Closed-loop deep brain stimulation parameter optimization system and method for patients with impaired consciousness

Abstract

The application provides a closed-loop deep brain stimulation parameter optimization system and method for patients with consciousness disorders, and belongs to the technical field of medical devices. The system comprises: a stimulation module, which is an implanted medical stimulation device, used to receive standardized instructions from a signal processing and control module, and to apply electrical stimulation to a target point in the brain of a patient according to preset stimulation parameters in the instructions; an electroencephalogram (EEG) monitoring module, which is a non-invasive medical monitoring device, used to collect EEG signals of the scalp of the patient in real time during or intermittently during electrical stimulation, and to transmit standardized digital signals to the signal processing and control module after signal conditioning; and the signal processing and control module, used to receive standardized digital signals from the EEG monitoring module and to perform preprocessing, calculate neuroelectrophysiological indicators of quantified consciousness states, realize automatic optimization, and output parameter adjustment instructions to the stimulation module. The application realizes personalized and accurate treatment, improves optimization efficiency and treatment effect, and also improves the degree of automation.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, and in particular to a closed-loop deep brain stimulation parameter optimization system and method for patients with impaired consciousness. Background Technology

[0002] Disorders of Consciousness (DoC) are a clinical syndrome characterized by severe impairment of both the level and content of consciousness due to serious brain injury. Deep brain stimulation (DBS), an emerging neuromodulation technique, modulates abnormal neural circuit activity by implanting electrodes into specific targets in the brain, such as the central thalamus, and applying electrical impulses. In recent years, increasing research has demonstrated the potential of DBS in promoting the recovery of consciousness in patients with disorders of consciousness.

[0003] Currently, in clinical practice, when applying DBS technology to patients with impaired consciousness, the setting of stimulation parameters (such as frequency, pulse width, voltage / current, stimulation duration, etc.) mainly relies on the following two methods: Based on parameters reported in the literature: Clinicians refer to the DBS parameter combinations that have been reported in published academic literature as effective for certain patients with disorders of consciousness for treatment. For example, using a frequency of 70 Hz and a pulse width of 90 microseconds.

[0004] Physician-based trial-and-error adjustment: The physician first sets an initial set of stimulus parameters for the patient. After a period of treatment (several days or weeks), the patient's changes in consciousness are assessed using clinical scales (such as the JFK Coma Recovery Scale-Revised, CRS-R). If the effect is unsatisfactory, the physician will adjust one or more parameters based on personal experience, and then observe and evaluate for another period of time. This process is repeated continuously in order to find the most effective combination of parameters for a particular patient.

[0005] Both of the above-mentioned existing technical solutions have obvious drawbacks: Lack of personalization and objectivity: Parameter-based approaches fail to account for the significant individual differences among patients with disorders of consciousness (such as etiology, disease course, location and extent of brain injury). A universal parameter reported in the literature may be completely ineffective for a specific patient, leading to poor treatment outcomes. This approach cannot achieve personalized and precise treatment.

[0006] The adjustment process is time-consuming, labor-intensive, and lacks sufficient evidence: While physician-based trial-and-error adjustments represent a step towards personalization, the process is extremely lengthy and inefficient. Each parameter adjustment requires a long observation period to assess its effectiveness, significantly increasing the time and financial burden on patients, their families, and the medical team. More importantly, the direction and pace of parameter adjustments rely heavily on the physician's subjective experience, lacking immediate and objective biological indicators as a basis for adjustment. This makes the entire process resemble blind trial and error, making it difficult to guarantee the efficient discovery of optimal parameters.

[0007] Significant feedback delay exists: both behavioral observation and scale assessment exhibit substantial lag. While the brain's electrophysiological responses to DBS stimulation are immediate, the translation of these responses into observable behavioral changes takes a considerable amount of time. Existing methods are unable to capture and utilize this immediate neurophysiological feedback signal to guide parameter optimization.

[0008] Therefore, this invention proposes a closed-loop deep brain stimulation parameter optimization system and method for patients with impaired consciousness. Summary of the Invention

[0009] This invention provides a closed-loop deep brain stimulation parameter optimization system and method for patients with impaired consciousness, in order to solve the aforementioned technical problems.

[0010] This invention provides a closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness, comprising: The stimulation module is an implantable medical stimulation device used to receive standardized instructions from the signal processing and control module and apply electrical stimulation to the target points in the patient's brain according to the preset stimulation parameters in the standardized instructions. The EEG monitoring module is a non-invasive medical monitoring device used to collect EEG signals from the patient's scalp in real time during or between electrical stimulation sessions. After signal conditioning, the signals are transmitted to the signal processing and control module as standardized digital signals. The signal processing and control module is used to receive standardized digital signals from the EEG monitoring module and perform preprocessing, calculate and quantify the neurophysiological indicators of the state of consciousness based on a preset algorithm, realize automatic optimization of stimulation parameters, and output parameter adjustment instructions to the stimulation module.

[0011] Preferably, the stimulation parameters include: frequency, pulse width, and amplitude.

[0012] Preferably, the implantable medical stimulation device includes: an implantable pulse generator, an extension lead, and a DBS electrode implanted in the brain at a target site, wherein the target site is the central thalamus.

[0013] Preferably, the neurophysiological indicators of the state of consciousness include one or more of the following: brain electrical complexity indicators and brain inter-brain functional connectivity indicators.

[0014] Preferably, the signal processing and control module includes: The initialization unit is used to set the parameter space to be optimized and the search step size, and at the same time, to set a set of initial stimulus parameters P_initial; The baseline measurement unit is used to acquire a baseline EEG signal by the EEG monitoring module without applying or with initial parameter stimulation, and to calculate the baseline consciousness index value Index_baseline based on the signal processing and control module. The parameter traversal and stimulation unit is used by the signal processing and control module to automatically generate a new parameter combination P_i in the parameter space according to the search step size, and send instructions to the stimulation module to make the stimulation module perform electrical stimulation several times according to P_i. The synchronous monitoring and calculation unit is used to continuously collect EEG signals during the short window period after stimulation using parameter combination P_i, and to process the continuously collected EEG signals in real time based on the signal processing and control module, and calculate the consciousness index value Index_i corresponding to parameter combination P_i and normalized by the baseline consciousness index value. The data storage and iteration unit is used to record the pairing of parameter combination P_i and corresponding consciousness index value Index_i, and return to the parameter traversal and stimulation unit to select the next untested parameter combination P_i+1 for testing, and repeat the execution operations of the synchronous monitoring and calculation unit and the data storage and iteration unit until all preset parameter combinations have been traversed. The optimal parameter determination unit is used to compare the consciousness index values ​​of all records after the traversal is completed and select the parameter combination P_optimal that produces the maximum consciousness index value. The application and output unit is used to use the found optimal parameter P_optimal as the recommended parameter for the patient's subsequent long-term DBS treatment.

[0015] This invention provides a method for optimizing closed-loop deep brain stimulation parameters for patients with impaired consciousness, comprising: Step 1: Receive standardized instructions from the signal processing and control module, and apply electrical stimulation to the target point in the patient's brain according to the preset stimulation parameters in the standardized instructions; Step 2: During or between electrical stimulation sessions, EEG signals from the patient's scalp are collected in real time, and after signal conditioning, standardized digital signals are transmitted to the signal processing and control module. Step 3: Receive the standardized digital signal from the EEG monitoring module and perform preprocessing. Calculate the neurophysiological indicators of the quantified state of consciousness based on a preset algorithm to achieve automatic optimization of stimulation parameters and output parameter adjustment instructions to the stimulation module.

[0016] Compared with the prior art, the beneficial effects of this application are as follows: 1. Personalized and precise treatment: This system directly uses the patient's own brain's real-time electrophysiological response to different stimulation parameters to determine the optimal parameters, eliminating the reliance on general parameters in literature or the subjective experience of physicians, and tailoring the most effective treatment plan for each patient.

[0017] 2. Significantly improves optimization efficiency: The traditional trial-and-error process that takes weeks or even months is shortened to an automated scanning process that takes only hours.

[0018] 3. Provides objective quantitative basis: It adopts quantitative consciousness indicators based on EEG calculation to replace the lagging and somewhat subjective behavioral assessment, so that the process and results of parameter optimization are based on objective and repeatable biological data.

[0019] 4. Potential to improve treatment outcomes: By systematically and comprehensively exploring the parameter space, it is possible to find the optimal parameter combination that may be missed by traditional trial and error methods, thereby potentially maximizing the effect of DBS on promoting the recovery of patient consciousness.

[0020] 5. High degree of system automation: The entire optimization process is executed automatically by the system, reducing human intervention and ensuring the standardization of the process and the reliability of the results.

[0021] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0022] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0023] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a structural diagram of a closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness, as described in an embodiment of the present invention. Figure 2 This is a flowchart of a closed-loop deep brain stimulation parameter optimization method for patients with impaired consciousness, as described in an embodiment of the present invention. Detailed Implementation

[0024] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0025] This invention provides a closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness, such as... Figure 1 As shown, it includes: The stimulation module is an implantable medical stimulation device used to receive standardized instructions from the signal processing and control module and apply electrical stimulation to the target points in the patient's brain according to the preset stimulation parameters in the standardized instructions. The EEG monitoring module is a non-invasive medical monitoring device used to collect EEG signals from the patient's scalp in real time during or between electrical stimulation sessions. After signal conditioning, the signals are transmitted to the signal processing and control module as standardized digital signals. The signal processing and control module is used to receive standardized digital signals from the EEG monitoring module and perform preprocessing, calculate and quantify the neurophysiological indicators of the state of consciousness based on a preset algorithm, realize automatic optimization of stimulation parameters, and output parameter adjustment instructions to the stimulation module.

[0026] Preferably, the stimulation parameters include: frequency, pulse width, and amplitude.

[0027] Preferably, the implantable medical stimulation device includes: an implantable pulse generator, an extension lead, and a DBS electrode implanted in the brain at a target site, wherein the target site is the central thalamus.

[0028] Preferably, the neurophysiological indicators of the state of consciousness include one or more of the following: brain electrical complexity indicators and brain inter-brain functional connectivity indicators.

[0029] In this embodiment, standardized instructions refer to control signals with a fixed format and unified encoding rules. These signals are used to clearly inform the stimulation module of the stimulation parameters, stimulation duration, stimulation mode, and other information to be executed, ensuring accurate interpretation and execution by the stimulation module. Specifically, a digital encoding method is used, with the instruction format including a start bit, parameter field, check bit, and end bit. The parameter field explicitly contains the binary encoding of stimulation parameters such as frequency, pulse width, and amplitude. The check bit is used to verify the accuracy of instruction transmission and avoid false triggering. For example, standardized instructions can adopt the UART communication protocol format, with a baud rate set to 9600bps, a start bit of 1 bit, data bits of 8 bits, a check bit of 1 bit (odd parity), and an end bit of 1 bit. A specific instruction can be represented as "0x01 0x0046 0x005A 0x0003 0x02", where 0x01 is the start bit, 0x0046 corresponds to a frequency of 70Hz, 0x005A corresponds to a pulse width of 90μs, 0x0003 corresponds to an amplitude of 3V, and 0x02 is the end bit.

[0030] In this embodiment, stimulation parameters refer to key parameters used to define the characteristics of electrical stimulation signals. These parameters directly affect the regulatory effect of electrical stimulation on neural circuits and are the core regulatory indicators of DBS treatment. The value range of these parameters is determined based on clinical safety data and neurophysiological research results to ensure that the intensity of electrical stimulation applied by the parameter combination is within an effective regulatory range and does not cause damage to brain tissue. For example, stimulation parameters mainly include frequency (unit: Hz), pulse width (unit: ... The amplitude (unit: V or mA) can be set within a safe range of 20-200Hz frequency and 40-450µA pulse width. The amplitude is 0.5-10V, and the specific value should be optimized within this range according to the individual patient's condition.

[0031] In this embodiment, the target point in the patient's brain refers to a specific brain region that plays a crucial role in consciousness regulation, as determined by neuroanatomical and neurophysiological studies. This is the core location for DBS electrode implantation. Specifically, the target point is precisely located using preoperative imaging examinations such as magnetic resonance imaging (MRI) and positron emission tomography (PET), combined with a neuronavigation system, to ensure accurate implantation of the electrode tip into this region. For example, in this invention, the target point is set as the central thalamus, a key hub for consciousness regulation. Preoperatively, high-resolution images of the patient's brain are obtained using a 3.0T MRI scan. The neuronavigation system is used to accurately map the target point coordinates onto the patient's skull positioning markers. During the procedure, microelectrode recording technology is used to verify the accuracy of the target point location, ensuring that the electrode implantation error does not exceed 1mm.

[0032] In this embodiment, electrical stimulation refers to a pulsed electrical signal applied to the target brain tissue via electrodes, meeting clinical treatment requirements. Its characteristics are determined by stimulation parameters and are used to regulate the electrical activity of nerve cells. Specifically, a square wave pulse is used as the electrical stimulation signal waveform. This waveform has steep rising and falling edges, and its parameters are easily and precisely controlled. The stimulation signal output is generated by an implantable pulse generator, transmitted to the DBS electrode via an extension wire, and then released to the target tissue from the electrode tip. For example, the electrical stimulation signal can be a biphasic square wave, where the amplitude and duration of the positive and negative phase pulses are equal to avoid charge accumulation in the brain tissue. The parameters of a particular electrical stimulation signal are a frequency of 100 Hz and a pulse width of 120 mm. Amplitude 2.5V, positive pulse duration 60 seconds The duration of the negative phase pulse is 60. The pulse interval is 9ms.

[0033] In this embodiment, the period during or between electrical stimulation sessions refers to the time between two electrical stimulation sessions during the application of electrical stimulation to the target point by the stimulation module. This is a critical time window for EEG signal acquisition, ensuring that the acquired signals reflect the impact of electrical stimulation on brain activity. Specifically, a synchronization control circuit synchronizes the stimulation module and the EEG monitoring module, clearly defining the acquisition window during stimulation (the time from the start to the end of stimulation) and the acquisition window during the interval (the time from the end of one stimulation to the start of the next). The duration of the acquisition window is set according to the stimulation parameters and signal analysis requirements. For example, if the duration of a single electrical stimulation is 0.1 seconds and the interval between two stimulations is 1 second, then the acquisition window during stimulation is 5 seconds, and the acquisition window during the interval is 10 seconds. The acquisition process and the stimulation process are kept in sync through a synchronous trigger signal; the rising edge of the synchronous trigger signal corresponds to the start of stimulation, and the falling edge corresponds to the end of stimulation.

[0034] In this embodiment, real-time acquisition of EEG signals from the patient's scalp refers to the continuous capture of weak brain electrical activity signals on the surface of the patient's scalp using scalp electrodes of the EEG monitoring module. The acquisition process is uninterrupted, ensuring signal continuity and timeliness. Specifically, a high-sampling-rate analog-to-digital converter chip is used, combined with an anti-interference acquisition circuit to avoid the influence of environmental noise and physiological artifacts on signal acquisition. The sampling rate is set according to the frequency range of the EEG signal to ensure complete preservation of the signal's characteristic information. For example, a sampling rate of 256Hz or 512Hz is used, with a sampling precision of 16 bits. The impedance between the electrode and the scalp is controlled below 50kΩ. The EEG signals obtained through continuous acquisition are time-series data, with each data point corresponding to the EEG voltage value at a specific moment, in units of... For example, the EEG signal value collected at a certain moment is 2.3. .

[0035] In this embodiment, signal conditioning processing refers to a series of processes such as amplification, filtering, noise reduction, and standardization of the acquired raw EEG signal to remove interference components in the signal, enhance the characteristics of the effective signal, and convert the raw signal into a standardized signal that meets the requirements of subsequent processing.

[0036] In this embodiment, preprocessing refers to the preliminary processing operations performed by the signal processing and control module after receiving the standardized digital signal to remove residual noise and enhance the effective signal characteristics, laying the foundation for subsequent feature extraction and index calculation. Specifically, the preprocessing algorithm includes operations such as smoothing filtering, baseline correction, and signal segmentation, which are implemented through software programming. The processing is executed in real time and does not affect the overall system response speed.

[0037] In this embodiment, the preset algorithm refers to a set of algorithms pre-stored in the signal processing and control module for calculating and quantifying neurophysiological indicators of the state of consciousness and for automatically optimizing stimulus parameters. It is crucial for the system to achieve its core functions. Specifically, the algorithms are implemented using C or Python programming languages, and have undergone extensive clinical data verification and optimization to ensure their accuracy, stability, and real-time performance. The algorithm programs are stored in the module's flash memory and can be updated and upgraded via external devices. For example, the preset algorithms include EEG complexity calculation algorithms (such as the Lempel-Ziv complexity algorithm), brain region functional connectivity calculation algorithms (such as coherence analysis algorithms), and parameter optimization algorithms (such as grid search algorithms). The computational complexity of these algorithms has been optimized to ensure real-time operation on the embedded processor.

[0038] In this embodiment, the neurophysiological indicators for quantifying the state of consciousness refer to quantitative indicators extracted from EEG signals that can objectively reflect the patient's level of consciousness. Their numerical changes directly reflect the impact of electrical stimulation on the patient's state of consciousness and are the core evaluation criteria for parameter optimization. Specifically, by analyzing the time, frequency, and spatial characteristics of EEG signals, feature parameters closely related to the state of consciousness are extracted. These parameters are then quantified into single or multiple indicator values ​​using methods such as weighted summation and entropy calculation. The range of these indicators has a clear correspondence with normal consciousness and states of impaired consciousness. For example, this indicator is a weighted sum including indicators such as EEG complexity and functional connectivity between brain regions, with a value range of 0-1. The closer the value is to 1, the higher the level of consciousness.

[0039] In this embodiment, the automatic optimization of stimulation parameters refers to the process by which the signal processing and control module automatically searches for the optimal combination of stimulation parameters based on calculated neurophysiological indices using a preset algorithm, without manual intervention, thus achieving automation and intelligence in parameter optimization. Specifically, an iterative search strategy is employed, traversing different parameter combinations within a preset parameter space at a set step size, calculating the neurophysiological indices corresponding to each parameter combination, and determining the optimal parameter combination by comparing the index values. The optimization process terminates when all parameter combinations have been traversed or a preset number of iterations has been reached. For example, the parameter space is set to a frequency of 20-100Hz (step size 10Hz) and a pulse width of 40-200Hz. (Step size 20μs), amplitude 1-5V (step size 0.5V), the system traverses all parameter combinations (9×9×9=729 groups in total) in a grid search manner, calculates the consciousness index value corresponding to each group of parameters, and selects the parameter combination with the largest index value as the optimal parameter.

[0040] In this embodiment, the parameter adjustment command refers to the control signal output by the signal processing and control module to the stimulation module after completing parameter optimization, which informs the stimulation module to switch to the optimal stimulation parameters. It has the same format and encoding rules as the standardized command. For example, if the optimal parameter combination is a frequency of 80Hz, a pulse width of 120μs, and an amplitude of 3V, the parameter adjustment command is represented as "0x01 0x0050 0x0078 0x0003 0x02", where 0x0050 corresponds to a frequency of 80Hz, 0x0078 corresponds to a pulse width of 120μs, and 0x0003 corresponds to an amplitude of 3V. After receiving the command, the stimulation module switches to this parameter combination within a few minutes and begins to apply electrical stimulation.

[0041] In this embodiment, the EEG complexity index refers to an indicator that quantifies the complexity of brain neural activity based on the time-series characteristics of EEG signals. Patients with higher levels of consciousness exhibit more complex brain neural activity, and this index value is typically larger, making it one of the core indicators reflecting the level of consciousness. Specifically, mature complexity calculation algorithms (such as Lempel-Ziv complexity, sample entropy, approximate entropy, etc.) are used to analyze and calculate the index from the preprocessed EEG signals. The algorithm is implemented in the signal processing and control module through software programming, and the calculation process is executed in real time to ensure rapid feedback of the index results. For example, using an algorithm based on Lempel-Ziv complexity to calculate this index involves the following steps: First, the preprocessed EEG signal (time series x1, x2, ..., xn) is converted into a binary sequence (by calculating and setting a threshold, signal values ​​greater than the threshold are recorded as 1, and those less than or equal to the threshold are recorded as 0), resulting in the binary sequence s1, s2, ..., sn; then, starting from the beginning of the sequence, subsequences are gradually expanded, and the number of different subsequences is counted; finally, the complexity value is calculated based on the sequence length and the number of different subsequences. In this embodiment, the functional connectivity index between brain regions quantifies the synchronicity or correlation of neural electrical activity between different brain regions. It reflects the efficiency of information transmission between brain regions. Patients with good consciousness have stronger functional connectivity between brain regions, and the index value is usually larger. Specifically, key brain regions related to consciousness regulation (such as the prefrontal cortex, parietal lobe, thalamus, etc.) are selected, and EEG signals of each brain region are extracted. Algorithms such as coherence analysis, phase synchronization analysis, and mutual information analysis are used to calculate the degree of correlation between signals from different brain regions to obtain the functional connectivity index. The calculation of the index is based on the frequency domain or time domain characteristics of the signal to ensure that it can accurately reflect the functional correlation between brain regions. For example, the prefrontal cortex (F3 and F4 channels) and the central thalamus (acquired via scalp electrodes or intracranial electrodes near the DBS electrodes) are selected as target brain regions. A coherence analysis algorithm is used to calculate functional connectivity indices. The specific steps are: performing Fourier transform on the EEG signals of the two brain regions to obtain their respective power spectral densities and cross-power spectral densities; calculating the coherence coefficient (range 0-1), where a larger coherence coefficient indicates stronger synchronization of neural electrical activity between the two brain regions; and combining different frequency bands (e.g., ...) Frequency band 8-13Hz The brain region functional connectivity index is obtained by weighted summation of the coherence coefficients in the frequency band (13-30Hz). The final consciousness index is obtained by weighted summation of the complexity index, functional connectivity index, and other time-frequency indices. This index ranges from 0 to 1; for example, the EEG complexity index for normal conscious volunteers is approximately 0.7-0.9, the index for the unresponsive arousal syndrome group among patients with consciousness disorders is approximately 0.3-0.5, and the index for the minimally conscious state group is 0.5-0.7.

[0042] The beneficial effects of the above technical solution are as follows: through the closed-loop collaborative work of the stimulation module, EEG monitoring module, and signal processing and control module, the automated and personalized optimization of DBS stimulation parameters is achieved. It does not rely on the subjective experience of physicians or general literature parameters, and can capture the neurophysiological response of patients to different stimulation parameters in real time, quickly find the most suitable combination of stimulation parameters for patients, significantly improve the accuracy and effectiveness of DBS treatment, shorten the parameter optimization cycle, and reduce the medical burden. At the same time, through non-invasive EEG monitoring and precise target stimulation, the safety and comfort of treatment are ensured.

[0043] This invention provides a closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness, wherein the signal processing and control module includes: The initialization unit is used to set the parameter space to be optimized and the search step size, and at the same time, to set a set of initial stimulus parameters P_initial; The baseline measurement unit is used to acquire a baseline EEG signal by the EEG monitoring module without applying or with initial parameter stimulation, and to calculate the baseline consciousness index value Index_baseline based on the signal processing and control module. The parameter traversal and stimulation unit is used by the signal processing and control module to automatically generate a new parameter combination P_i in the parameter space according to the search step size, and send instructions to the stimulation module to make the stimulation module perform electrical stimulation several times according to P_i. The synchronous monitoring and calculation unit is used to continuously collect EEG signals during the short window period after stimulation using parameter combination P_i, and to process the continuously collected EEG signals in real time based on the signal processing and control module, and calculate the consciousness index value Index_i corresponding to parameter combination P_i and normalized by the baseline consciousness index value. The data storage and iteration unit is used to record the pairing of parameter combination P_i and corresponding consciousness index value Index_i, and return to the parameter traversal and stimulation unit to select the next untested parameter combination P_i+1 for testing, and repeat the execution operations of the synchronous monitoring and calculation unit and the data storage and iteration unit until all preset parameter combinations have been traversed. The optimal parameter determination unit is used to compare the consciousness index values ​​of all records after the traversal is completed and select the parameter combination P_optimal that produces the maximum consciousness index value. The application and output unit is used to use the found optimal parameter P_optimal as the recommended parameter for the patient's subsequent long-term DBS treatment.

[0044] In this embodiment, the boundary range of the parameter space, the search step size, and the initial stimulation parameters are preset through software programming. The parameter space and search step size are set based on clinical safety data and optimization accuracy requirements. The initial stimulation parameters can be selected from commonly used clinical parameters or personalized initial values ​​determined based on the patient's preoperative assessment data. The initialization information is stored in the module's memory and can be modified through external devices. For example, the parameter space is set to a frequency of 20-100Hz, a pulse width of 40-200μs, and an amplitude of 1-5V, and the search step size is set to a frequency of 10Hz and a pulse width of 20μs. The amplitude is 0.5V, and the initial stimulation parameters P_initial are set to a frequency of 70Hz and a pulse width of 90. The amplitude is 2.5V (a commonly used clinical parameter). The initialization unit automatically reads these settings after the system starts up, providing basic parameters for subsequent optimization processes.

[0045] In this embodiment, the EEG monitoring module is controlled to collect EEG signals within a set time window. The duration of the collection time window is set according to the signal stability requirements (usually 30 seconds to 5 minutes). If initial parameter stimulation is applied during the collection process, synchronization control is used to ensure that the stimulation and collection time are synchronized. The baseline consciousness index value is calculated using the same algorithm as for subsequent parameter stimulation to ensure consistent evaluation criteria. For example, if no electrical stimulation is applied to collect the baseline EEG signal, the collection time is set to 2 minutes. The EEG monitoring module continuously collects at a sampling rate of 256Hz. After the collection is completed, the signal processing and control module preprocesses the baseline EEG signal and uses the Lempel-Ziv complexity algorithm and coherence analysis algorithm to calculate the EEG complexity index and the brain region functional connectivity index, respectively. The two indices are weighted and summed (with weights of 0.6 and 0.4, respectively) to obtain the baseline consciousness index value Index_baselinee. For example, if the calculated Index_baseline=0.45, then...

[0046] In this embodiment, a grid search algorithm is used to generate untested parameter combinations sequentially according to frequency, pulse width, and amplitude. The generation order of parameter combinations can be preset (e.g., first fix the frequency and pulse width, then iterate through the amplitude; then change the pulse width, iterate through the amplitude, and so on). After generating the parameter combinations, they are converted into a standardized instruction format and sent to the stimulation module through the communication interface. At the same time, the tested parameter combinations are recorded to avoid repeated testing. For example, the parameter space is a frequency of 20-100Hz (step size 10Hz) and a pulse width of 40-200Hz. (Step length 20) The amplitude is 1-5V (step size 0.5V). Parameter combinations are generated in the order of "frequency → pulse width → amplitude". The first parameter combination P1 is a frequency of 20Hz and a pulse width of 40Hz. The amplitude is 1V, and the second parameter combination P2 is a frequency of 20Hz and a pulse width of 40. The amplitude is 1.5V... The nth parameter combination Pn is a frequency of 100Hz and a pulse width of 200. With an amplitude of 5V, after generating P_i, it is encoded into a standardized instruction "0x010x00140x00280x00010x02" (corresponding to P1) and sent to the stimulation module. After receiving it, the stimulation module performs electrical stimulation according to the combination of these parameters.

[0047] In this embodiment, the stimulation module and the EEG monitoring module are synchronized by a synchronous trigger signal. The rising edge of the trigger signal corresponds to the start of stimulation. The acquisition window during stimulation is 5 seconds, and the acquisition window during the interval is 10 seconds. After preprocessing, the acquired EEG signals are used to calculate the consciousness index value using the same algorithm as the baseline index calculation. The baseline consciousness index value is then normalized to eliminate the influence of individual baseline differences.

[0048] In this embodiment, flash memory is used as the data storage medium, with the storage format being "parameter combination ID-frequency-pulse width-amplitude-consciousness index value Index_i-collection timestamp". The paired data for each parameter combination is stored independently for easy subsequent querying and comparison. Iterative control is implemented through software logic. The system automatically detects whether there are untested parameter combinations. If so, the next parameter combination is selected, and parameter traversal and stimulation unit execution are triggered. If not, the iteration ends. For example, the parameter combination P_i (frequency 50Hz, pulse width 120Hz) is recorded. The amplitude (3V) and the corresponding pairing data with Index_i=0.5 are stored as "ID001-50Hz-120". After storing "-3V-1.51-20240520103025", the system detects an untested parameter combination P_i+1 (frequency 50Hz, pulse width 120μs, amplitude 3.5V), automatically returns to the parameter traversal and stimulation unit, and triggers the stimulation and data of P_i+1. The principle of the index_i+1 corresponding to P_i+1 is similar to that of the index_i corresponding to P_i, and will not be elaborated here.

[0049] In this embodiment, a bubble sort algorithm is used to sort all stored consciousness indicator values ​​in descending order. The parameter combination corresponding to the first-ranked indicator value is selected as the optimal parameter combination. If multiple parameter combinations correspond to the same maximum indicator value, the combination whose frequency, pulse width, and amplitude are closer to the range of commonly used clinical parameters is selected, or additional secondary evaluation indicators (such as stimulation energy consumption) are used for screening. For example, after all parameter combinations have been traversed, a total of 729 pairs of data were recorded. Through sorting, the parameter combination P_optimal (frequency 80Hz, pulse width 160Hz, amplitude 160Hz) was found to be optimal. The maximum value of the consciousness index corresponding to the amplitude 3.5V is Index_optimal=0.78. The index values ​​of other parameter combinations are all less than this value. Therefore, P_optimal is determined to be the optimal parameter combination.

[0050] In this embodiment, the optimal parameter combination is stored in a dedicated storage area of ​​the signal processing and control module. Simultaneously, the parameters are output to external terminal devices (such as medical monitors or computers) via wireless communication (e.g., Bluetooth, WiFi) or wired communication (e.g., USB). Output formats include text descriptions, numerical codes, and charts, facilitating physician viewing and use. The system can also automatically send the optimal parameters to the stimulation module, allowing it to directly switch to those parameters for long-term treatment. For example, the application and output unit stores the optimal parameter combination P_optimal (frequency 80Hz, pulse width 160...). The amplitude (3.5V) is stored in a dedicated area, and the parameters are simultaneously output to the physician's tablet via Bluetooth. The output is displayed as "Optimal stimulation parameters: frequency 80Hz, pulse width 160". "Amplitude 3.5V, corresponding to a consciousness index value of 0.78." After the physician confirms, the system can automatically control the stimulation module to switch to this parameter combination and begin long-term DBS treatment.

[0051] The beneficial effects of the above technical solution are as follows: by refining the signal processing and control module into seven functionally defined units, the parameter optimization process is automated, standardized, and efficient. From initialization settings to baseline measurement, parameter traversal, synchronous monitoring, data storage, optimal parameter determination, and application output, a complete parameter optimization closed loop is formed. The search and determination of optimal parameters can be completed without manual intervention, significantly shortening the parameter optimization cycle and improving optimization accuracy. At the same time, the collaborative work of each unit ensures the stability and reliability of the optimization process, providing patients with personalized and precise DBS treatment parameters.

[0052] This invention provides a closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness, wherein the electroencephalogram (EEG) monitoring module includes: The dynamic gain preamplifier unit is used to dynamically configure the amplification factor based on the real-time peak voltage of the latest EEG signal, employing a low-noise instrumentation amplifier. The pre-amplified signal is obtained, where, This represents the absolute value of the peak voltage of the EEG signal. The signal-to-noise ratio adaptive interference suppression unit receives the pre-amplified signal and combines it with the interference reference signal synchronously output by the stimulation module. It then employs a dynamic notch filtering algorithm based on real-time signal-to-noise ratio to obtain the interference-free EEG signal. The iterative update formula for the filter weights is as follows: ,in, The notch filter weights are for the (k+1)th iteration. The weight for the k-th iteration is... For adaptive step size, Let be the real-time signal-to-noise ratio of the signal at time k. The signal is the pre-amplified EEG signal at time k. The interference reference signal at time k is based on Feedback adjusts the weight update rate; In this embodiment, for : Window function selection: Hanning window, window length = 1024 sampling points (corresponding to 4 seconds, sampling rate 256Hz).

[0053] Frequency range division: signal frequency range 0-128Hz, noise frequency range limited to 0-4Hz (baseline noise) and 50Hz±2Hz (power frequency interference).

[0054] Calculation steps: Perform a Fourier transform on the EEG signal in the window at time k to obtain the power spectral density P(f); signal power The key frequency band is 8-80Hz; Noise power ; Signal-to-noise ratio .

[0055] The artifact removal unit is used to construct the covariance matrix manifold of the multi-channel EEG signal and calculate the Riemann distance between each signal component in the de-interference EEG signal and the pure EEG reference covariance matrix to identify artifacts. The Riemann distance formula is: ,in, Let i be the covariance matrix of the i-th signal component. Covariance matrix of pure EEG benchmark The Riemann distance; For matrix trace operations; This refers to matrix determinant operations. Number of EEG acquisition channels; when When the corresponding component is identified as an artifact, the artifact component is removed from the de-interference EEG signal to obtain a clean EEG signal. This is a preset threshold.

[0056] In this embodiment, the pure EEG benchmark covariance matrix Construction: Resting-state EEG signals were collected from 50 healthy volunteers (aged 20-60 years), with each sample collected for 5 minutes at a sampling rate of 256 Hz. Remove artifacts (electromyography, electrooculography) from each signal and calculate the single-case covariance matrix; The mean of the covariance matrices of 50 single cases is obtained. For example, a dimension of 16×16 corresponds to 16 channels of acquisition.

[0057] ROC curve analysis was used to determine the EEG signals (including known artifact fragments) of 20 patients with impaired consciousness. When the value is 0.8, the artifact detection sensitivity is ≥90% and the specificity is ≥88%.

[0058] In this embodiment, 40 is the decibel value corresponding to the peak voltage of the target output signal (i.e., the target output peak voltage is 10). ,20lg(10 / 1 =20dB, with a margin of 40dB for subsequent processing.

[0059] In this embodiment, a programmable notch filter chip or dynamic notch filtering is implemented through digital signal processing (DSP). The interference reference signal is synchronously output by the stimulation module (reflecting the frequency and phase characteristics of the stimulation signal). The real-time signal-to-noise ratio (SNR) is obtained by calculating the ratio of signal power to noise power. The filter weights are adjusted in real time according to a preset iterative update formula. The weight update rate is determined by the real-time SNR; the lower the SNR, the faster the weight update rate, in order to quickly adapt to changes in interference. The initial value is set to 0.1, and the adaptive step size ranges from 0.001 to 0.01, depending on the system stability. The value range is 0-20dB. The amplitude range is 0-1, and the signal is standardized.

[0060] In this embodiment, the rule for determining the adaptive step size is as follows: when When the signal-to-noise ratio is <5dB (low signal-to-noise ratio), the value is 0.01 (fast weight update). When 5dB≤ When the signal-to-noise ratio is ≤15dB (medium signal-to-noise ratio), the value is 0.005 (equalization update). when When the signal-to-noise ratio is >15dB (high signal-to-noise ratio), the value is 0.001 (slow update to maintain stability).

[0061] Interference reference signal ( ): Output synchronously by the stimulation module, with the same frequency and phase as the electrical stimulation signal, and an amplitude of 1 / 1000 of the stimulation amplitude (unit: V).

[0062] In this embodiment, the interference-free EEG signal is decomposed into multiple channels (e.g., using principal component analysis, PCA) to obtain multiple signal components, and the covariance matrix of each signal component is calculated.

[0063] The beneficial effects of the above technical solution are as follows: Through detailed design of the three core units of the EEG monitoring module, high-precision and high-fidelity acquisition and processing of EEG signals are achieved. The dynamic gain pre-amplification unit ensures effective amplification of EEG signals of different amplitudes, avoiding saturation and insufficient amplitude. The signal-to-noise ratio adaptive interference suppression unit can accurately suppress external interference such as stimulus interference and power frequency interference. The artifact removal unit effectively removes physiological artifacts and environmental artifacts. The three units work together to significantly improve the quality of EEG signals, providing a reliable data foundation for the accurate calculation of subsequent neurophysiological indicators and parameter optimization, and ensuring that the system can accurately capture the patient's neurophysiological response.

[0064] This invention provides a closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness, wherein the electroencephalogram (EEG) monitoring module further includes: The historical signal correction unit is used to retrieve n pairs of pre-historical signals output by the EEG monitoring module and post-historical signals received by the signal processing and control module from the historical database, and to lock the change segment, occurrence time, and change amplitude of each pair of signals to generate a change vector. Calculate the transmission correction factor: ,in, Let the magnitude of the change vector of the j-th pair of signals be denoted as 'j'. This is the reference amplitude of the EEG signal; use Correct transmission distortion of pure EEG signals; The multi-channel differential processing unit is used to perform multi-channel differential processing on the corrected signal according to the acquisition channel to obtain the first signal. Combined with EEG signal sampling frequency The first signal is calibrated to obtain the second signal. ; The feature fusion unit extracts the signal features of each second signal, constructs a two-dimensional feature-attribute array according to the differential attributes, and parses the corresponding array to determine the weight of each second signal. ,in, Let V be the amplitude variance of the k-th second signal; Let be the amplitude variance mass charge of all second signals; based on The latest EEG signal is obtained by weighted fusion of all the second signals; A sliding baseline dynamic normalization unit is used to normalize the latest EEG signal to obtain a normalized analog signal, wherein the normalization formula is: ,in, The normalized EEG signal at time t; The signal after feature fusion at time t; Let be the moving average of the baseline signal m1 seconds before time t; Let m1 seconds before time t be the sliding standard deviation of the baseline signal. The analog-to-digital converter unit is used to convert standardized analog signals into standardized digital signals and transmit them to the signal processing and control module.

[0065] In this embodiment, typically n=50-100 sets of recent data are selected, and the change vector is the difference between the later historical signal and the earlier historical signal, with units of... , It is usually set to 100. The amplitude was determined based on the typical amplitude of clinical EEG signals.

[0066] In this embodiment, a bipolar differential processing method is used to perform differential operations on the signals of adjacent channels (such as differential between channel 1 and channel 2, differential between channel 3 and channel 4, etc.) to obtain the first signal. Differential processing can effectively suppress common-mode interference; amplitude calibration is performed based on the ratio of the EEG signal's sampling frequency to the standard sampling frequency (set to 256Hz) to obtain the second signal. This ensures that the amplitudes of signals collected at different sampling frequencies are comparable.

[0067] In this embodiment, the rows in the feature-attribute two-dimensional array represent the second signal index, and the columns represent features such as amplitude variance, mean, and frequency, as shown in Table 1: Table 1 Two-dimensional array of features and attributes In this embodiment, the sliding window duration m1 is set (usually 1-5 seconds), and the moving average of the baseline signal within m1 seconds before the current time t is calculated in real time. With sliding standard deviation The current signal value is converted into a standardized signal using a standardization formula. The amplitude range of the standardized signal is usually [-3, 3], which facilitates subsequent processing and analysis.

[0068] In this embodiment, a 24-bit ADC chip with a sampling rate of 256Hz and a standardized analog signal are selected. =1.2375 (amplitude range [-3, 3]), when converted to a digital signal, the analog signal amplitude is mapped to the range of 24-bit binary two's complement (-8388608 to 8388607), and the mapping formula is as follows. Substituting the values ​​into the calculation, we get 3460281, which is then converted into 24-bit binary two's complement as "001101001011000110101101", and transmitted to the signal processing and control module via the SPI interface.

[0069] The beneficial effects of the above technical solution are as follows: by adding five functional units, the signal processing flow of the EEG monitoring module is further improved; the historical signal correction unit compensates for transmission distortion and improves signal accuracy; the multi-channel differential processing and feature fusion unit integrates the effective information of multi-channel signals, enhancing signal reliability and information content; the sliding baseline dynamic standardization unit eliminates baseline drift and individual differences, making signals more comparable; and the analog-to-digital conversion unit realizes high-precision conversion of analog signals to digital signals, providing high-quality digital signals for subsequent processing by the signal processing and control module, ensuring the integrity and accuracy of data transmission and processing throughout the system.

[0070] This invention provides a closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness, wherein the signal processing and control module includes: The signal enhancement unit is used to construct a dynamic threshold based on the real-time variance of the signal and to adaptively smooth and correct abnormal noise points in the standardized digital signal. Determine the quality assessment factor of the smoothed signal. When the quality assessment factor is lower than the preset value, start the enhanced FastICA algorithm, dynamically adjust the iteration step size to separate residual artifacts, and retain high-quality effective signal components. The reconstruction unit is used to decompose the preserved signal into multiple scales based on wavelet packet decomposition, and to reconstruct the key frequency band signals related to the state of consciousness after energy enhancement. The feature extraction unit is used to extract features from the reconstructed signal and normalize all extracted features to construct a consciousness state feature set, wherein the consciousness state feature set includes: complexity features, frequency band energy features and brain network connectivity features. The indicator determination unit is used to dynamically adjust the weights of each feature by combining the entropy weight method with the patient's preoperative personalized baseline parameters, and obtain the quantitative indicator of consciousness state by weighted summation.

[0071] In this embodiment, for example, the signal quality evaluation factor is set to be obtained by weighting the signal-to-noise ratio (weight 0.5), kurtosis (weight 0.3), and skewness (weight 0.2), and the evaluation threshold is 0.8 (after standardization); the variance of the standardized digital signal is calculated in real time, and the variance is 0.8. The dynamic threshold is set to ±3 × variance = ±2.4. Points in the signal that exceed ±2.4 are identified as abnormal noise points and corrected using 3-point median filtering; the signal-to-noise ratio of the corrected signal is calculated to be 12dB (0.6 after standardization), and the kurtosis is calculated to be... =4.5 (0.75 after standardization), skewness = 0.8 (0.8 after standardization), quality assessment factor = 0.6×0.5+0.75×0.3+0.8×0.2=0.3+0.225+0.16=0.685<0.8, start the enhanced FastICA algorithm, set the initial iteration step size = 0.1, and dynamically adjust the step size according to the signal separation effect (increase the step size if the separation effect is poor, and decrease the step size if the effect is good), separate out and remove the electromyographic artifact components, and retain high-quality effective signals.

[0072] In this embodiment, a suitable wavelet basis function (such as the db4 wavelet) is selected, and the number of wavelet packet decomposition layers is set (usually 3-5 layers, determined according to the frequency range of the EEG signal). The preserved signal is decomposed into multiple low-frequency and high-frequency sub-band signals. By analyzing the correlation between signals in different frequency bands and the state of consciousness, key frequency bands (such as...) are determined. Frequency band 8-13Hz Frequency band 13-30Hz The frequency band is 30-80Hz, which is closely related to brain activity. The sub-band signals of the key frequency band are enhanced (e.g., multiplied by an enhancement factor of 1.2-1.5), while the energy of the sub-band signals of non-key frequency bands remains unchanged. Then, all sub-band signals are reconstructed into a complete signal through wavelet packet inverse transform. For example, using the db4 wavelet as the wavelet basis function, with a decomposition layer of 4, the preserved signal is decomposed into 16 sub-band signals, corresponding to frequency ranges of 0-8Hz, 8-16Hz, 16-24Hz, 24-32Hz…120-128Hz. Clinical data validation confirms that 8-16Hz… 16-32Hz (frequency band) 32-64Hz (frequency band) The three frequency bands are the key frequency bands. The sub-band signals corresponding to these three frequency bands are multiplied by an enhancement factor of 1.3, and the enhancement factor of the sub-band signals of other frequency bands is set to 1.0. The 16 sub-band signals are reconstructed by wavelet packet inverse transform to obtain the energy-enhanced signal.

[0073] In this embodiment, the complexity features include five parameters: normalized sample entropy (F1), approximate entropy (F2), peak value (F3), variance (F4), and kurtosis (F5); the frequency band energy features include normalized sample entropy (F1), approximate entropy (F2), peak value (F3), variance (F4), and kurtosis (F5). Frequency band (0-4Hz) energy (F6), Frequency band (4-8Hz) energy (F7) Frequency band (8-13Hz) energy (F8), Frequency band (13-30Hz) energy (F9), Frequency band (30-80Hz) energy (F10) The peak power spectral density (F11) of the frequency band, consisting of 6 parameters; brain network connectivity features, including the normalized coherence coefficients of the prefrontal and parietal lobes (F12), the prefrontal and thalamic lobes (F13), the parietal and thalamic lobes (F14), and the whole-brain average phase synchronization (F15), consisting of 4 parameters, yielded the following set of consciousness state features: {F1=0.6,F2=0.55,F3=0.7,F4=0.65,F5=0.4,F6=0.3,F7=0.35,F8=0.8,F9=0.75,F10=0.6,F11=0.7,F12=0.6,F13=0.7,F14=0.65,F15=0.5}.

[0074] In this embodiment, the objective weight of each feature is first calculated using the entropy weight method (based on the information entropy of the feature; the smaller the information entropy, the greater the weight, indicating that the feature has a stronger ability to distinguish the state of consciousness). Personalized baseline parameters of the patient before surgery (such as age, etiology, course of disease, brain injury site, preoperative CRS-R score, etc.) are collected. A correlation model between the baseline parameters and feature weights is established through multiple regression analysis. The weights of each feature are dynamically adjusted according to the patient's specific baseline parameters. The adjusted feature weights are weighted and summed with the normalized feature values ​​to obtain a quantitative index of the state of consciousness. For example, the initial weights of 15 features are calculated using the entropy weight method: {W1=0.08,W2=0.07,W3=0.06,W4=0.05,W5=0.04,W6=0.03,W7=0.04,W8=0.12,W9=0.11,W10=0.09,W11=0.10,W12=0.07,W13=0.08,W14=0.06,W15=0.04}; The patient's preoperative personalized baseline parameters were: age 45 years, etiology: traumatic brain injury, duration of illness: 6 months, brain injury site: right frontoparietal lobe, and preoperative CRS-R score: 10. The adjustment coefficients for each feature were calculated using an association model: {K1=1.0,K2=1.0,K3=1.1,K4=1.0,K5=0.9,K6=0.9,K7=1.0,K8=1.2,K9=1.1,K10=1.0,K11=1.1,K12=1.0,K13=1.0,K14=1.0,K15=1.0}. The adjusted weights were Wi'=Wi×Ki. The normalized feature values ​​and the adjusted weights were then weighted and summed to obtain the quantitative index of consciousness state.

[0075] The beneficial effects of the above technical solution are as follows: through the coordinated work of four units—signal enhancement, reconstruction, feature extraction, and index determination—a precise transformation from standardized digital signals to quantitative indicators of consciousness state is achieved. The signal enhancement and reconstruction units effectively improve signal quality and the significance of effective features, avoiding interference from noise and artifacts in subsequent analysis. The feature extraction unit comprehensively covers physiological features related to consciousness state, ensuring the completeness and relevance of features. The index determination unit assigns objective weights to features using the entropy weighting method, combined with dynamic adjustments to the patient's preoperative personalized baseline parameters, enabling the quantitative indicators to accurately reflect the individual patient's level of consciousness. This provides a scientific and reliable evaluation basis for stimulus parameter optimization, ensuring the accuracy of the optimization direction.

[0076] This invention provides a method for optimizing closed-loop deep brain stimulation parameters for patients with impaired consciousness, such as... Figure 2 As shown, it includes: Step 1: Receive standardized instructions from the signal processing and control module, and apply electrical stimulation to the target point in the patient's brain according to the preset stimulation parameters in the standardized instructions; Step 2: During or between electrical stimulation sessions, EEG signals from the patient's scalp are collected in real time, and after signal conditioning, standardized digital signals are transmitted to the signal processing and control module. Step 3: Receive the standardized digital signal from the EEG monitoring module and perform preprocessing. Calculate the neurophysiological indicators of the quantified state of consciousness based on a preset algorithm to achieve automatic optimization of stimulation parameters and output parameter adjustment instructions to the stimulation module.

[0077] The beneficial effects of the above technical solution are as follows: through the closed-loop collaborative work of the stimulation module, EEG monitoring module, and signal processing and control module, the automated and personalized optimization of DBS stimulation parameters is achieved. It does not rely on the subjective experience of physicians or general literature parameters, and can capture the neurophysiological response of patients to different stimulation parameters in real time, quickly find the most suitable combination of stimulation parameters for patients, significantly improve the accuracy and effectiveness of DBS treatment, shorten the parameter optimization cycle, and reduce the medical burden. At the same time, through non-invasive EEG monitoring and precise target stimulation, the safety and comfort of treatment are ensured.

[0078] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness, characterized in that, include: The stimulation module is an implantable medical stimulation device used to receive standardized instructions from the signal processing and control module and apply electrical stimulation to the target points in the patient's brain according to the preset stimulation parameters in the standardized instructions. The EEG monitoring module is a non-invasive medical monitoring device used to collect EEG signals from the patient's scalp in real time during or between electrical stimulation sessions. After signal conditioning, the signals are transmitted to the signal processing and control module as standardized digital signals. The signal processing and control module is used to receive standardized digital signals from the EEG monitoring module and perform preprocessing, calculate and quantify the neurophysiological indicators of the state of consciousness based on a preset algorithm, realize automatic optimization of stimulation parameters, and output parameter adjustment instructions to the stimulation module.

2. The closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness according to claim 1, characterized in that, The stimulation parameters include: frequency, pulse width, and amplitude.

3. The closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness according to claim 1, characterized in that, The implantable medical stimulation device includes: an implantable pulse generator, an extension lead, and a DBS electrode implanted in the brain at a target site, the target site being the central thalamus.

4. The closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness according to claim 1, characterized in that, The neurophysiological indicators of the state of consciousness include one or more of the following: brain electrical complexity indicators and brain inter-brain functional connectivity indicators.

5. The closed-loop deep brain stimulation parameter optimization system for patients with impaired consciousness according to claim 1, characterized in that, The signal processing and control module includes: The initialization unit is used to set the parameter space to be optimized and the search step size, and at the same time, to set a set of initial stimulus parameters P_initial; The baseline measurement unit is used to acquire a baseline EEG signal by the EEG monitoring module without applying or with initial parameter stimulation, and to calculate the baseline consciousness index value Index_baseline based on the signal processing and control module. The parameter traversal and stimulation unit is used by the signal processing and control module to automatically generate a new parameter combination P_i in the parameter space according to the search step size, and send instructions to the stimulation module to make the stimulation module perform electrical stimulation several times according to P_i. The synchronous monitoring and calculation unit is used to continuously collect EEG signals during the short window period after stimulation using parameter combination P_i, and to process the continuously collected EEG signals in real time based on the signal processing and control module, and calculate the consciousness index value Index_i corresponding to parameter combination P_i and normalized by the baseline consciousness index value. The data storage and iteration unit is used to record the pairing of parameter combination P_i and corresponding consciousness index value Index_i, and return to the parameter traversal and stimulation unit to select the next untested parameter combination P_i+1 for testing, and repeat the execution operations of the synchronous monitoring and calculation unit and the data storage and iteration unit until all preset parameter combinations have been traversed. The optimal parameter determination unit is used to compare the consciousness index values ​​of all records after the traversal is completed and select the parameter combination P_optimal that produces the maximum consciousness index value. The application and output unit is used to use the found optimal parameter P_optimal as the recommended parameter for the patient's subsequent long-term DBS treatment.

6. A method for optimizing closed-loop deep brain stimulation parameters for patients with impaired consciousness, characterized in that, include: Step 1: Receive standardized instructions from the signal processing and control module, and apply electrical stimulation to the target point in the patient's brain according to the preset stimulation parameters in the standardized instructions; Step 2: During or between electrical stimulation sessions, EEG signals from the patient's scalp are collected in real time, and after signal conditioning, standardized digital signals are transmitted to the signal processing and control module. Step 3: Receive the standardized digital signal from the EEG monitoring module and perform preprocessing. Calculate the neurophysiological indicators of the quantified state of consciousness based on a preset algorithm to achieve automatic optimization of stimulation parameters and output parameter adjustment instructions to the stimulation module.

Images