Self-adaptive awakening system and method for disturbance of consciousness based on traditional Chinese medicine thumbtack needle fusion nerve regulation

Abstract

The invention provides a disturbance of consciousness self-adaptive awakening system and method based on traditional Chinese medicine thumbtack needle fusion nerve regulation and control, and belongs to the technical field of nerve regulation and control and rehabilitation engineering. The system comprises a central control and processing unit, a stimulation subsystem, an acquisition subsystem, a magnetic artifact suppression module and a traditional Chinese medicine state monitoring module; the awareness network state of a patient is evaluated in real time in combination with qEEG and fNIRS signals and traditional Chinese medicine acupoint physiological data, stimulation parameters are optimized through a self-adaptive parameter engine, personalized treatment is achieved, and the system adopts a double-station phase locking collaborative stimulation strategy and is combined with a magnetic artifact suppression technology to ensure data quality; the method comprises the steps of initialized network topology identification, execution of double-station stimulation, artifact suppression, network state inversion, data judgment and staged optimization, and awareness and consciousness integration of a patient is promoted.

Description

Technical Field

[0001] This invention belongs to the field of neuromodulation and rehabilitation engineering technology, specifically relating to an adaptive awakening system and method for consciousness disorders that integrates traditional Chinese medicine acupuncture with neuromodulation. Background Technology

[0002] Existing repetitive transcranial magnetic stimulation (rTMS) techniques for promoting arousal in patients with chronic disorders of consciousness (pDoC) typically employ an open-loop treatment model. For example, patent publication CN112451829A discloses an ultrasound-based arousal system that works by first conducting a pre-treatment assessment, then target area treatment, and finally a post-treatment assessment. This sequential, offline model cannot dynamically adjust the treatment plan based on real-time changes in the patient's brain state during treatment. It is precisely this separate assessment and treatment approach that leads to a lack of individualized treatment plans and unstable efficacy.

[0003] Another key challenge lies in the problem of "measurability during stimulation." rTMS generates strong transient electromagnetic artifacts during operation, which severely contaminates the weak physiological signals acquired simultaneously, such as EEG and fNIRS, making it difficult to achieve "simultaneous stimulation and estimation of network state." Although some solutions (such as WO2017037107A1) attempt to integrate EEG electrodes and TMS coils at the hardware level, they do not provide a complete system-level artifact suppression solution that can ensure high signal-to-noise ratio data during stimulation, particularly lacking hardware-level active cancellation circuitry.

[0004] Existing closed-loop modulation schemes have limitations. For example, although the relatively similar Chinese patent application CN113040790A proposes to use EEG+fNIRS synchronous monitoring for closed-loop modulation of neural stimulation, its technical solution does not reveal how to solve the problem of strong artifacts in rTMS, nor does it include a dual-station phase-locked synergistic stimulation strategy for the core pathology of pDoC (multiple network disconnections), and it does not introduce biochemical markers reflecting neurotransmitter levels as gating variables.

[0005] Current neuromodulation techniques have failed to effectively integrate with traditional Chinese medicine (TCM) therapies. TCM acupuncture, particularly press-needle therapy (a type of intradermal needle), is widely used clinically as an adjunct treatment for various neurological disorders. Its core mechanism lies in regulating the flow of Qi and blood in the meridians and the functions of the internal organs through continuous and stable microstimulation of specific acupoints (such as Baihui, Sishencong, Shuigou, and Neiguan), achieving the goal of "opening the orifices and awakening the mind." However, the effectiveness of this therapy usually relies on the qualitative assessment of the physician. How to quantify its effects in real-time and objectively, and scientifically and in real-time combine it with modern neuromodulation techniques such as rTMS to form a synergistic treatment plan, remains an unsolved technical challenge. Current technologies generally lack a method and device that can incorporate the immediate physiological effects of TCM interventions as feedback signals into a closed-loop control system.

[0006] Therefore, existing technologies generally suffer from problems such as being open-loop, blind, lacking individualization and pathway-level targeting, and failing to achieve synergistic effects through the integration of traditional Chinese and Western medicine. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a closed-loop neuromodulation system and method that can monitor multimodal biological information of pDoC patients in real time, synchronously and accurately during rTMS stimulation, including central nervous activity and peripheral acupoint physiological responses, and adaptively optimize stimulation parameters online accordingly.

[0008] The technical solution of this invention is as follows: A traditional Chinese medicine acupuncture combined with neuromodulation adaptive arousal system for consciousness disorders includes: The central control and processing unit has a software core running inside. The stimulation subsystem includes a dual-channel phase-locked controller synchronized by a clock source, and two power channels that drive the posterior cortical coil and the anterior cortical coil, respectively. The acquisition subsystem includes a multimodal acquisition controller also synchronized by the clock source, and a sensor and amplifier for acquiring at least one brain function signal, including qEEG signal and fNIRS signal; The magnetic artifact suppression module includes an artifact reference probe and an analog artifact cancellation circuit, wherein the analog artifact cancellation circuit is connected between the amplifier and the analog-to-digital converter of the acquisition subsystem. The TCM status monitoring module is used to monitor changes in local subcutaneous physiological parameters caused by acupuncture at specific acupoints on the patient, and convert these parameter changes into digital signals that are input to the central control and processing unit. The central control and processing unit is configured to uniformly receive and process digital data streams from the acquisition subsystem, externally input biochemical markers, and acupoint physiological state signals from the traditional Chinese medicine state monitoring module. It estimates the state of the consciousness network in real time through the network-loop online inversion module and multimodal information fusion algorithm inside the central control and processing unit, and calculates optimized stimulation parameters by integrating all input information through the adaptive parameter engine. Then, it sends the optimized stimulation parameters to the stimulation subsystem through digital parameter instructions to form closed-loop control.

[0009] Furthermore, the stimulation subsystem employs a dual-station phase-locked collaborative stimulation strategy, and the dual-channel phase-locked controller generates a dual-station phase difference δ as a control dimension that can be optimized online, for pathway-level regulation targeting the multi-network disconnection pathology of patients with chronic consciousness disorders.

[0010] Furthermore, the magnetic artifact suppression module adopts an integrated architecture consisting of hardware gating, front-end analog cancellation, and back-end digital suppression.

[0011] Furthermore, the TCM state monitoring module includes a laser Doppler blood flow probe, which is placed next to acupoints such as Baihui, Sishencong, Shuigou, or Neiguan to monitor the microcirculation blood flow index as a quantitative indicator and as a real-time feedback variable for closed-loop regulation.

[0012] Furthermore, the core software includes a data, security, and session management module, which is used to store treatment data, monitor security interlocking mechanisms, and save an individualized network model at the end of the session.

[0013] Furthermore, the adaptive parameter engine employs a phased strategy, including a pro-awakening phase for the thalamus-posterior cortex network and a pro-integration phase for the frontoparietal network. It evaluates network state changes and optimizes stimulation parameters, including stimulation frequency f, stimulation intensity I, and bi-station phase difference δ, through a reward function. Before updating the stimulation parameters, the adaptive parameter engine also determines whether to perform parameter updates or adjust the update step size based on preset thresholds of biochemical markers and the change in the acupoint microcirculation blood flow index relative to the baseline.

[0014] Furthermore, the network-loop online inversion module uses a dynamic causal modeling method to estimate the thalamus-posterior cortex weight W, the frontoparietal connectivity strength C, and the global integration index G as target variables for closed-loop optimization.

[0015] A closed-loop adaptive neural modulation method applied to the above system, executed by a central control and processing unit, is used for signal processing and parameter control of the stimulation subsystem and the acquisition subsystem; it includes the following steps: Step 1, Initialization Phase: Perform fNIRS and qEEG resting-state network topology identification, assess the patient's initial conscious network state, set initial stimulation parameters, apply press needles and install TCM sensors; Step 2: Perform dual-station phase-locked stimulation: Generate high-power pulses based on the current parameters and stimulate the posterior and anterior cortices respectively; Step 3, Synchronous Acquisition and Artifact Suppression: Acquire qEEG and fNIRS signals during the stimulation interval, and simultaneously acquire artifact reference signals for analog domain differential cancellation; Step 4, Online Network-Loop Inversion: Estimate the state parameters of the consciousness network, including the thalamus-posterior cortex weight W, the frontoparietal connection strength C, and the global integration index G; Step 5: Input multimodal biomarkers, including fused biochemical biomarkers and physiological signals from traditional Chinese medicine acupoints; Step 6: Data quality assessment: If the data is unqualified, discard the collected data and keep the current stimulation parameters unchanged; if the data is qualified, determine the current treatment stage. Step 7, Phased Decision Making and Optimization: During the arousal promotion phase, evaluate the arousal promotion reward function and optimize the stimulus parameters to determine whether the arousal threshold has been reached; if the arousal threshold has been reached, switch to the integration promotion phase; during the integration promotion phase, evaluate the integration promotion reward function and optimize the stimulus parameters. Step 8: Update the stimulation parameters for the next round and determine whether the session has ended. If not, execute the process in a loop; if it has ended, save the data and model.

[0016] Furthermore, the arousal-promoting reward function is: , in, This is an arousal response, representing the system's arousal level or response intensity. The change in the weights of the thalamus-posterior cortex network reflects the degree of activation of arousal-related pathways; The change in the global integration index reflects the improvement in the overall network integration level of the brain; Given the current stimulus intensity, These are the weighting coefficients. This is the baseline term, used to set the initial bias for rewards; This is used to modulate the positive contribution of changes in thalamic-posterior cortical weighting to arousal responses. Used to adjust for the positive contribution of changes in the global integration index to arousal response. Used to control the inhibitory effect of the squared term of stimulus intensity on reward, and to prevent adverse reactions caused by excessive stimulation; The integration incentive function is: , in, The integration response represents the system's level of response to information integration or coordination capabilities; The changes in connection strength reflect the degree of enhancement of the frontal and parietal networks and cross-regional information integration pathways. This is the forehead-to-crown connection strength value measured at the current moment. The value of the frontoparietal connection strength measured at the previous moment; Changes in the global integration index represent an improvement in the brain's overall integration ability. The global integration index is calculated at the current moment. The global integration index calculated at the previous moment; Stimulus intensity For phase difference change, The phase difference between the two stations at the current moment. The phase difference between the two stations at the previous moment is used to characterize the synchronicity adjustment amplitude of the bi-station coordinated stimulation. Excessive phase difference changes indicate asynchrony between different neural regions and effective integration of inhibitory information. These are the weighting coefficients. For baseline bias term, and This is used to enhance the positive contribution of changes in connection strength and global integration index to promoting consciousness integration. Used to mitigate the risks associated with excessive stimulation intensity. It is used to suppress the asynchrony caused by excessive phase difference adjustment and to ensure the coordination and physiological rationality of dual-station stimulation.

[0017] A computer-readable storage medium storing computer instructions, which, when executed by a processor, cause the processor to perform the above-described method steps and output updated control parameters—stimulation frequency f, stimulation intensity I, and bi-station phase difference δ—to the stimulation subsystem in the form of digital parameter instructions.

[0018] The beneficial effects of this invention are: This invention presents an adaptive arousal system and method for consciousness disorders fused with traditional Chinese medicine acupuncture and neuromodulation. It employs an integrated artifact-immune architecture of "hardware gating + front-end analog cancellation + back-end digital inhibition," successfully solving the problem of electromagnetic artifact interference during repetitive transcranial magnetic stimulation (rTMS), thus transforming from a traditional offline assessment mode to real-time online closed-loop control. This allows the system to dynamically adjust stimulation parameters based on real-time changes in the patient's brain state, avoiding the risk of blind treatment. By using the thalamus-posterior cortex network and frontoparietal network state as closed-loop targets and introducing a phased adaptive strategy of "promoting arousal-promoting integration," this invention achieves pathway-level precise regulation of the multi-network disconnection pathology in pDoC patients. This paradigm is the first to use quantifiable network parameters (such as weights W, connectivity C, and global integration index G) as the optimization basis, avoiding "one-size-fits-all" interventions. The dual-channel phase-locked co-stimulation subsystem uses the dual-station phase difference δ as a novel control dimension that can be optimized online, enabling rapid matching of the patient's brain network segments. The individualized resonance window design optimizes the energy utilization of stimulation, reduces ineffective pulses, shortens treatment time, and lowers energy consumption, resulting in significant economic benefits. By incorporating biochemical markers and TCM state markers as gating variables into the closed loop, multi-dimensional information fusion and dose-response matching are achieved. Simultaneously, built-in data quality judgment and safety interlock mechanisms (such as automatic pause for abnormal signals) ensure a high degree of safety during treatment. Compared to existing methods, this effectively reduces the incidence of side effects and improves individualized precision, making it more suitable for clinical environments. By using quantitative physiological indicators of TCM acupuncture therapy (such as the acupoint microcirculation blood flow index) as real-time feedback variables, this invention organically integrates modern neuromodulation with traditional TCM wisdom. This synergistic mechanism may enhance the patient's responsiveness to rTMS through the pre-regulatory effect of acupuncture (such as improving blood circulation). This invention not only fills the gap in the integration of TCM and Western medicine but also provides a novel overall solution for pDoC treatment, with broad prospects for social application. Attached Figure Description

[0019] Figure 1 This is a structural block diagram of an adaptive arousal system for consciousness disorders that integrates traditional Chinese medicine acupuncture with neuromodulation, according to the present invention.

[0020] Figure 2 This is a flowchart of a phased closed-loop adaptive control method for an adaptive awakening method for consciousness disorders that integrates traditional Chinese medicine acupuncture with neuromodulation, according to the present invention. Detailed Implementation

[0021] This invention proposes an adaptive awakening system and method for prolonged disorders of consciousness (pDoC) that integrates traditional Chinese medicine acupuncture and neuromodulation. Designed primarily for patients with pDoC, this system combines modern neuromodulation techniques (such as repetitive transcranial magnetic stimulation, rTMS) with traditional Chinese medicine acupuncture. By real-time monitoring of multimodal biosignals (such as qEEG, fNIRS, biochemical markers, and acupoint physiological responses), it achieves individualized closed-loop regulation, thereby improving the precision and stability of treatment efficacy. The system emphasizes the synergistic effect of traditional Chinese and Western medicine, avoiding the open-loop, blind, and pathway-level targeting problems common in existing technologies. The following two embodiments describe the system's hardware architecture, connectivity, working principle, and methodological flow step by step.

[0022] Example 1: As Figure 1 As shown, the system of this invention adopts a modular design. The core architecture includes a central control and processing unit (host), a stimulation subsystem, and an acquisition subsystem. The magnetic artifact suppression module is configured in the front-end acquisition link (including an artifact reference probe, an artifact amplifier, and an analog denoising signal unit) to eliminate magnetic interference in real time. The TCM status monitoring module is integrated in the central control and processing unit to process acupoint physiological signals. These modules and subsystems are globally synchronized through a high-precision master clock signal to ensure real-time acquisition, processing, and feedback of signals during rTMS stimulation. At the same time, the system incorporates TCM acupoint monitoring to form a complete closed-loop link.

[0023] Central Control and Processing Unit: The central computer is the core of the system's control and decision-making. Its internal software integrates a Network-Loop Online Inversion Module (NCM) and an adaptive parameter engine. Operators set treatment plans through a user interface and can input offline-detected biochemical indicators (such as dopamine concentration) from the biochemical marker input port. The computer is also connected to a Traditional Chinese Medicine Status Monitor (TCM Status Monitor) for real-time reception and processing of data from TCM sensors. The TCM sensors include laser Doppler blood flow, near-infrared tissue oxygenation, skin temperature, skin conductance response, micro-vibration, or local pressure sensors. The master clock and synchronization unit provide a high-precision time reference for the system, ensuring strict synchronization between the stimulation subsystem and the acquisition subsystem. The output of this unit sends digital parameter instructions to the stimulation subsystem, and the input receives the digital data stream from the acquisition subsystem. The digital data stream serves as feedback or gating variables for closed-loop optimization. In terms of working principle, this unit uniformly processes qEEG, fNIRS, biochemical markers, and TCM signals. It estimates the state of the consciousness network (such as the thalamus-posterior cortex weight W) through algorithms and calculates optimization parameters (such as the stimulation frequency f). The safety interlock mechanism is built into the data management module. If an abnormality is detected (such as excessive signal noise), the system will automatically suspend operation to ensure patient safety.

[0024] Stimulation Subsystem: This subsystem is responsible for generating precise magnetic stimulation pulses. It includes a master clock and synchronization unit (using a high-stability crystal oscillator OCXO as the clock source), a dual-channel phase-locked controller, channel A driver, and channel B driver. The connection is designed as follows: the master clock signal is distributed to all modules to ensure synchronization accuracy at the microsecond level; digital parameter commands are transmitted from the central unit to the phase-locked controller, which generates a phase-precise waveform. After being amplified by the channel A driver and the channel B driver, it forms a high-power stimulation pulse, which is used to drive the first cortical coil for modulating the thalamus-posterior cortical network and the second cortical coil for modulating the frontoparietal network, respectively. The working principle is to achieve phase-locked synergistic stimulation of two stations (posterior cortex to promote arousal and anterior cortex to promote integration). For the pathological conditions of multiple network disconnections in pDoC patients (such as thalamic-posterior cortex disconnection), the phase difference δ of the two stations is used as a new control dimension (range 0°-180°) and can be optimized online to enhance network resonance. Stimulation parameters example: stimulation frequency f is 1-10Hz, stimulation intensity I is 80-120% of the motion threshold, pulse width is 200-500μs, online resolution of phase difference δ of the two stations is ≤1°, and phase-locking error is ≤0.5°.

[0025] Acquisition Subsystem: Under the precise timing control of the multimodal acquisition controller, this subsystem acquires raw analog signals from the brain via hardware gating / trigger signals during the intervals between stimulation artifacts. Simultaneously, a clean artifact signal is acquired by the qEEG sensing cap and fNIRS electrodes. These two signals are then processed by a preamplifier and an artifact reference amplifier, respectively, before being sent to an analog artifact removal signal processing unit for differential cancellation. The processed signals are then amplified and filtered by the main amplification and filtering unit, and finally converted into a digital data stream by an analog-to-digital converter before being transmitted back to the central computer for analysis. The working principle is to simultaneously acquire electroencephalogram (qEEG, sampling rate 500-1000 Hz, covering α and β bands) and near-infrared spectroscopy (fNIRS, wavelengths 760 nm and 850 nm, monitoring blood oxygen saturation) signals for real-time assessment of brain activity. Specific parameter examples: the qEEG sensing cap uses a 64-channel dry electrode system, and the fNIRS electrodes are positioned in the frontal and parietal lobes to ensure coverage of key conscious networks.

[0026] Magnetic Artifact Suppression Module: This module is designed to suppress electromagnetic artifact interference during stimulation, enabling synchronous acquisition of qEEG and fNIRS signals. It includes: an artifact reference probe, an artifact reference amplifier, an analog artifact removal signal processing unit, a main amplification and filtering unit, and an analog-to-digital converter. The connection adopts a differential structure: the artifact reference probe independently acquires clean artifact signals, which are then canceled in real-time with the original signal in the analog domain. The processed signal is filtered (bandpass 10-100Hz) and converted before being transmitted back as a digital stream. The working principle is based on a "three-level suppression" architecture: hardware gating blocks peak artifacts, front-end analog cancellation reduces residual interference, and back-end digital filtering refines the data, ensuring a signal-to-noise ratio >20dB, enabling measurability during stimulation.

[0027] Traditional Chinese Medicine (TCM) Status Monitoring Module: This module integrates TCM elements into the system, including TCM sensors (such as laser Doppler blood flow probes). The connection relationship is as follows: the TCM sensors are placed next to specific acupoints of the patient (such as Baihui, Sishencong, Shuigou, Neiguan) to continuously monitor changes in local microcirculation and other physiological indicators caused by acupuncture stimulation. The digital signals are then transmitted to the TCM status monitor for processing and input into the central computer as one of the real-time feedback variables for closed-loop regulation. The monitoring frequency is 1-5Hz.

[0028] Example 2: As Figure 2As shown, the method of this invention adopts an iterative closed-loop process, forming a dynamically optimized treatment cycle from initialization to completion. This process is guided by consciousness network parameters (thalamus-posterior cortex weight W, frontoparietal connectivity strength C, and global integration index G), integrates multimodal information, ensures that the treatment gradually shifts from promoting arousal to promoting integration, and naturally incorporates TCM feedback, making the whole process more in line with individual patient differences. Among them, the global integration index G is a quantitative indicator used to assess the overall integration level of the brain network, playing an important role in the fields of neuroscience, consciousness research, and rehabilitation engineering. This index originates from brain network analysis theory and aims to reflect the efficiency of information flow and coordination between different brain regions, and is particularly suitable for the diagnosis, monitoring, and treatment optimization of patients with prolonged disorders of consciousness (pDoC).

[0029] Initialization Phase: After system startup, baseline assessment is first performed: resting-state fNIRS and qEEG signals are acquired to identify the initial network topology, such as calculating the thalamus-posterior cortex weight W=0.3, frontoparietal connectivity strength C=0.2, and global integration index G=0.4. This step provides a benchmark for subsequent optimization. Next, initial parameters are set, such as f=5 Hz, I=100% motion threshold, and δ=0°. Simultaneously, press needles are applied and TCM sensors are installed: the physician applies press needles to acupoints, and the sensors monitor blood flow changes in real time. This phase ensures that the treatment is based on the patient's actual condition and avoids blind stimulation. Main closed-loop iterative process: Execute dual-station phase-locked stimulation: Generate pulses based on current parameters, targeting the posterior cortex for arousal and the anterior cortex for integration. Next, perform synchronous acquisition and artifact suppression: Acquire signals during the intervals and cancel artifacts using simulated differential methods. Then, proceed to network-loop online inversion (NCM): Estimate network parameters using dynamic causal modeling, with the following formula: , , in, The optimal parameter estimation vector represents the set of parameters obtained by minimizing the objective function, which reflects the current state of the patient's consciousness network. Let be the vector of parameters to be estimated. The observation vector consists of multimodal physiological signals, including: qEEG power spectrum (reflecting the intensity of cortical electrical activity), fNIRS blood oxygenation changes (reflecting cerebral blood flow and metabolic levels), and TCM blood flow index (measured by a laser Doppler blood flow probe and used to quantify microcirculation changes at acupoints). This is a forward model of the system, used to generate a model based on a given stimulus input. and parameters Predict the corresponding multimodal physiological responses; this model characterizes the effects of stimulus parameters (such as frequency, intensity, and bi-station phase difference) on brain network dynamics; The stimulus input vector contains the control parameters of transcranial magnetic stimulation (stimulation frequency f, stimulation intensity I, and bi-station phase difference δ), representing the external adjustable control variables of the system. is the regularization coefficient, with a value of 0.05, used to balance the fitting error term and the regularization term, prevent overfitting, and ensure the model's generalization ability among individuals; Regularization terms are used to constrain parameters. To ensure the smoothness and physical interpretability of parameter estimation, the complexity or magnitude can be determined using L2 norm or Tikhonov regularization. It is a non-linear activation function (which can be in the form of Sigmoid or tanh) used to simulate the saturation characteristics of neural responses; This is a neural state vector, representing the average activation level of each network node (such as the thalamus, posterior cortex, and frontoparietal cortex) at the current moment; This is a bias term used to compensate for system background noise or resting-state differences. Subsequently, multimodal biomarkers are input: biochemical biomarkers (such as dopamine > threshold) and TCM state biomarkers (such as blood flow index > 20% of baseline). Data quality is assessed: if the signal-to-noise ratio is insufficient, the data is discarded and the parameters are maintained, returning to the stimulation step; if the data quality is satisfactory, the adaptive parameter engine determines the current treatment stage. Phased Decision Making and Optimization: Determining the Current Phase: If W < 0.5, proceed to the Arousal Promotion Phase (Yes Path): Evaluating the Reward Function: , in, This is an arousal response, representing the system's arousal level or response intensity. The change in the weights of the thalamus-posterior cortex network reflects the degree of activation of arousal-related pathways; The change in the global integration index reflects the improvement in the overall network integration level of the brain; Given the current stimulus intensity, These are the weighting coefficients. This is the baseline term, used to set the initial bias for rewards; This is used to modulate the positive contribution of changes in thalamic-posterior cortical weighting to arousal responses. Used to adjust for the positive contribution of changes in the global integration index to arousal response. Used to control the inhibitory effect of the squared term of stimulus intensity on reward, and to prevent adverse reactions caused by excessive stimulation; Optimize parameters: Update using gradient ascent: , in, The updated parameter vector, which is the stimulus plan adjusted by the system according to the reward gradient, will be executed by the central control unit in the next stimulus cycle to continuously improve the arousal or integration effect. Let be the parameter vector at the current moment, representing the combination of stimulus parameters for the system at the t-th iteration; The learning rate controls the step size for each parameter update, with a value of 0.05. For reward function For parameters The gradient vector represents the sensitivity of each parameter to the system's arousal or integration response in the current state. This gradient is obtained from the multimodal network inversion results via the Jacobian matrix and is used to indicate whether parameters should be increased or decreased in the direction of reward enhancement. Threshold determination: If W > 0.6, switch to the integration promotion phase; otherwise, in the integration promotion phase (No path): evaluate the reward function. , in, The integration response represents the system's level of response to information integration or coordination capabilities; The changes in connection strength reflect the degree of enhancement of the frontal and parietal networks and cross-regional information integration pathways. This is the forehead-to-crown connection strength value measured at the current moment. The value of the frontoparietal connection strength measured at the previous moment; Changes in the global integration index represent an improvement in the brain's overall integration ability. The global integration index is calculated at the current moment. The global integration index calculated at the previous moment; Stimulus intensity For phase difference change, The phase difference between the two stations at the current moment. The phase difference between the two stations at the previous moment is used to characterize the synchronicity adjustment amplitude of the bi-station coordinated stimulation. Excessive phase difference changes indicate asynchrony between different neural regions and effective integration of inhibitory information. These are the weighting coefficients. For baseline bias term, and This is used to enhance the positive contribution of changes in connection strength and global integration index to promoting consciousness integration. Used to mitigate the risks associated with excessive stimulation intensity. Used to suppress asynchrony caused by excessive phase difference adjustment, ensuring the coordination and physiological rationality of dual-station stimulation; Optimize parameters: Update using the same gradient ascent method described above; Finally, update the parameters for the next round; Cycle control and termination: Determining the end of the session: If the treatment duration exceeds 30 minutes or G>0.8, the data and model, including the individualized network model and treatment log, are saved and the session ends. A report is generated at this stage for subsequent analysis.

[0030] Example 3: A computer-readable storage medium storing computer instructions, which, when executed by a processor, cause the processor to perform the above-described method steps and output the updated control parameters—stimulation frequency f, stimulation intensity I, and bi-station phase difference δ—to the stimulation subsystem in the form of digital parameter instructions.

[0031] Through the above embodiments, the present invention achieves a smooth transition from evaluation to optimization. After integrating traditional Chinese medicine, the global integration index G is significantly improved, and the stability of the therapeutic effect is enhanced. The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the embodiments of the present invention can have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A self-adaptive arousal system for consciousness disorders integrating traditional Chinese medicine acupuncture and neuromodulation, characterized in that, include: The central control and processing unit has a software core running inside. The stimulation subsystem includes a dual-channel phase-locked controller synchronized by a clock source, and two power channels that drive the posterior cortical coil and the anterior cortical coil, respectively. The acquisition subsystem includes a multimodal acquisition controller also synchronized by the clock source, and a sensor and amplifier for acquiring at least one brain function signal, including qEEG signal and fNIRS signal; The magnetic artifact suppression module includes an artifact reference probe and an analog artifact cancellation circuit, wherein the analog artifact cancellation circuit is connected between the amplifier and the analog-to-digital converter of the acquisition subsystem. The TCM status monitoring module is used to monitor changes in local subcutaneous physiological parameters caused by acupuncture at specific acupoints on the patient, and convert these parameter changes into digital signals that are input to the central control and processing unit. The central control and processing unit is configured to uniformly receive and process digital data streams from the acquisition subsystem, externally input biochemical markers, and acupoint physiological state signals from the traditional Chinese medicine state monitoring module. Through the network-loop online inversion module within the central control and processing unit, it estimates the state of the consciousness network in real time based on the fusion information of the digital data streams, externally input biochemical markers, and acupoint physiological state signals. An adaptive parameter engine then calculates optimized stimulation parameters by integrating all input information and sends these parameters to the stimulation subsystem via digital parameter commands, forming a closed-loop control.

2. The system according to claim 1, characterized in that, The stimulation subsystem employs a dual-station phase-locked collaborative stimulation strategy. The dual-channel phase-locked controller generates a dual-station phase difference δ as a control dimension that can be optimized online, and is used for pathway-level regulation of the multi-network disconnection pathology in patients with chronic consciousness disorders.

3. The system according to claim 1, characterized in that, The magnetic artifact suppression module adopts an integrated architecture consisting of hardware gating, front-end analog cancellation, and back-end digital suppression.

4. The system according to claim 1, characterized in that, The TCM status monitoring module includes a laser Doppler blood flow probe, which is placed next to acupoints such as Baihui, Sishencong, Shuigou, or Neiguan to monitor the microcirculation blood flow index as a quantitative indicator and as a real-time feedback variable for closed-loop regulation.

5. The system according to claim 1, characterized in that, The core software includes a data, security, and session management module, which is used to store treatment data, monitor security interlocking mechanisms, and save an individualized network model at the end of the session.

6. The system according to claim 1, characterized in that, The adaptive parameter engine employs a phased strategy, including a pro-awakening phase for the thalamus-posterior cortex network and a pro-integration phase for the frontoparietal network. It evaluates network state changes and optimizes stimulation parameters, including stimulation frequency f, stimulation intensity I, and bi-station phase difference δ, through a reward function. Before updating the stimulation parameters, the adaptive parameter engine also determines whether to perform parameter updates or adjust the update step size based on preset thresholds of biochemical markers and the change in the acupoint microcirculation blood flow index relative to the baseline.

7. The system according to claim 1, characterized in that, The network-loop online inversion module uses a dynamic causal modeling method to estimate the thalamus-posterior cortex weight W, frontoparietal connectivity C, and global integration index G as target variables for closed-loop optimization.

8. A closed-loop adaptive neural modulation method applied to the system of claim 1, characterized in that, Executed by a central control and processing unit, the method performs signal processing and parameter control on the stimulation subsystem and the acquisition subsystem; the method includes: Step 1, Initialization Phase: Perform fNIRS and qEEG resting-state network topology identification, assess the patient's initial conscious network state, set initial stimulation parameters, apply press needles and install TCM sensors; Step 2: Perform dual-station phase-locked stimulation: Generate high-power pulses based on the current parameters and stimulate the posterior and anterior cortices respectively; Step 3, Synchronous Acquisition and Artifact Suppression: Acquire qEEG and fNIRS signals during the stimulation interval, and simultaneously acquire artifact reference signals for analog domain differential cancellation; Step 4, Online Network-Loop Inversion: Estimate the state parameters of the consciousness network, including the thalamus-posterior cortex weight W, the frontoparietal connection strength C, and the global integration index G; Step 5: Input multimodal biomarkers, including fused biochemical biomarkers and physiological signals from traditional Chinese medicine acupoints; Step 6: Data quality assessment: If the data is unqualified, discard the collected data and keep the current stimulation parameters unchanged; if the data is qualified, determine the current treatment stage. Step 7, Phased Decision Making and Optimization: During the arousal promotion phase, evaluate the arousal promotion reward function and optimize the stimulus parameters to determine whether the arousal threshold has been reached; if the arousal threshold has been reached, switch to the integration promotion phase; during the integration promotion phase, evaluate the integration promotion reward function and optimize the stimulus parameters. Step 8: Update the stimulation parameters for the next round and determine whether the session has ended. If not, execute the process in a loop; if it has ended, save the data and model.

9. The method according to claim 8, characterized in that, The arousal-promoting reward function is: ， in, This is an arousal response, representing the system's arousal level or response intensity. The change in the weights of the thalamus-posterior cortex network reflects the degree of activation of arousal-related pathways; The change in the global integration index reflects the improvement in the overall network integration level of the brain; Given the current stimulus intensity, These are the weighting coefficients. This is the baseline term, used to set the initial bias for rewards; This is used to modulate the positive contribution of changes in thalamic-posterior cortical weighting to arousal responses. Used to adjust for the positive contribution of changes in the global integration index to arousal response. It is used to control the inhibitory effect of the squared term of stimulus intensity on reward and prevent adverse reactions caused by excessive stimulation; The integration incentive function is: ， in, The integration response represents the system's level of response to information integration or coordination capabilities; The changes in connection strength reflect the degree of enhancement of the frontal and parietal networks and cross-regional information integration pathways. This is the forehead-to-crown connection strength value measured at the current moment. The value of the frontoparietal connection strength measured at the previous moment; Changes in the global integration index represent an improvement in the brain's overall integration ability. The global integration index is calculated at the current moment. The global integration index calculated at the previous moment; Stimulus intensity For phase difference change, The phase difference between the two stations at the current moment. The phase difference between the two stations at the previous moment is used to characterize the synchronicity adjustment amplitude of the bi-station coordinated stimulation. Excessive phase difference changes indicate asynchrony between different neural regions and effective integration of inhibitory information. These are the weighting coefficients. For baseline bias term, and This is used to enhance the positive contribution of changes in connection strength and global integration index to promoting consciousness integration. Used to mitigate the risks associated with excessive stimulation intensity. It is used to suppress the asynchrony caused by excessive phase difference adjustment and to ensure the coordination and physiological rationality of dual-station stimulation.

10. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed by the processor, the processor performs the steps of the method described in claim 8 or 9, and outputs the updated control parameters—stimulation frequency f, stimulation intensity I, and bi-station phase difference δ—to the stimulation subsystem in the form of digital parameter instructions.

Images