sensing unit contains multiple thin-film thermocouple nodes. The independent microelectrode contacts, miniature Fabry-Perot pressure sensor targets, and thin-film thermocouple nodes are arranged alternately on the outer surface of the insulated flexible probe body according to a three-dimensional spiral trajectory, forming a topologically interwoven structure.

[0007] Preferably, the multi-source signal synchronous acquisition module establishes a physical signal connection with the multimodal integrated sEEG probe to receive local field potential signals, intracranial pressure data, and brain tissue temperature data. The multi-source signal synchronous acquisition module is internally configured with a global clock unit, a front-end conditioning array, an analog-to-digital converter array, and a data frame encapsulation unit. The front-end conditioning array converts the input physical sensing signals into corresponding analog sequences. The analog-to-digital converter array performs parallel hold and quantization triggered by the rising edge of the system reference clock, generating digital measurement values. The data frame encapsulation unit constructs a synchronous data frame matrix according to a preset discrete time period, adds an absolute timestamp, and continuously outputs it to the central processing and control module.

[0008] Preferably, the central processing and control module is configured with a microenvironment impedance tensor reconstruction submodule, an outer loop parameter calibration submodule, and an inner loop shaping calculation submodule. The peripheral steady-state monitoring module continuously acquires peripheral electrocardiogram signals and sends them to the central processing and control module.

[0009] Preferably, the microenvironment impedance tensor reconstruction submodule performs three-dimensional spline interpolation on discrete intracranial pressure data and brain tissue temperature data to generate continuous spatial pressure gradient fields and continuous spatial temperature gradient fields. The microenvironment impedance tensor reconstruction submodule calculates the three-dimensional conductivity tensor matrix of the tissue surrounding the target site based on a microenvironment dielectric constant compensation model. The microenvironment dielectric constant compensation model sets the real-time three-dimensional conductivity tensor matrix as the product of the baseline conductivity tensor matrix, a temperature compensation scalar factor, and a pressure porosity compensation scalar factor; wherein, the temperature compensation scalar factor is linearly positively correlated with the deviation of the continuous spatial temperature gradient field relative to the baseline body temperature, and the pressure porosity compensation scalar factor is linearly positively correlated with the deviation of the continuous spatial pressure gradient field relative to the baseline intracranial pressure.

[0010] Preferably, the outer loop parameter calibration submodule extracts the RR interval sequence of the cardiac cycle from the peripheral ECG signal. After cubic spline interpolation resampling and discrete Fourier transform processing, it calculates the ratio of the preset low-frequency power value to the high-frequency power value to generate a steady-state characterization coefficient. Based on the steady-state deviation between the steady-state characterization coefficient and the preset autonomic nervous system steady-state target value, the outer loop parameter calibration submodule executes a discrete position proportional-integral control algorithm to calculate the updated inner loop trigger threshold and total stimulation duty cycle. The outer loop parameter calibration submodule extracts the absolute timestamp of the R-wave peak of the cardiac cycle to calculate the average RR interval. Combined with the absolute timestamp of the current R-wave peak, it constructs a time-phase mapping model to convert the continuous system absolute time into the cardiovascular dynamics instantaneous phase. When the cardiovascular dynamics instantaneous phase is within the preset systolic target phase interval, the outer loop parameter calibration submodule controls the output of a high-level time-domain gating signal.

[0011] Preferably, the inner-loop shaping calculation submodule performs discrete short-time Fourier transform processing on the local field potential signal, calculates the ratio of the pathological frequency band energy value to the physiological baseline frequency band energy value, and generates a pathological frequency band relative energy ratio feature. When the pathological frequency band relative energy ratio feature of any channel within a continuous time window exceeds the updated inner-loop trigger threshold and reaches the upper limit of the fault tolerance count, the inner-loop shaping calculation submodule activates the multi-channel current allocation solution operation. The inner-loop shaping calculation submodule transforms the multi-channel current allocation solution operation into a quadratic programming optimization problem with constraints of total charge balance equality and single-channel current safety limit inequality. The objective function of the optimization problem is set to minimize the deviation between the actual spatial potential gradient and the preset white matter target ideal activation electric field spatial distribution model, with an additional energy penalty term for the current vector. Simultaneously, the inner-loop shaping submodule establishes a physical mapping constraint model based on the anisotropic Poisson equation. This model defines that, in each coordinate component of the three-dimensional Cartesian coordinate system, the sum of the partial derivatives of the product of the elements of the three-dimensional conductivity tensor matrix and the partial derivatives of the potential distribution corresponds to the current amplitude distribution allocated to each independent electrode contact. The inner-loop shaping submodule then obtains the optimal multi-channel current output vector that minimizes the objective function value by executing the interior-point method or the effective set method.

[0012] Preferably, the dynamic spatial shaping pulse generation module generates a basic stimulus sequence based on preset pulse width and pulse frequency parameters. The high-speed hardware and gate network configured within the dynamic spatial shaping pulse generation module receives the basic stimulus sequence and the time-domain gating signal, performs pulse state binding processing, and outputs a gated stimulus sequence. Only within the time window when the time-domain gating signal remains high, the dynamic spatial shaping pulse generation module drives the active matrix network according to the optimal multi-channel current output vector to distribute the gated stimulus sequence to the microelectrode array to execute targeted physical electric field output.

[0013] Preferably, during the non-stimulation time interval, the central processing and control module drives the probe's transmitting electrode contacts to release a subthreshold high-frequency sinusoidal carrier wave and simultaneously acquires the voltage response signal. The microenvironment impedance tensor reconstruction submodule calculates the actual transfer impedance array based on the ratio of the complex amplitude of the voltage response signal to the complex amplitude of the injected current parameter. It then compares the actual transfer impedance array with the theoretical transfer impedance array obtained by substituting it into the forward electromagnetic simulation mapping operator to obtain the impedance residual function value. When the impedance residual function value is greater than the calibrated trigger threshold, the microenvironment impedance tensor reconstruction submodule executes an optimization algorithm to extract the corrected parameters and update the microenvironment dielectric constant compensation model.

[0014] The present invention, by adopting the above technical solution, can bring the following beneficial effects:

[0015] 1. This invention constructs a topologically interlaced structure by alternating microelectrodes, pressure sensors, and temperature sensors along a three-dimensional spiral trajectory on the outer surface of an integrated sEEG probe. After a single implantation into the target white matter region, spatially discrete field potential, intracranial pressure, and brain temperature data are simultaneously acquired along the probe's axial and radial directions. This effectively avoids the risk of tissue damage caused by multiple probe implantation and provides a multi-source physical quantity input reference for the subsequent control module, achieving a high-precision simultaneous sensing effect of multimodal signals of the brain microenvironment in the acute phase.

[0016] 2. This invention generates a continuous physical gradient field by performing interpolation operations on intracranial pressure and brain temperature data. Combined with a dielectric constant compensation model, it reconstructs the three-dimensional conductivity tensor matrix of the tissue surrounding the target point in real time. The matrix is ​​then substituted into the Poisson equation to solve for the multi-channel current distribution ratio with charge constraints, adaptively offsetting the tissue impedance shift phenomenon caused by cerebral edema and metabolic abnormalities. This drives the active matrix network to output pulse stimulation with specific parameters to the microelectrode array, achieving a high-precision targeted electric field dynamic spatial shaping effect under complex microenvironment disturbances.

[0017] 3. This invention calculates the steady-state coefficient of heart rate variability by extracting peripheral electrocardiogram features, and updates the inner loop trigger threshold and duty cycle based on its deviation by executing a closed-loop control algorithm. At the same time, it constructs a phase mapping model based on the systolic peak of the cardiac cycle to generate a time-domain gating signal for the gate circuit, so that the stimulation pulse is locked in the specific time window of tissue micro-expansion caused by blood flow pulsation. This overcomes the limitations of a single feedback dimension and realizes the dual regulation effect of cross-domain spatiotemporal coupling between central nervous system stimulation and peripheral autonomic nervous function. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall system of the present invention;

[0019] Figure 2 This is a topological schematic diagram of the multimodal integrated sEEG probe of the present invention;

[0020] Figure 3 This is a flowchart illustrating the overall closed-loop control process of the system of this invention.

[0021] Figure 4 This is the logic diagram for impedance tensor derivation based on multi-source gradient field interpolation in this invention;

[0022] Figure 5 This is a flowchart of the current distribution solution and output control of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please refer to the appendix. Figure 1-5 The present invention provides a closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure and brain temperature monitoring, comprising: a multimodal integrated sEEG probe 100, a multi-source signal synchronous acquisition module 200, a peripheral steady-state monitoring module 300, a central processing and control module 400 and a dynamic spatial shaping pulse generation module 500.

[0025] The multimodal integrated sEEG probe 100 includes an insulating flexible probe body with a cylindrical structure. The outer surface of the insulating flexible probe body is provided with a microelectrode array 110, a pressure sensing unit 120 and a temperature sensing unit 130.

[0026] The microelectrode array 110 contains multiple independent electrode contacts for acquiring local field potential signals and outputting electrical stimulation pulses; the pressure sensing unit 120 contains multiple miniature Fabry pressure sensor targets for acquiring intracranial pressure data; and the temperature sensing unit 130 contains multiple thin-film thermocouple nodes for acquiring brain tissue temperature data.

[0027] Independent microelectrode contacts, miniature Fabry pressure sensor target surfaces, and thin-film thermocouple nodes are alternately arranged on the outer surface of the insulating flexible probe body 101 according to a three-dimensional spiral trajectory, which forms a topological interlaced structure in space.

[0028] Let the central axis of the insulating flexible probe body 101 be the Z-axis of the spatial coordinate system, and let the cross-sectional radius of the insulating flexible probe body 101 be... In a topologically interwoven structure, the three-dimensional spatial coordinate parameters of any sensing node on the surface of the multimodal integrated sEEG probe 100 The following helical parametric equations are satisfied:

[0029]

[0030] in, Represents the coordinate components of any sensing node on the X-axis of the spatial coordinate system; Represents the coordinate components of any sensing node on the Y-axis of the spatial coordinate system; Represents the coordinate components of any sensing node on the Z-axis of the spatial coordinate system; Indicates the first The position parameter of the helix angle where each sensing node is located; The pitch parameter represents the topologically interleaved structure; Represents pi;

[0031] The independent microelectrode contacts in the microelectrode array 110, the miniature Fabry pressure sensor target surface in the pressure sensing unit 120, and the thin film thermocouple node in the temperature sensing unit 130 constitute a ternary node group. On the outer surface of the insulating flexible probe body 101, multiple ternary node groups are arranged sequentially along the Z-axis.

[0032] In the Within each ternary node group, the helix angle position parameters of the independent microelectrode contacts, the micro Fabry-Perot pressure sensor target surface, and the thin-film thermocouple node are respectively defined as... , and , , and The following phase distribution equation is satisfied:

[0033]

[0034] in, Indicates the first The starting reference angle of a ternary node group; This represents the fixed angle difference within a ternary node group;

[0035] The center-to-center distance between the independent microelectrode contacts within the ternary node group and the adjacent microfapper pressure sensor target surface, as well as the center-to-center distance between the microfapper pressure sensor target surface and the adjacent thin-film thermocouple node, are all kept consistent.

[0036] The topologically interleaved structure allows the various sensing nodes of the multimodal integrated sEEG probe 100 to be distributed in a dispersed manner on the insulating flexible probe body 101. The topologically interleaved structure ensures that after the multimodal integrated sEEG probe 100 is implanted in the target white matter fiber bundle region, it can simultaneously acquire spatially discrete electrophysiological signals, intracranial pressure data, and brain tissue temperature data along the axial and radial dimensions of the insulating flexible probe body 101.

[0037] The pressure sensing unit 120 and the temperature sensing unit 130 are topologically interleaved with the microelectrode array 110 along the axial direction of the insulating flexible probe body. The topologically interleaved distribution structure formed by the alternating arrangement of the pressure sensing unit 120, the temperature sensing unit 130 and the microelectrode array 110 enables the multimodal integrated sEEG probe 100 to simultaneously acquire the intracranial pressure gradient and the brain tissue temperature gradient distributed along the axial direction of the insulating flexible probe body.

[0038] The multi-source signal synchronous acquisition module 200 establishes a physical signal connection with the multimodal integrated sEEG probe 100. The multi-source signal synchronous acquisition module 200 is used to receive the local field potential signal acquired by the microelectrode array 110, the intracranial pressure data acquired by the pressure sensing unit 120, and the brain tissue temperature data acquired by the temperature sensing unit 130, and to perform amplification and analog-to-digital conversion processing.

[0039] The multi-source signal synchronous acquisition module 200 internally includes a global clock unit 201, a front-end conditioning array 202, an analog-to-digital converter array 203, and a data frame encapsulation unit 204;

[0040] The front-end conditioning array 202 establishes a physical connection with the multimodal integrated sEEG probe 100. The front-end conditioning array 202 includes a high-throughput electrophysiological interface, an interferometric optical demodulator, and a thermocouple cold junction compensation circuit.

[0041] The peripheral steady-state monitoring module 300 includes a surface ECG electrode patch and is used to continuously acquire the patient's peripheral ECG signals.

[0042] The central processing and control module 400 is communicatively connected to the multi-source signal synchronous acquisition module 200 and the peripheral steady-state monitoring module 300, respectively. The central processing and control module 400 is internally configured with a micro-environment impedance tensor reconstruction submodule 410, an outer loop parameter calibration submodule 420 and an inner loop shaping calculation submodule 430.

[0043] The microenvironment impedance tensor reconstruction submodule 410 is used to receive intracranial pressure data and brain tissue temperature data. Based on the intracranial pressure data and brain tissue temperature data, the microenvironment impedance tensor reconstruction submodule 410 constructs a three-dimensional physical gradient field and calculates the three-dimensional conductivity tensor matrix of the tissue surrounding the target point based on a microenvironment dielectric constant compensation model. The specific formula for the microenvironment dielectric constant compensation model is as follows:

[0044]

[0045] in, Indicates at time Spatial coordinates The real-time three-dimensional conductivity tensor matrix at the location; Represents the reference conductivity tensor matrix; Indicates at time Spatial coordinates A continuous spatial temperature gradient field at a given location; Indicates the baseline body temperature; Indicates the temperature impedance coefficient; Indicates at time Spatial coordinates Continuous spatial pressure gradient field at the location; Indicates baseline intracranial pressure; This represents the transformation coefficient for edema porosity; the above operation is a scalar-tensor multiplication;

[0046] The outer loop parameter calibration submodule 420 is used to receive peripheral electrocardiogram signals. The outer loop parameter calibration submodule 420 extracts cardiovascular dynamic characteristic data based on the peripheral electrocardiogram signals, calculates global autonomic nervous homeostasis indicators, and outputs outer loop calibration weights and time-domain gating signals.

[0047] The inner-loop shaping submodule 430 is used to receive the local field potential signal and the outer-loop calibration weight. The inner-loop shaping submodule 430 extracts pathological frequency band features through time-frequency analysis, and when the triggering condition is met, it performs multi-channel current allocation calculations by combining the three-dimensional conductivity tensor matrix, generating a spatial shaping output vector containing the current amplitude and polarity of each microelectrode contact within the microelectrode array 110. The inner-loop shaping submodule 430 calculates the multi-channel current allocation by solving the Poisson equation, specifically:

[0048]

[0049] in, Represents the spatial distribution of electric potential; This indicates the total number of independent electrode contacts included in the microelectrode array 110; Indicates assignment to the first The current amplitude of each independent electrode contact; Indicates the first The spatial coordinates of an independent electrode contact; Represents the Dirac function;

[0050] The dynamic spatial shaping pulse generation module 500 is connected to the central processing and control module 400 and the multimodal integrated sEEG probe 100. The dynamic spatial shaping pulse generation module 500 includes an active matrix network. The dynamic spatial shaping pulse generation module 500 drives the active matrix network to apply a multi-channel independently regulated electrical stimulation sequence to the microelectrode array 110 based on the spatial shaping output vector generated by the inner loop shaping calculation submodule 430 and the time-domain gating signal output by the outer loop parameter calibration submodule 420.

[0051] Based on the above system architecture and the configuration of each hardware module, this invention implements a dynamic collaborative closed-loop control strategy. Furthermore, this invention provides a workflow for a closed-loop neuromodulation system based on intracranial EEG, intracranial pressure, and brain temperature monitoring, comprising the following steps:

[0052] In step S100, the multi-source signal synchronous acquisition module 200 acquires local field potential signals, intracranial pressure data, and brain tissue temperature data through the multimodal integrated sEEG probe 100, while the peripheral steady-state monitoring module 300 simultaneously acquires peripheral electrocardiogram signals.

[0053] Step S200: The microenvironment impedance tensor reconstruction submodule 410 receives intracranial pressure data and brain tissue temperature data. The microenvironment impedance tensor reconstruction submodule 410 performs spatial spline interpolation on the discrete intracranial pressure data and brain tissue temperature data to generate a continuous three-dimensional physical gradient field. The microenvironment impedance tensor reconstruction submodule 410 combines the microenvironment dielectric constant compensation model to calculate the three-dimensional conductivity tensor matrix of the tissue surrounding the target point.

[0054] Step S300: The outer loop parameter calibration submodule 420 receives the peripheral ECG signal, extracts the RR interval of the peripheral ECG signal, calculates the low-frequency power and high-frequency power, calculates the ratio of low-frequency power to high-frequency power, generates a heart rate variability index, calculates the deviation between the heart rate variability index and the preset steady-state target value, and updates the inner loop trigger threshold based on the deviation. The update formula for the inner loop trigger threshold is:

[0055] in, This represents the inner loop trigger threshold in the m-th discrete sampling period; Indicates the initial trigger threshold; This represents the proportional control gain coefficient; This represents the integral control gain coefficient; This represents the deviation between the heart rate variability index and the preset steady-state target value in the m-th discrete sampling period; Indicates the discrete-time sampling interval; The iterative index variable representing the historical discrete sampling period;

[0056] In step S400, the outer loop parameter calibration submodule 420 analyzes the peripheral electrocardiogram signal, extracts the systolic peak phase in the cardiac cycle, and generates a time-domain gating signal based on the systolic peak phase.

[0057] In step S500, the inner ring shaping calculation submodule 430 receives the local field potential signal, performs time-frequency analysis on the local field potential signal, extracts pathological frequency band feature values, and compares the pathological frequency band feature values ​​with the inner ring trigger threshold. When the pathological frequency band feature value is greater than the inner ring trigger threshold, the inner ring shaping calculation submodule 430 triggers the spatial shaping operation logic.

[0058] In step S600, the inner ring shaping calculation submodule 430 combines the three-dimensional conductivity tensor matrix output by the micro-environment impedance tensor reconstruction submodule 410 to perform multi-channel current distribution calculation by solving the anisotropic Poisson equation. The inner ring shaping calculation submodule 430 generates a spatial shaping output vector containing the amplitude and polarity of the independent electrode contact current inside the microelectrode array 110.

[0059] In step S700, the dynamic spatial shaping pulse generation module 500 receives the spatial shaping output vector generated by the inner loop shaping calculation submodule 430 and the time-domain gating signal output by the outer loop parameter calibration submodule 420. The dynamic spatial shaping pulse generation module 500 generates a basic stimulus sequence according to the preset pulse width and pulse frequency parameters.

[0060] In step S800, the dynamic spatial shaping pulse generation module 500 performs logical operations on the basic stimulus sequence and the time-domain gating signal, and outputs the gated stimulus sequence. The output formula of the gated stimulus sequence is as follows:

[0061] in, Indicates at time The output gated stimulus sequence; Indicates at time The basic stimulus sequence; Indicates at time The cardiovascular dynamics of the systolic peak phase; and These represent the start and end phases of the time-domain gated window, respectively. ,when In The value is 1 if it falls within the interval, and 0 otherwise.

[0062] In step S900, the dynamic spatial shaping pulse generation module 500 drives the internal active matrix network to distribute the gated stimulation sequence to the microelectrode array 110 of the multimodal integrated sEEG probe 100 according to the spatial shaping output vector, and performs physical electric field output on the target point.

[0063] Steps S100 to S900 above constitute the macroscopic process of the closed-loop neuromodulation based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring of the present invention. To fully realize this modulation system, the core processing mechanisms involved in the main process, such as multi-source signal synchronization, impedance reconstruction, parameter calibration, feature extraction, and electric field shaping, will be described in detail below. First, regarding the signal acquisition stage in step S100 above, this embodiment provides a synchronous acquisition mechanism for multi-source physical quantities and electrophysiological signals, including the following steps:

[0064] Step S101: The global clock unit 201 generates the system reference clock tick and distributes the system reference clock tick to the front-end conditioning array 202 and the analog-to-digital converter array 203.

[0065] Step S102: The microelectrode array 110 acquires the local field potential signal in the tissue microenvironment, the pressure sensing unit 120 acquires the optical phase shift signal caused by physical pressure, and the temperature sensing unit 130 acquires the thermodynamic differential voltage signal.

[0066] Step S103: The front-end conditioning array 202 receives the local field potential signal, the optical phase shift signal, and the thermodynamic differential voltage signal. The high-throughput electrophysiological interface in the front-end conditioning array 202 performs impedance matching and amplification on the local field potential signal and outputs an electrophysiological simulation sequence. The interferometric optical demodulator in the front-end conditioning array 202 converts the optical phase shift signal into a pressure simulation sequence. The thermocouple cold junction compensation circuit in the front-end conditioning array 202 converts the thermodynamic differential voltage signal into a temperature simulation sequence.

[0067] In step S104, the analog-to-digital converter array 203 triggers multiple sampling channels in parallel at the rising edge of the system reference clock. The analog-to-digital converter array 203 performs parallel hold and quantization on the electrophysiological simulation sequence, pressure simulation sequence and temperature simulation sequence to generate digital measurement values ​​aligned with the system reference clock.

[0068] Step S105: The data frame encapsulation unit 204 receives the digital measurement values ​​output by the analog-to-digital converter array 203. The data frame encapsulation unit 204 assembles synchronous data frames according to a preset discrete time period. The structure of the synchronous data frame satisfies the following matrix mapping equation:

[0069]

[0070] in, Indicates at discrete sampling time The generated synchronization data frame matrix; Indicates at discrete sampling time The local field potential column vector; Indicates at discrete sampling time Intracranial pressure column vector; Indicates at discrete sampling time The column vector of brain tissue temperature;

[0071] Local field potential column vector The dimension is , The intracranial pressure column vector corresponds to the total number of independent microelectrode contacts contained in the microelectrode array 110. The dimension is , The total number of miniature Fabry-Perot pressure sensor targets contained in the pressure sensing unit 120, and the brain tissue temperature column vector. The dimension is , The total number of thin-film thermocouple nodes included in the corresponding temperature sensing unit 130;

[0072] Step S106, the data frame encapsulation unit 204 encapsulates the current discrete sampling time As an absolute timestamp, it is appended to the synchronization data frame matrix. The frame header;

[0073] Step S107: The multi-source signal synchronous acquisition module 200 will generate a synchronous data frame matrix with absolute timestamps. Continuous output to the central processing and control module 400, synchronous data frame matrix Provides time-aligned physical quantity input data for the calculation of the three-dimensional conductivity tensor matrix.

[0074] After acquiring the synchronization data frame matrix, in order to execute the above step S200, this embodiment provides an impedance tensor passive estimation model based on multi-source gradient field interpolation, the execution steps of which include:

[0075] Step S201: The microenvironment impedance tensor reconstruction submodule 410 receives the synchronous data frame matrix output by the multi-source signal synchronous acquisition module 200. The synchronous data frame matrix contains the intracranial pressure column vector and the brain tissue temperature column vector at discrete sampling times.

[0076] Step S202: The microenvironment impedance tensor reconstruction submodule 410 parses the intracranial pressure column vector and the brain tissue temperature column vector. The intracranial pressure column vector contains digital measurement values ​​of multiple spatially discrete miniature Fabry pressure sensor target surfaces, and the brain tissue temperature column vector contains digital measurement values ​​of multiple spatially discrete thin-film thermocouple nodes.

[0077] Step S203: The microenvironment impedance tensor reconstruction submodule 410 performs three-dimensional spline interpolation on the discretely distributed digital measurements. The microenvironment impedance tensor reconstruction submodule 410 calculates and generates continuous spatial temperature gradient fields and continuous spatial pressure gradient fields. The basic mathematical function expression for the three-dimensional spline interpolation operation is:

[0078]

[0079] in, Indicates at time spatial coordinate vector The values ​​of the physical field in continuous space at a given location correspond to either the temperature gradient field or the pressure gradient field in continuous space, respectively. This represents the total number of sensor nodes participating in the interpolation operation; Indicates at time No. The weighting coefficients of each sensor node; Indicates the first The physical space coordinates of each sensor node; Represents radial basis functions. Spatial coordinate vector With the Physical space coordinates of each sensor node The Euclidean distance between them; Representing spatial coordinate vectors respectively Coordinate components in three orthogonal directions; Indicates at time The coefficient of the linear trend term; Indicates at time The polynomial basis constants, the microenvironment impedance tensor reconstruction submodule 410 constructs and solves a system of linear algebraic equations based on the digital measurements of multiple sensing nodes, and determines the weighting coefficients and polynomial basis coefficients.

[0080] Step S204: The microenvironment impedance tensor reconstruction submodule 410 retrieves the patient's preoperative anisotropic diffusion tensor data from the preset medical image storage unit. The microenvironment impedance tensor reconstruction submodule 410 constructs a reference conductivity tensor matrix based on the anisotropic diffusion tensor data. The reference conductivity tensor matrix is... symmetric matrix;

[0081] Step S205: The microenvironment impedance tensor reconstruction submodule 410 extracts the temperature value in the continuous spatial temperature gradient field and calculates the temperature compensation scalar factor; the microenvironment impedance tensor reconstruction submodule 410 extracts the pressure value in the continuous spatial pressure gradient field and calculates the pressure porosity compensation scalar factor.

[0082] Step S206: The microenvironment impedance tensor reconstruction submodule 410 performs matrix multiplication and nonlinear modulation operations on the reference conductivity tensor matrix, temperature compensation scalar factor, and pressure porosity compensation scalar factor to calculate the real-time three-dimensional conductivity tensor matrix. The specific formula for the nonlinear modulation operation is as follows:

[0083]

[0084] in, Indicates at time spatial coordinate vector The real-time three-dimensional conductivity tensor matrix at the location; Represents the reference conductivity tensor matrix; Indicates at time spatial coordinate vector A continuous spatial temperature gradient field at a given location; Indicates the baseline body temperature; This represents the temperature impedance coefficient, and is the calculation parameter for the corresponding temperature compensation scalar factor; Indicates at time spatial coordinate vector Continuous spatial pressure gradient field at the location; Indicates baseline intracranial pressure; This represents the edema porosity transformation coefficient, corresponding to the calculation parameters of the pressure porosity compensation scalar factor; this nonlinear modulation operation corrects the overall amplitude of conductivity through the scalar factor, and the anisotropic structural characteristics of local tissues are determined by the reference conductivity tensor matrix. Keep it fixed;

[0085] Step S207: The microenvironment impedance tensor reconstruction submodule 410 outputs the real-time three-dimensional conductivity tensor matrix to the inner loop shaping calculation submodule 430. The real-time three-dimensional conductivity tensor matrix serves as the input parameter for the inner loop shaping calculation submodule 430 to perform multi-channel current distribution calculation.

[0086] Furthermore, to ensure the long-term accuracy of the microenvironment dielectric constant compensation model, this embodiment also provides an active impedance detection and adaptive calibration mechanism based on subthreshold high-frequency carriers, including the following steps:

[0087] Step S208: The central processing and control module 400 monitors the working status of the inner loop shaping calculation submodule 430, extracts the time gap when no electrical stimulation pulse is output, and generates an active detection command during the non-stimulation time gap.

[0088] Step S209: The dynamic spatial shaping pulse generation module 500 receives an active detection command and drives the preset transmitting electrode contacts in the microelectrode array 110 of the multimodal integrated sEEG probe 100 to release a subthreshold high-frequency sinusoidal carrier wave to the tissue surrounding the target point. The current amplitude of the subthreshold high-frequency sinusoidal carrier wave is lower than the activation threshold of the neuronal action potential, and the frequency range is set in a preset high-frequency band (e.g., 1kHz-10kHz range, to avoid the main frequency band of effective physiological EEG signals).

[0089] Step S210: The multi-source signal synchronous acquisition module 200 synchronously acquires the voltage response signal corresponding to the subthreshold high-frequency sinusoidal carrier through the preset receiving electrode contacts in the microelectrode array 110, and transmits the voltage response signal to the microenvironment impedance tensor reconstruction submodule 410.

[0090] Step S211: The micro-environment impedance tensor reconstruction submodule 410 extracts the voltage response signal and the injection current parameters of the subthreshold high-frequency sinusoidal carrier. The micro-environment impedance tensor reconstruction submodule 410 extracts the complex amplitude at the carrier frequency through discrete Fourier transform or lock-in amplification technology, and divides the complex amplitude of the voltage response signal by the complex amplitude of the injection current parameters to calculate the actual transfer impedance array. The calculation formula for the matrix elements in the actual transfer impedance array is as follows:

[0091]

[0092] in, Indicates at carrier frequency The actual transfer impedance matrix elements are injected from the i-th transmitting electrode contact and measured at the j-th receiving electrode contact; This indicates that the j-th receiving electrode contact extracts the signal at the carrier frequency. The complex amplitude of the voltage response signal at the location; This indicates the injected current corresponding to the i-th transmitting electrode contact at the carrier frequency. Complex amplitude at the location;

[0093] Step S212: The micro-environment impedance tensor reconstruction submodule 410 obtains the real-time three-dimensional conductivity tensor matrix generated by multi-source gradient field interpolation. The micro-environment impedance tensor reconstruction submodule 410 substitutes the real-time three-dimensional conductivity tensor matrix into the positive electromagnetic simulation mapping operator constructed based on the quasi-steady-state Laplace equation to calculate the theoretical transfer impedance array.

[0094] Step S213: The microenvironment impedance tensor reconstruction submodule 410 compares the actual transferred impedance array with the theoretical transferred impedance array and calculates the impedance residual function value. The formula for calculating the impedance residual function is as follows:

[0095] in, Indicates at time The impedance residual function value; Indicates at time The calculated actual transfer impedance array; Indicates at time spatial coordinate vector The real-time three-dimensional conductivity tensor matrix is ​​generated by physical gradient field calculation. A positive electromagnetic simulation mapping operator representing the physical spatial structure of the multimodal integrated sEEG probe 100 and the surrounding tissue. Represents the square operation of the 2-norm of a matrix;

[0096] Step S214: The microenvironment impedance tensor reconstruction submodule 410 determines whether the impedance residual function value is greater than the preset calibration trigger threshold. When the impedance residual function value is greater than the calibration trigger threshold, the microenvironment impedance tensor reconstruction submodule 410 executes the Levenberg-Marquardt optimization algorithm to minimize the impedance residual function value as the convergence target.

[0097] Step S215: The microenvironment impedance tensor reconstruction submodule 410 extracts the corrected temperature impedance coefficient and edema porosity transformation coefficient based on the iterative convergence result of the Levenberg-Marquardt optimization algorithm. The microenvironment impedance tensor reconstruction submodule 410 uses the corrected temperature impedance coefficient and edema porosity transformation coefficient to update the corresponding parameters in the microenvironment dielectric constant compensation model. The updated parameters are applied to the calculation of the real-time three-dimensional conductivity tensor matrix at subsequent sampling times.

[0098] Regarding the outer loop parameter calibration step in step S300 above, this embodiment provides a slow calibration strategy for inner loop parameters based on peripheral autonomic nervous system homeostasis indices, including the following steps:

[0099] Step S301: The peripheral steady-state monitoring module 300 continuously acquires the patient's peripheral electrocardiogram (ECG) signals and sends the peripheral ECG signals to the outer loop parameter calibration submodule 420 inside the central processing and control module 400.

[0100] Step S302: The outer loop parameter calibration submodule 420 performs long-time-base waveform analysis on the peripheral electrocardiogram signal to extract the continuous cardiac cycle RR interval sequence. The outer loop parameter calibration submodule 420 performs cubic spline interpolation resampling processing on the RR interval sequence to generate a heart rate variability discrete time sequence with equal time intervals. The outer loop parameter calibration submodule 420 performs discrete Fourier transform processing on the heart rate variability discrete time sequence to calculate the power spectral density of the heart rate variability signal.

[0101] Step S303: The outer loop parameter calibration submodule 420 performs discrete frequency domain summation on the power spectral density within a preset low-frequency range to obtain the low-frequency power value. The outer loop parameter calibration submodule 420 performs discrete frequency domain summation on the power spectral density within a preset high-frequency range to obtain the high-frequency power value.

[0102] Step S304: The outer loop parameter calibration submodule 420 divides the low-frequency power value by the high-frequency power value to calculate and generate the steady-state characterization coefficient. The formula for calculating the steady-state characterization coefficient is as follows:

[0103]

[0104] in, Indicates the first Steady-state characterization coefficients for each slow calibration cycle; Indicates the first Low-frequency power value for each slow calibration cycle; Indicates the first High-frequency power value for each slow calibration cycle;

[0105] Step S305, outer loop parameter calibration submodule 420 calculates the steady-state deviation between the steady-state characterization coefficient and the preset heart rate variability reference ratio. The formula for calculating the steady-state deviation is:

[0106]

[0107] in, Indicates the first Steady-state deviation over one slow calibration cycle; This indicates the preset heart rate variability reference ratio;

[0108] Step S306: The outer loop parameter calibration submodule 420, based on the steady-state deviation, executes the discrete position proportional-integral control algorithm to calculate the updated inner loop trigger threshold. The formula for calculating the inner loop trigger threshold is:

[0109]

[0110] in, Indicates the first The inner loop trigger threshold is updated after each slow calibration cycle; This represents the initial trigger threshold of the inner loop shaping submodule 430; This represents the proportional control gain coefficient relative to the trigger threshold; This represents the integral control gain coefficient with respect to the trigger threshold; This represents the discrete time interval of the slow calibration cycle; Discrete iterative index variable representing the historical slow calibration period;

[0111] Step S307: The outer loop parameter calibration submodule 420, based on the steady-state deviation, executes a discrete position proportional-integral control algorithm to calculate the updated total stimulus duty cycle. The formula for calculating the total stimulus duty cycle is:

[0112]

[0113] in, Indicates the first The total duty cycle of the stimulus is updated after a slow calibration cycle; This indicates the initial total duty cycle of the dynamic spatial shaping pulse generation module 500; This represents the proportional control gain coefficient for the duty cycle; This represents the integral control gain coefficient with respect to the duty cycle. This represents the discrete time interval of the slow calibration cycle;

[0114] Step S308: The outer loop parameter calibration submodule 420 determines whether the updated total duty cycle of the stimulus is within the preset safe duty cycle range, and at the same time determines whether the updated inner loop trigger threshold is within the preset safe threshold range. When the updated total duty cycle of the stimulus or the inner loop trigger threshold exceeds its corresponding preset safe range, the outer loop parameter calibration submodule 420 performs anti-windup processing, truncates its value to the nearest safe range boundary value, and simultaneously freezes the integral term accumulation operation of the corresponding parameter in the background, stops the accumulation of steady-state deviation, until the steady-state deviation reverses and the output value falls back to the safe range, thereby preventing response stagnation caused by historical errors.

[0115] Step S309: The outer loop parameter calibration submodule 420 sends the updated inner loop trigger threshold to the inner loop shaping calculation submodule 430, and the outer loop parameter calibration submodule 420 sends the updated total stimulus duty cycle to the dynamic spatial shaping pulse generation module 500.

[0116] In step S310, the inner loop shaping operator module 430 uses the updated inner loop trigger threshold to execute the trigger condition comparison logic of the pathological frequency band characteristics of the local field potential signal. The dynamic spatial shaping pulse generation module 500 limits the effective output time ratio of the multi-channel electrical stimulation sequence within a control cycle according to the updated total duty cycle of the stimulation.

[0117] Corresponding to the feature comparison and trigger determination in step S500 above, this embodiment provides an extraction logic for abnormal sEEG features and a logic for determining trigger conditions, including the following steps:

[0118] Step S501: The inner loop shaping submodule 430 receives the synchronization data frame matrix output by the multi-source signal synchronous acquisition module 200, and the inner loop shaping submodule 430 extracts the local field potential column vector in the synchronization data frame matrix.

[0119] Step S502: The local field potential column vector contains the electrophysiological digital measurement values ​​measured at discrete sampling times by each independent electrode contact of the microelectrode array 110. The inner loop shaping calculation submodule 430 performs bandpass filtering and power frequency notch processing on the electrophysiological digital measurement values ​​corresponding to each independent electrode contact to obtain the local field potential discrete time series.

[0120] Step S503: The inner loop shaping submodule 430 performs Discrete Short-Time Fourier Transform (DSFT) processing on the local field potential discrete-time sequence. The DFT processing converts the time-domain represented electrophysiological signal into a time-frequency domain represented spectrum matrix. The specific calculation formula for the DFT processing is as follows:

[0121]

[0122] in, Indicates the index in the time-shifted window. With discrete frequency index Complex spectrum value at; Indicates the time point after translation. The discrete-time series value of the local field potential at the location; Indicates length is Discrete data window functions; This represents the total number of sampling points included in the discrete data window function; The base of the natural logarithm; Represents the imaginary unit; Represents pi;

[0123] Step S504: The inner ring shaping submodule 430 calculates the power spectral density within a specific frequency range based on the time-frequency domain representation spectral matrix. The inner ring shaping submodule 430 performs discrete frequency domain summation on the preset pathological frequency band to obtain the energy value of the pathological frequency band. The inner ring shaping submodule 430 performs discrete frequency domain summation on the preset physiological baseline frequency band to obtain the energy value of the physiological baseline frequency band.

[0124] Step S505: The inner ring shaping calculation submodule 430 divides the energy value of the pathological frequency band by the energy value of the physiological baseline frequency band to calculate and generate the relative energy ratio feature of the pathological frequency band. The calculation formula for the relative energy ratio feature of the pathological frequency band is:

[0125]

[0126] in, Indicates the index in the time-shifted window. Characteristics of the relative energy ratio of the pathological frequency band at the location; and These represent the lower limit and upper limit of the discrete frequency index for the preset pathological frequency band, respectively. and These represent the lower limit and upper limit of the discrete frequency index of the preset physiological baseline frequency band, respectively; Indicates the index in the time-shifted window. With discrete frequency index The power spectral density value at that location;

[0127] Step S506: The inner ring shaping calculation submodule 430 obtains the inner ring trigger threshold updated and output by the outer ring parameter calibration submodule 420. The inner ring shaping calculation submodule 430 compares the relative energy ratio characteristics of the pathological frequency band of each independent electrode contact with the inner ring trigger threshold.

[0128] Step S507: The inner loop shaping calculation submodule 430 counts the number of consecutive time windows in which the relative energy ratio of the pathological frequency band is greater than the inner loop trigger threshold. When the number of time windows in which the relative energy ratio of the pathological frequency band of any channel is greater than the inner loop trigger threshold reaches the preset fault tolerance count limit, the inner loop shaping calculation submodule 430 determines that the closed loop control trigger condition is met.

[0129] Step S508: After the closed-loop control trigger condition is met, the inner loop shaping calculation submodule 430 generates a spatial shaping trigger command, which activates the multi-channel current distribution solution operation based on the three-dimensional conductivity tensor matrix.

[0130] Step S509: When the relative energy ratio of the pathological frequency bands of all channels is less than or equal to the inner loop trigger threshold, or the number of time windows that are continuously greater than the inner loop trigger threshold has not reached the upper limit of the fault tolerance count, the inner loop shaping calculation submodule 430 determines that the closed loop control trigger condition is not met, the inner loop shaping calculation submodule 430 clears the feature statistics cache, and continues to perform the feature extraction operation of the local field potential discrete time sequence at the next discrete sampling time.

[0131] Once the closed-loop control trigger condition is met, in order to achieve spatial shaping in steps S600 to S900, this embodiment provides a current distribution solution and output mechanism based on the reconstructed conductivity tensor, including the following steps:

[0132] Step S601: The inner loop shaping calculation submodule 430 receives the spatial shaping trigger command and obtains the real-time three-dimensional conductivity tensor matrix output by the micro-environment impedance tensor reconstruction submodule 410.

[0133] Step S602: The inner ring shaping calculation submodule 430 extracts the white matter fiber bundle orientation of the target point based on preoperative diffusion tensor imaging, calculates the projection angle between the preset spatial electric field gradient vector and the white matter fiber bundle orientation, and combines the spatial threshold decay distribution equation based on the neuronal axon activation mechanism to map and construct the preset ideal activation electric field spatial distribution model of the white matter target point. The inner ring shaping calculation submodule 430 transforms the multi-channel current distribution solution process into a quadratic programming optimization problem with physical partial differential equation constraints, total charge balance constraints and safety limit threshold constraints.

[0134] Step S603, Inner Loop Shaping Calculation Submodule 430: Establishes a physical mapping constraint model connecting the multi-channel current output vector and the spatial potential distribution. The physical mapping constraint model is based on the anisotropic Poisson equation, and the specific formula is as follows: in, and These represent the coordinate axis components of a three-dimensional rectangular coordinate system; Indicates at time spatial coordinate vector Elements of the real-time three-dimensional conductivity tensor matrix at the specified location; Indicates at time spatial coordinate vector Electric potential distribution at the location; This indicates the total number of independent electrode contacts included in the microelectrode array 110; and defines... The number is even to ensure complete charge conservation and compatibility of the solution matrix under the multipole electric field configuration; Indicates at time Assigned to the The current amplitude of each independent electrode contact; Indicates the first The spatial physical coordinates of an independent electrode contact; Represents the Dirac function;

[0135] Step S604: The inner loop shaping operator submodule 430 establishes the objective function for the quadratic programming optimization problem. The formula for calculating the objective function is as follows:

[0136]

[0137] in, Indicates at time The objective function value for electric field shaping, This represents the actual spatial electric field distribution generated by the distributed current at time t, and satisfies... ; This represents the spatial region of the target white matter fiber bundle; Represents spatial coordinate vectors A spatial distribution model of the ideal activation electric field at the white matter target site; This represents the energy penalty weighting coefficient; Indicates at time Multi-channel current output vector ; Represents the square operation of the 2-norm of a vector;

[0138] Step S605: The inner loop shaping calculation submodule 430 sets the constraints for multi-channel current distribution. The constraints include equality constraints and inequality constraints, and the specific formulas are as follows:

[0139]

[0140] in, The safe limit threshold for the amplitude of a single-channel current is represented by an equality constraint used to limit the total charge injected into the tissue to maintain balance.

[0141] Step S606: The inner loop shaping operator submodule 430 performs the interior point method or effective set method to solve the objective function. The inner loop shaping operator submodule 430 calculates the optimal multi-channel current output vector that minimizes the objective function value. The optimal multi-channel current output vector includes the current amplitude and polarity distribution ratio of each independent electrode contact inside the microelectrode array 110.

[0142] Step S607: The inner loop shaping operator module 430 sends the optimal multi-channel current output vector to the dynamic space shaping pulse generation module 500.

[0143] Step S608: The dynamic space shaping pulse generation module 500 adjusts the output parameters of the numerically controlled constant current source of the corresponding channel in the internal active matrix network according to the optimal multi-channel current output vector.

[0144] Step S609: The dynamic spatial shaping pulse generation module 500 drives the microelectrode array 110 of the multimodal integrated sEEG probe 100 to synchronously output a multi-channel electrical stimulation sequence in the spatial region of the target white matter fiber bundle. The dynamic spatial shaping pulse generation module 500 uses the principle of spatial electric field superposition to change the three-dimensional electric field geometric distribution of the target area.

[0145] Finally, to achieve precise time-domain matching between step S400 and the final electric field output, this embodiment provides a microsecond-level phase-locked gate mechanism for electric field output based on cardiovascular dynamics phase, including the following steps:

[0146] Step S401: The peripheral steady-state monitoring module 300 continuously collects peripheral electrocardiogram signals and sends the peripheral electrocardiogram signals to the outer loop parameter calibration submodule 420 inside the central processing and control module 400.

[0147] Step S402: The outer loop parameter calibration submodule 420 performs morphological filtering and peak detection operations on the peripheral electrocardiogram signal. The outer loop parameter calibration submodule 420 extracts the R wave peak in the cardiac cycle and records the absolute timestamps of the current R wave peak and the historical R wave peak.

[0148] Step S403: The outer loop parameter calibration submodule 420 calculates the average RR interval based on the absolute timestamps of historical R-wave peaks. Combining the absolute timestamp of the current R-wave peak with the average RR interval, the outer loop parameter calibration submodule 420 constructs a time-phase mapping model. The outer loop parameter calibration submodule 420 converts the continuous absolute system time into a cardiovascular dynamic instantaneous phase with periodic characteristics. The formula for calculating the cardiovascular dynamic instantaneous phase mapping is:

[0149]

[0150] To prevent calculated phase jumps caused by heart rate variability at the boundary between adjacent cardiac cycles, the outer loop parameter calibration submodule 420, upon receiving a new absolute timestamp of the R-wave peak, regardless of the currently calculated... Regardless of the value, it is forcibly reset to zero, and a new absolute timestamp is used as the reference. Restart the mapping calculation for the next cycle;

[0151] in, Indicates the absolute time of the current system The instantaneous phase of cardiovascular dynamics at the location; Indicates the most recent (the) The absolute timestamp of the R-wave peak occurrence; This represents the average RR interval calculated based on the absolute timestamps of historical R-wave peaks; Represents pi;

[0152] Step S404: The outer ring parameter calibration submodule 420 sets the cardiovascular dynamics target phase interval. The target phase interval is defined by the preset start phase value and end phase value according to the stable window period of cerebral blood flow perfusion.

[0153] Step S405: The outer loop parameter calibration submodule 420 generates a time-domain gating signal. The outer loop parameter calibration submodule 420 limits the level state of the time-domain gating signal through logic judgment. When the instantaneous phase of cardiovascular dynamics is within the target phase interval, the outer loop parameter calibration submodule 420 controls the time-domain gating signal to output a high level. When the instantaneous phase of cardiovascular dynamics is outside the target phase interval, the outer loop parameter calibration submodule 420 controls the time-domain gating signal to output a low level.

[0154] Step S406: The dynamic spatial shaping pulse generation module 500 receives the time-domain gating signal output by the outer loop parameter calibration submodule 420, and the dynamic spatial shaping pulse generation module 500 simultaneously receives the basic stimulus sequence and the optimal multi-channel current output vector generated by the inner loop shaping calculation submodule 430.

[0155] Step S407: The dynamic spatial shaping pulse generation module 500 is internally configured with a high-speed hardware and gate circuit network. The dynamic spatial shaping pulse generation module 500 connects the basic stimulus sequence to the first input terminal of the high-speed hardware and gate circuit network, connects the time-domain gating signal to the second input terminal of the high-speed hardware and gate circuit network, and performs pulse state binding processing.

[0156] In step S408, the high-speed hardware and gate circuit network restricts the physical output channel state of the dynamic spatial shaping pulse generation module 500. The dynamic spatial shaping pulse generation module 500 only drives the microelectrode array 110 of the multimodal integrated sEEG probe 100 to release the modulation electric field to the target point according to the optimal multi-channel current output vector within the time window during which the time-domain gate signal is maintained at a high level. The dynamic spatial shaping pulse generation module 500 locks the electric field output within the preset time window of the cardiovascular dynamics phase.

Claims

1. A closed-loop neural modulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring, characterized in that, include: Multimodal integrated sEEG probe (100), multi-source signal synchronous acquisition module (200), peripheral steady-state monitoring module (300), central processing and control module (400) and dynamic spatial shaping pulse generation module (500). The multimodal integrated sEEG probe (100) is used to acquire local field potential signals, intracranial pressure data, and brain tissue temperature data. The multi-source signal synchronous acquisition module (200) establishes a physical signal connection with the multimodal integrated sEEG probe (100) to receive and synchronously process the local field potential signal, the intracranial pressure data, and the brain tissue temperature data. The peripheral steady-state monitoring module (300) is used to continuously acquire peripheral electrocardiogram signals; The central processing and control module (400) is communicatively connected to the multi-source signal synchronous acquisition module (200) and the peripheral steady-state monitoring module (300). The central processing and control module (400) is internally configured with a micro-environment impedance tensor reconstruction submodule (410), an outer loop parameter calibration submodule (420), and an inner loop shaping calculation submodule (430). The microenvironment impedance tensor reconstruction submodule (410) calculates the three-dimensional conductivity tensor matrix of the tissue surrounding the target point based on the intracranial pressure data and the brain tissue temperature data; the outer loop parameter calibration submodule (420) extracts cardiovascular dynamics feature data based on the peripheral electrocardiogram signal, calculates and outputs the inner loop trigger threshold and time-domain gating signal; The inner ring shaping calculation submodule (430) extracts pathological frequency band features based on the local field potential signal, and when the pathological frequency band features are greater than the inner ring trigger threshold, it combines the three-dimensional conductivity tensor matrix to perform multi-channel current allocation calculation to generate a spatial shaping output vector. The dynamic spatial shaping pulse generation module (500) is connected to the central processing and control module (400) and the multimodal integrated sEEG probe (100) respectively, and is used to drive the multimodal integrated sEEG probe (100) to perform physical electric field output on the target point according to the spatial shaping output vector and the time-domain gating signal.

2. The closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring according to claim 1, characterized in that, The multimodal integrated sEEG probe (100) includes: An insulating flexible probe body, wherein a microelectrode array (110), a pressure sensing unit (120) and a temperature sensing unit (130) are disposed on the outer surface of the insulating flexible probe body. The microelectrode array (110) includes multiple independent microelectrode contacts, the pressure sensing unit (120) includes multiple miniature Fabry pressure sensor targets, and the temperature sensing unit (130) includes multiple thin-film thermocouple nodes. The independent microelectrode contacts, the miniature Fabry-Perot pressure sensor target surface, and the thin-film thermocouple nodes are arranged alternately on the outer surface of the insulating flexible probe body according to a three-dimensional spiral trajectory, forming a topologically interwoven structure.

3. The closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring according to claim 1, characterized in that, The multi-source signal synchronous acquisition module (200) internally includes: Global clock unit (201), front-end conditioning array (202), analog-to-digital converter array (203), and data frame encapsulation unit (204). The front-end conditioning array (202) converts the received local field potential signal, the optical phase shift signal corresponding to the intracranial pressure data, and the thermodynamic differential voltage signal corresponding to the brain tissue temperature data into analog sequences. The analog-to-digital converter array (203) performs parallel hold and quantization on the analog sequence at the rising edge of the system reference clock generated by the global clock unit (201) to generate digital measurement values; The data frame encapsulation unit (204) constructs a synchronous data frame matrix according to a preset discrete time period, and appends the discrete sampling time as an absolute timestamp to the frame header of the synchronous data frame matrix, and continuously outputs it to the central processing and control module (400).

4. The closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring according to claim 1, characterized in that, The microenvironment impedance tensor reconstruction submodule (410) performs three-dimensional spatial spline interpolation on the discrete intracranial pressure data and brain tissue temperature data to generate a continuous spatial pressure gradient field and a continuous spatial temperature gradient field. The microenvironment impedance tensor reconstruction submodule (410) calculates the three-dimensional conductivity tensor matrix based on the microenvironment dielectric constant compensation model. The specific formula of the microenvironment dielectric constant compensation model is as follows: in, Indicates at time spatial coordinate vector The real-time three-dimensional conductivity tensor matrix at the location; Represents the reference conductivity tensor matrix; Indicates at time spatial coordinate vector A continuous spatial temperature gradient field at a given location; Indicates the baseline body temperature; Indicates the temperature impedance coefficient; Indicates at time spatial coordinate vector Continuous spatial pressure gradient field at the location; Indicates baseline intracranial pressure; The scalar factor represents the transformation coefficient of edema porosity. In the above formula, the scalar factor is scaled proportionally only to the overall amplitude of the reference conductivity tensor matrix to keep the anisotropic structural characteristics of the local tissue unchanged.

5. The closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring according to claim 1, characterized in that, The outer loop parameter calibration submodule (420) extracts the cardiac cycle RR interval sequence of the peripheral electrocardiogram signal, performs cubic spline interpolation resampling and discrete Fourier transform processing, calculates the ratio of the preset low-frequency power value to the high-frequency power value, and generates a steady-state characterization coefficient; the outer loop parameter calibration submodule (420) calculates the steady-state deviation between the steady-state characterization coefficient and the preset autonomic nervous system steady-state target value; The outer loop parameter calibration submodule (420) executes a discrete position proportional-integral control algorithm based on the steady-state deviation, calculates the updated inner loop trigger threshold and sends it to the inner loop shaping calculation submodule (430), and at the same time calculates the updated total stimulus duty cycle and sends it to the dynamic spatial shaping pulse generation module (500).

6. The closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring according to claim 1, characterized in that, The outer loop parameter calibration submodule (420) extracts the absolute timestamp of the R wave peak of the peripheral electrocardiogram signal's central cardiac cycle and calculates the average RR interval; The outer ring parameter calibration submodule (420) combines the absolute timestamp of the current R wave peak with the average RR interval to construct a time-phase mapping model, converting the continuous system absolute time into the cardiovascular dynamics instantaneous phase; When the instantaneous phase of the cardiovascular dynamics is within the preset systolic target phase interval, the outer loop parameter calibration submodule (420) controls the time-domain gating signal to output a high level; When the instantaneous phase of the cardiovascular dynamics is outside the target phase interval of the systolic phase, the time-domain gating signal is controlled to output a low level.

7. The closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring according to claim 1, characterized in that, The inner ring shaping submodule (430) performs discrete short-time Fourier transform processing on the local field potential signal to generate a time-frequency domain representation spectrum matrix; The inner ring shaping operator submodule (430) calculates the ratio of the energy value of the pathological frequency band to the energy value of the physiological baseline frequency band based on the time-frequency domain representation spectrum matrix, and generates the relative energy ratio feature of the pathological frequency band. When the number of time windows in which the relative energy ratio of the pathological frequency band of any channel is continuously greater than the inner loop trigger threshold reaches the preset fault tolerance count limit, the inner loop shaping calculation submodule (430) determines that the closed loop control trigger condition is met and generates a spatial shaping trigger command to activate the multi-channel current distribution solution operation.

8. The closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring according to claim 1, characterized in that, The inner loop shaping submodule (430) transforms the multi-channel current distribution solution into a quadratic programming optimization problem with physical partial differential equation constraints, total charge balance constraints and safety limit threshold constraints. The physical partial differential equation constraint establishes a physical mapping constraint model connecting the multi-channel current output vector and the spatial potential distribution. The specific formula is as follows: in, and These represent the coordinate axis components of a three-dimensional rectangular coordinate system; Indicates at time spatial coordinate vector The elements of the three-dimensional conductivity tensor matrix at that location; Indicates at time spatial coordinate vector Electric potential distribution at the location; This indicates the total number of independent electrode contacts; Indicates at time Assigned to the The current amplitude of each independent electrode contact; Indicates the first The spatial physical coordinates of an independent electrode contact; Represents the Dirac function; The inner loop shaping operator module (430) performs the interior point method or the effective set method to solve the objective function and obtain the optimal multi-channel current output vector that minimizes the objective function value, which is then used as the space shaping output vector.

9. The closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring according to claim 1, characterized in that, During the non-stimulation time interval, the central processing and control module (400) drives the multimodal integrated sEEG probe (100) to release subthreshold high-frequency sinusoidal carrier waves to the tissue around the target point. The multi-source signal synchronous acquisition module (200) synchronously acquires the voltage response signal corresponding to the subthreshold high-frequency sinusoidal carrier and transmits it to the micro-environment impedance tensor reconstruction submodule (410); the micro-environment impedance tensor reconstruction submodule (410) calculates the actual transfer impedance array by dividing the complex amplitude of the voltage response signal by the complex amplitude of the injected current parameter, and compares it with the theoretical transfer impedance array calculated based on the forward electromagnetic simulation mapping operator to obtain the impedance residual function value; When the impedance residual function value is greater than the calibrated trigger threshold, an optimization algorithm is executed to extract the corrected parameters to update the microenvironment dielectric constant compensation model.

10. The closed-loop neuromodulation system based on intracranial electroencephalography, intracranial pressure, and brain temperature monitoring according to claim 1, characterized in that, The dynamic spatial shaping pulse generation module (500) is internally configured with high-speed hardware and gate circuit network; The dynamic spatial shaping pulse generation module (500) generates a basic stimulus sequence according to preset pulse width and pulse frequency parameters; the dynamic spatial shaping pulse generation module (500) connects the basic stimulus sequence and the time-domain gating signal to the high-speed hardware and gate circuit network to perform pulse state binding processing and outputs a gated stimulus sequence; The dynamic spatial shaping pulse generation module (500) only distributes the gated stimulation sequence to the microelectrode array (110) of the multimodal integrated sEEG probe (100) to perform physical electric field output within the time window during which the time-domain gated signal is maintained at a high level, based on the active matrix network driven by the spatial shaping output vector.

Images