Implantable closed-loop nerve stimulation system and control method thereof

Abstract

The invention provides an implantable closed-loop nerve stimulation system and a control method thereof.The implantable closed-loop nerve stimulation system comprises a local real-time control device, a local optimization control device and a cloud server which are connected in sequence, and a patient feedback device connected with the local optimization control device, the local real-time control device is implanted into the brain and is configured to calculate the signal confidence coefficient of the electroencephalogram signal in the first time window to detect an abnormal electroencephalogram signal, and adjust the length of the first time window based on the first instruction; the local optimization control device is an extracranial portable device configured to: process a patient feedback event within a second time window greater than the first time window, determine a first control policy based on at least one of an electrode impedance change rate, a signal confidence, and the patient feedback event; and the cloud server is configured to analyze the first data in a third time window greater than the second time window and generate an alarm message and / or a second control strategy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of brain-computer interface regulation, in particular, the present application relates to an implantable closed-loop neural stimulation system and a control method thereof. BACKGROUND

[0002] Epilepsy is a nervous system disease, which usually needs long-term neural regulation treatment. Since the implantable closed-loop neural stimulation system can respond in real time according to the brain electrical activity of the patient, it has been widely used. The existing implantable device usually triggers electrical stimulation with fixed parameters after detecting preset features (such as high-frequency oscillation, sharp wave), and the abnormal detection threshold and response strategy are usually manually set by doctors in outpatient service, which lacks dynamic adaptation ability.

[0003] However, clinical practice shows that the brain electrical characteristics, seizure patterns and subjective feelings of patients have significant individual differences and time-varying. For example, the abnormal brain electrical signals of the same patient may change significantly in different physiological cycles (such as menstrual period, sleep stage) or disease progression stages. If the system cannot adaptively adjust the stimulation strategy, it is easy to cause missed detection or mis-stimulation, and then affect the efficacy or cause discomfort. In addition, the subjective feedback (such as "uncomfortable", "effective") of the patient is usually only used as reference information for doctors to adjust the stimulation parameters, and the existing system does not consider the subjective feedback of the patient to adaptively adjust the stimulation strategy.

[0004] Therefore, there is an urgent need for an implantable closed-loop neural stimulation system that can realize personalized and adaptive closed-loop neural regulation. SUMMARY

[0005] The present application proposes an implantable closed-loop neural stimulation system and a control method thereof to overcome the shortcomings of the prior art. Based on a system composed of a local real-time control device, a local optimization control device and a cloud server, combined with signal confidence, electrode impedance change rate and patient feedback events, the system can realize efficient cooperation of multiple time granularities, adaptive switching of working modes, personalized comfort adjustment of stimulation parameters, while ensuring the safety and timeliness of the control strategy.

[0006] In a first aspect, the embodiments of the present application provide an implantable closed-loop neural stimulation system, comprising: a local real-time control device, a local optimization control device and a cloud server connected in sequence, and a patient feedback device connected with the local optimization control device, wherein, The local real-time control device is implanted in the intracranial and is configured to: calculate the signal confidence of the collected brain electrical signals in multiple channels in real time within a first time window to determine whether there is an abnormal first brain electrical signal, and trigger pulse stimulation when it is determined that there is an abnormal first brain electrical signal; periodically calculate an electrode impedance change rate, and send the signal confidence and the electrode impedance change rate to the local optimization control device, and adjust the length of the first time window based on the first instruction sent by the local optimization control device; The local optimization control device is an extracranial portable device, configured to: process a trigger signal from the patient feedback device to obtain a patient feedback event within a second time window greater than the first time window, and determine a first control strategy based on at least one of the electrode impedance change rate, the signal confidence and the patient feedback event, the first control strategy including: whether to switch the current working mode, and sending the first instruction to the local real-time control device when it is determined to switch the current working mode; The cloud server is configured to: perform analysis on first data within a third time window greater than the second time window and generate an alarm message and / or a second control strategy, wherein the first data includes: physiological data and treatment-related data of the patient within a first time unit, and a set of patient feedback-related data.

[0007] In a second aspect, the embodiments of the present application provide a control method, which is applied to the implantable closed-loop neural stimulation system as described in the first aspect, and the method comprises: The local real-time control device implanted in the intracranial calculates the signal confidence of the collected brain electrical signals in multiple channels in real time within a first time window to determine whether there is an abnormal first brain electrical signal, and triggers pulse stimulation when it is determined that there is an abnormal first brain electrical signal; periodically calculates an electrode impedance change rate, and sends the signal confidence and the electrode impedance change rate to the local optimization control device, and adjusts the length of the first time window based on the first instruction sent by the local optimization control device; The local optimization control device is an extracranial portable device, configured to: process a trigger signal from the patient feedback device to obtain a patient feedback event within a second time window greater than the first time window, and determine a first control strategy based on at least one of the electrode impedance change rate, the signal confidence and the patient feedback event, the first control strategy including: whether to switch the current working mode, and sending the first instruction to the local real-time control device when it is determined to switch the current working mode; The cloud server analyzes the first data and generates alarm messages and / or a second control strategy within a third time window that is longer than the second time window. The first data includes: the patient's physiological data and treatment-related data within the first time unit, as well as a collection of patient feedback-related data.

[0008] The beneficial technical effects of the technical solutions provided in this application include: The solution in this application is based on a system consisting of a local real-time control device, a local optimized control device, and a cloud server. By combining signal confidence, electrode impedance change rate, and patient feedback events, it achieves efficient coordination of multiple time granularities, adaptive switching of working modes, and personalized comfort adjustment of stimulation parameters, while ensuring the safety and timeliness of the control strategy.

[0009] Additional aspects and advantages of this application will be set forth in part in the description which follows, and will become apparent from the description or may be learned by practice of this application. Attached Figure Description

[0010] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a schematic diagram of the structure of an implantable closed-loop neurostimulation system provided in an embodiment of this application; Figure 2 A schematic diagram of another implantable closed-loop neurostimulation system provided in an embodiment of this application; Figure 3 This is a flowchart illustrating a control method for an implantable closed-loop neurostimulation system provided in an embodiment of this application. Detailed Implementation The embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the embodiments described below with reference to the accompanying drawings are exemplary descriptions for explaining the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions of the embodiments of this application.

[0011] Those skilled in the art will understand that, unless specifically stated otherwise, the terms "described" and "the" as used herein may also include plural forms. It should be further understood that the term "comprising" as used in this application's specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude implementations of other features, information, data, steps, operations, elements, components, and / or combinations thereof supported by this art. It should be understood that when we say an element is "connected" or "coupled" to another element, the element may be directly connected or coupled to the other element, or it may mean that the element and the other element are connected through an intermediate element. Furthermore, "connected" or "coupled" as used herein may include wireless connections or wireless coupling. The term "and / or" as used herein refers to at least one of the items defined by the term; for example, "A and / or B" may be implemented as "A," or as "B," or as "A and B."

[0012] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0013] Existing implantable closed-loop neurostimulation systems typically use fixed thresholds to detect EEG abnormalities and trigger electrical stimulation with preset parameters. This strategy is manually set by physicians and lacks adaptability to individual patient differences and dynamic changes in the patient's condition. In clinical practice, the quality of a patient's EEG signal, electrode status, and subjective feelings fluctuate significantly with physiological cycles and disease stages, leading to potential missed detections (in low-sensitivity modes) or false stimulation (in high-sensitivity modes). Furthermore, patient feedback is mostly used for offline parameter tuning and fails to be integrated with real-time physiological data to drive local intelligent decision-making; cloud-based analysis results are also difficult to safely and efficiently provide closed-loop feedback to the device. Therefore, the existing architecture lacks a multi-timescale collaborative mechanism for millisecond-level response, second-level feedback, and daily / weekly optimization, resulting in a fragmented perception-decision-execution chain and preventing truly personalized, safe, and reliable adaptive neuromodulation.

[0014] In view of the shortcomings of the prior art, the embodiments of this application aim to solve the following technical problems: How to build a closed-loop architecture that supports efficient collaboration across multiple time granularities, enabling millisecond-level real-time stimulation, second-level local feedback processing, and daily / weekly cloud-based strategy optimization to work together organically; How to achieve dynamic switching of working modes based on objective physiological signals (such as signal confidence and electrode impedance change rate) to avoid missed detection or false stimulation caused by fixed thresholds; How to incorporate patients' subjective feedback events (such as "discomfort" and "prodromal symptoms") into the local decision-making logic to enable the system to have "humanized" comfort adjustment capabilities; How can we ensure that cloud-generated control strategies can be safely and reliably deployed to the local optimization layer without compromising the security and timeliness of the underlying real-time stimuli?

[0015] The technical solution of this application and how it solves the above-mentioned technical problems are described in detail below with specific embodiments. It should be noted that the following embodiments can be referenced, borrowed, or combined with each other, and the same terms, similar features, and similar implementation steps in different embodiments will not be described again.

[0016] The implantable closed-loop neurostimulation system and its control method proposed in this application will be described in detail below with reference to the accompanying drawings.

[0017] In some embodiments, an implantable closed-loop neurostimulation system is provided, such as Figure 1 As shown, the implantable closed-loop neurostimulation system 10 includes: a local real-time control device 11, a local optimization control device 12, and a cloud server 13, which are connected in sequence, and a patient feedback device 14 connected to the local optimization control device 12. Specifically, the local real-time control device 11 is implanted intracranially, and the local optimization control device 12 is an extracranial portable device.

[0018] In some embodiments, the local real-time control device 11 is configured to: calculate the signal confidence of the acquired EEG signals from multiple channels in real time within a first time window to determine whether an abnormal first EEG signal exists, and trigger pulse stimulation when an abnormal first EEG signal is determined to exist. The local real-time control device 11 is also configured to: periodically calculate the rate of change of electrode impedance, send the signal confidence and the rate of change of electrode impedance to a local optimization control device 12, and adjust the length of the first time window based on a first instruction sent by the local optimization control device 12.

[0019] In some embodiments, the signal confidence level is calculated by a local real-time control device based on the signal-to-noise ratio of the EEG signal, multiple channel consistency parameters, and parameters characterizing physiological artifacts.

[0020] Optionally, the signal confidence level (SCI) can be calculated using the following formula: SCI = ω1×SNR + ω2 × α + ω3 × (1-β).

[0021] Wherein, SNR is the signal-to-noise ratio, α is the consistency parameter of multiple channels, β is the parameter characterizing physiological artifacts, and ω1, ω2, and ω3 are the weights corresponding to SNR, α, and β, respectively. Their values ​​can be adjusted according to specific application scenarios, and ω1 + ω2 + ω3 = 1.

[0022] Optionally, SNR = Ps / Pn, where Ps is the signal power and Pn is the noise power. In practical applications, EEG signals can be preprocessed, such as through filtering, to more accurately separate the signal from the noise.

[0023] Optionally, multiple channel consistency parameters are used to characterize the consistency of multi-channel signals and can be determined in various ways, such as using the Pearson correlation coefficient or the consistency index. This embodiment uses the Pearson correlation coefficient as an example to illustrate the process of determining multiple channel consistency parameters. Assuming there are two channels, signals X and Y, their correlation coefficient rXY can be calculated using the following formula: .

[0024] in, and Since rXY is the average value of signal X and signal Y respectively, it can be used as the consistency parameter of the two channels.

[0025] Optionally, for cases with more than two channels, the average correlation coefficient among all channels can be calculated as a consistency parameter for multiple channels.

[0026] Optionally, the presence of known physiological artifacts (such as eye movements, electromyography, etc.) can be determined using machine learning algorithms (such as Support Vector Machine (SVM), Random Forest (RF), etc.) to detect artifacts and assign corresponding scores. The scores range from 0 to 1, where 0 indicates no artifacts were detected and 1 indicates significant artifacts are present.

[0027] Optionally, the local real-time control device 11 can continuously extract features from EEG signals across multiple channels. These features include signal-to-noise ratio (SNR), inter-channel Pearson correlation coefficient, and eye-tracking / electromyography artifact model matching score. Then, the signal confidence index (SCI) is calculated in real-time using a weighted fusion formula, with a value ranging from 0 to 100. When SCI ≥ 80, the system determines that a high-confidence abnormal EEG event exists and triggers the stimulation process; when 60 ≤ SCI < 80, it is marked as a low-confidence event and only logged; when SCI < 60, it is considered noise or an artifact.

[0028] In some embodiments, periodically calculating the rate of change of electrode impedance may include: measuring once after each pulse stimulation (e.g., every 5–60 seconds, depending on the attack frequency), or, when there is no pulse stimulation, measuring by injecting a small test current at fixed intervals (e.g., every 100 ms). After the pulse plateau has stabilized (avoiding transients), the inter-electrode voltage V(t) (e.g., acquired via a high-precision ADC) and the output current I(t) (e.g., acquired via an on-chip sampling resistor and transimpedance amplifier) ​​are sampled simultaneously, and the impedance Z(t) can be calculated using the following formula: Z(t) = V(t) / I(t). The rate of change of impedance dZ / dt can be calculated using the following formula:

[0029] in, (That is, the impedance is updated every 100 ms), and the unit is Ω / s.

[0030] In some embodiments, the local optimization control device 12 is configured to process trigger signals from the patient feedback device to obtain patient feedback events within a second time window that is greater than the first time window.

[0031] Optionally, patient feedback events include the timestamp and event type corresponding to the trigger signal.

[0032] Optionally, the types of patient feedback events include, but are not limited to, "aura or discomfort" events, "effective or comfortable" events, etc.

[0033] In the above embodiments, the local optimization control device 12 is further configured to determine a first control strategy based on at least one of electrode impedance change rate, signal confidence, and patient feedback events.

[0034] Optionally, the first control strategy includes: whether to switch the current operating mode, and when it is determined that the current operating mode should be switched, sending a first instruction to the local real-time control device 11. Optionally, the first instruction is used to instruct the local real-time control device 11 to switch the current operating mode to achieve adjustment of the length of the first time window.

[0035] In some embodiments, the operating modes include: normal monitoring mode, high sensitivity mode, cautious mode, feedback-driven mode, and safety pause mode.

[0036] Optionally, the high-sensitivity mode switching trigger condition is: the signal confidence level is greater than or equal to 85, and it lasts for a fourth preset duration. The fourth preset duration can be, for example, 5 seconds, but is not limited to this.

[0037] Optionally, the trigger condition for switching to cautious mode is: the signal confidence level is less than or equal to 35, and no electrode impedance change rate is continuously exceeded by a preset threshold for a second preset duration. The second preset duration can be, for example, 3 seconds, but is not limited to this.

[0038] Optionally, the switching condition for the feedback-driven mode is: the presence of a patient feedback event of any event type.

[0039] Optionally, the trigger condition for switching to the safety pause mode is: within a fifth preset time period, the rate of change in electrode impedance is continuously detected to exceed a preset threshold, and the event type of the patient feedback event is the first event type. The fifth preset time period can be, for example, 5 seconds, but is not limited to this. The patient feedback event can occur within 5 seconds before or after the impedance abnormality.

[0040] Optionally, different operating modes correspond to different lengths of the first time window, and / or different lengths of the second time window.

[0041] In some embodiments, the working mode can be defined as follows: 1. Standard monitoring mode (can be referred to as S0 mode). In S0 mode, the first time window = 200 ms (default), and the second time window = 2s. Stimulation function can be enabled in S0 mode.

[0042] 2. High-sensitivity mode (can be denoted as S1 mode). In S1 mode, the first time window = 80 ms, and the second time window = 500 ms. The trigger condition for switching to S1 is: SCI ≥ 85. S1 mode is used to capture weak but high-confidence attacks.

[0043] 3. Cautionary Mode (S2 mode): In S2 mode, the first time window is 400 ms (desensitization), and the second time window is 200 ms (initiating artifact analysis). The trigger condition for switching to S2 is: SCI ≤ 35 and no impedance anomaly (i.e., no electrode impedance change rate continuously exceeding the preset threshold for the second preset duration is detected). S2 mode is used to avoid false stimulation.

[0044] 4. Feedback-driven mode (can be referred to as S3 mode). In S3 mode, the first time window = maintain the current value (do not force change), and the second time window = 200 ms (prioritize processing patient feedback). The trigger condition for switching to S3 is: receiving any patient feedback event (regardless of whether the SCI exceeds the abnormal threshold).

[0045] 5. Safety Pause Mode (can be referred to as S4 mode): In S4 mode, the output of pulse stimulation is disabled within the first time window, and the second time window is 100 ms (for emergency reporting to the cloud). The trigger condition for switching to S4 is: detecting an impedance abnormality (e.g., dZ / dt>2Ω / s) and receiving a "premonition or discomfort" event.

[0046] In some embodiments, switching working modes may include the following requests: If SCI ≥ 85 and lasts for 5 seconds, switch from S0 mode to S1 mode; If SCI ≤ 35 and there is no impedance anomaly, switch from S0 mode to S2 mode; If any patient feedback event is received, such as a “prodromal or discomfort” event or a “effective or comfortable” event, switch from S0 mode, S1 mode or S2 mode to S3 mode; If SCI ≤ 35 is found in S3 mode, switch from S3 mode to S2 mode; If both impedance anomaly and "prodromal or uncomfortable" event are met simultaneously in S3 mode, switch from S3 mode to S4 mode. If the doctor remotely deactivates the alarm or the impedance returns to normal for 24 hours, the system will switch from S4 mode to S0 mode.

[0047] All of the above switching is performed by the local optimization control device, which sends the length of the corresponding first time window to the local real-time control device via the radio frequency link.

[0048] In some embodiments, the cloud server 13 is configured to: analyze the first data and generate alarm messages and / or a second control strategy within a third time window that is longer than the second time window.

[0049] Optionally, the first data includes: the patient's physiological data and treatment-related data within the first time unit, as well as a collection of patient feedback-related data.

[0050] Optionally, the first time unit can be at least one of week, month, or year.

[0051] Optionally, the second control strategy is generated by the cloud server based on the first data.

[0052] Optionally, the second control strategy includes a set of rules for optimizing the local optimization control device in determining the first control strategy. It should be understood that the second control strategy includes a set of rules for optimizing the judgment logic of the local optimization control device in determining the first control strategy.

[0053] Optionally, the second control strategy includes at least one of the following: a threshold for characterizing signal confidence abnormalities, a priority for switching working modes, and a mapping relationship between patient feedback events and stimulus parameters.

[0054] Optionally, a threshold characterizing abnormal signal confidence can be set, for example, a high-sensitivity mode trigger threshold of 85. If the patient's average frequency of attacks over the past 7 days is >3 times / day, the high-sensitivity mode trigger threshold can be lowered from 85 to 80, thereby improving detection sensitivity.

[0055] Optionally, the priority of switching operating modes can include: when the switching trigger conditions for both S1 and S3 modes are met simultaneously, priority is given to entering S3 mode (feedback driven), which can resolve model conflicts. Alternatively, in S4 mode, if the impedance returns to normal and no new "discomfort" events persist for 24 hours, it can automatically return to S0 mode, thereby reducing physician intervention.

[0056] Optionally, the mapping relationship between patient feedback events and stimulation parameters may include the correspondence between the event type, occurrence frequency, and at least one of the associated electrode channels and the adjustment of stimulation parameters (including but not limited to the direction and adjustment ratio of the stimulation current, pulse width, or frequency), as shown in Table 1 below. When the event type of the patient feedback event is the first event type, the adjustment amplitude is negative to reduce the stimulation intensity.

[0057] Table 1

[0058] In some embodiments, the above mapping relationship can be personalized by the cloud server 13 based on the patient's historical feedback data, and loaded into the local optimization control device 12 after digital signature verification.

[0059] Optional physiological data includes, but is not limited to: electroencephalogram (EEG) signals, heart rate data, and motion data.

[0060] Optional treatment-related data include, but are not limited to: electrode impedance spectroscopy (EIS), stimulation parameter logs, SCI records, etc.

[0061] Optionally, patient feedback-related data includes, but is not limited to: trigger signals and the patient feedback events corresponding to the trigger signals. In some embodiments, such as Figure 2As shown, the local real-time control device 11 may include: multiple electrodes 111, a processor 112, and a controller 113. Specifically, the multiple electrodes 111 are used to acquire EEG signals from multiple channels. The processor 112 is used to determine whether there is an abnormal first EEG signal among the EEG signals in the multiple channels based on signal confidence, and when an abnormal first EEG signal is determined to exist, it sends abnormal event information to the controller 113. Optionally, the abnormal event information includes the channel identifier corresponding to the first EEG signal. When receiving the abnormal event information, the controller 113 is used to match initial stimulation parameters in a preset stimulation strategy based on signal confidence, and to safely limit the initial stimulation parameters based on the current impedance of the target electrode group, then adjust the limited stimulation parameters based on a second instruction, apply pulse stimulation to the target electrode group, and monitor the impedance of the target electrode group in real time.

[0062] Optionally, the controller can be configured to: calculate the maximum permissible output current based on real-time measured electrode impedance values, ensuring that the stimulation voltage does not exceed a preset safe voltage limit (e.g., 10 V); if the current corresponding to the initial stimulation parameters exceeds this maximum permissible value, it is forcibly limited to the aforementioned maximum permissible current. This safety limiting mechanism is independent of patient feedback events and takes precedence over any comfort adjustments to ensure that electrical stimulation is always within an electrochemical safety window.

[0063] Optionally, the initial pulse parameters include at least one of the following: stimulation current (e.g., current amplitude of 0.5–3.0 mA), pulse width (e.g., 60–450 μs), stimulation duration (e.g., 100–500 ms), and frequency (e.g., 100–130 Hz).

[0064] Optionally, the target electrode set is determined based on the channel identifier.

[0065] Optionally, the controller 113 applies pulsed stimulation to a target electrode combination determined based on the spatial distribution characteristics of the first EEG signal. Optionally, the spatial distribution characteristics include the energy proportion of the abnormal signal in each channel, the salience of bipolar leads, and source localization results. Channel identifiers are generated based on the spatial distribution characteristics.

[0066] Optionally, the second instruction is generated by the local optimization control device based on patient feedback events of the first event type within a first preset time period.

[0067] Optionally, the second instruction is generated based on the number of patient feedback events of the first event type occurring within a first preset duration. For example, the more times the patient feedback events of the first event type occur, the greater the reduction in the stimulation current indicated by the second instruction. Optionally, the first preset duration can be 24 hours, but is not limited to this. Optionally, the second instruction can be determined based on the correspondence shown in Table 1 above.

[0068] For example, if a patient presses the "Discomfort" button on the patient feedback device, the local optimization control device 12 can record: the event type is the first event type, the timestamp is T0, the current operating mode is S1 mode, and the electrode channels involved include Ch3-Ch5. Referring to Table 1 above, the stimulation current corresponding to the first event type is reduced by 15%. If there has been one first event type in the past 24 hours, the total reduction in stimulation current = 15% × (1 + 0.2 × 1) = 18%, thus obtaining the second instruction to reduce the stimulation current by 18%.

[0069] In some embodiments, the local optimization control device 12 can maintain a sliding time window (i.e., a first preset duration, defaulting to 24 hours) to count the number of patient feedback events of the first event type. If an event occurs once within this window, a second instruction is generated to reduce the stimulation current by 15%; if an event occurs ≥2 times, the reduction is increased to 25%.

[0070] In some embodiments, the preset stimulation strategy may include a correspondence (or a mapping relationship) between SCI range intervals and stimulation parameters, as shown in Table 2 below.

[0071] Table 2

[0072] In the above embodiment, after the local real-time control layer determines that there is an abnormal first EEG signal, it first selects the initial stimulation parameters from the preset stimulation strategy based on the signal confidence level, then receives the second instruction from the local optimization control layer, adjusts the stimulation intensity according to the patient's recent discomfort events, and finally, before the stimulation output, it performs a safety limit on the stimulation parameters based on the real-time electrode impedance value to ensure that the output current does not exceed the electrode safety threshold.

[0073] Optionally, the second instruction is generated based on the patient's "aura / discomfort" feedback events in the past 24 hours.

[0074] Optionally, when the current impedance of the target electrode assembly exceeds 5 kΩ, the stimulation current amplitude is limited to below 1.0 mA.

[0075] In some embodiments, such as Figure 2 As shown, the controller 13 may include an impedance monitoring circuit. Optionally, the impedance monitoring circuit is configured to stop the output of pulse stimulation by controlling a hardware enable signal to go low when an impedance abnormality of the electrode is detected.

[0076] Optionally, the impedance monitoring circuit may include an analog comparator. The input of this comparator is connected to the electrode sampling point, and the reference terminal is set with open-circuit and short-circuit threshold voltages. The output of the comparator is directly coupled to the enable port of the H-bridge drive circuit for the stimulation pulse output via a hardware enable signal line. When the electrode impedance is detected to exceed the safe range, the comparator immediately pulls down the hardware enable signal, thereby physically cutting off the stimulation current output path without intervention from the controller 13, ensuring patient safety. Since the response time of the hardware cutoff mechanism is less than 10 microseconds, much faster than software interruption processing (typically >100 microseconds), this scheme can effectively prevent tissue burns caused by high-voltage breakdown or short circuits due to electrode open circuits.

[0077] In some embodiments, the patient feedback device includes a first button and a second button, wherein the priority of a first event type corresponding to the first button is higher than the priority of a second event type corresponding to the second button.

[0078] Optionally, the patient feedback event of the first event type can be an "apprehension or discomfort" event.

[0079] Optionally, the second event type of patient feedback event can be an "effective or comfortable" event.

[0080] In some embodiments, the local real-time control device 11 is further configured to: send an electrode abnormality signal to the local optimization control device 12 when the electrode impedance change rate continuously exceeds a preset threshold for a second preset duration. The local optimization control device 12 is further configured to: when it receives the electrode abnormality signal and the event type of the patient feedback event is a first event type, perform the following operations: Send a third instruction to the local real-time control device 11 to indicate that the output of pulse stimulation is disabled; Adjust the length of the second time window to the lower limit of the second time window length; An encrypted event is sent to cloud server 13, triggering cloud server 13 to send a notification message to the doctor's terminal.

[0081] Optionally, the second preset duration can be 3 seconds, but it is not limited to this.

[0082] Optionally, the local real-time control device 11 is configured to periodically monitor electrode impedance, calculate its rate of change, and generate an electrode abnormality event and report it to the local optimization control device 12 when the rate of change of impedance continuously exceeds 2.0 Ω / s for more than 3 seconds. The local optimization control device 12 is configured to determine that a dual triggering condition is met if a "prodromal / discomfort" type event from the patient feedback device also exists within the same time window when the electrode abnormality event is received. In this case, the local optimization control device 12 can send a third instruction to the local real-time control device 11 to disable the output of pulse stimulation, and simultaneously adjust the length of the second time window to the lower limit of the second time window length (e.g., 200 milliseconds). It can also send an encrypted event to the cloud server 13, triggering the cloud server 13 to send a notification message to the doctor's terminal.

[0083] In some embodiments, treatment-related data includes impedance spectroscopy data, which can be uploaded to a cloud server 13 via a local optimization control device 12. The cloud server 13 is configured to send an alarm message to the doctor's terminal when it determines that the impedance of any electrode exhibits a monotonically increasing trend within a third preset time period, based on the charge transfer resistance and double-layer capacitance included in the impedance spectroscopy data.

[0084] Optionally, the impedance spectrum data may include electrochemical impedance spectroscopy (EIS) data periodically acquired by the local real-time control device 11. The electrochemical impedance spectroscopy data is generated by the local real-time control device 11 applying a multi-frequency AC test signal with an amplitude not exceeding 10 mV to the electrode during non-stimulation periods, measuring its voltage response, generating a complex impedance spectrum, and uploading it to the cloud server 13 via the local optimization control device 12.

[0085] Optionally, the cloud server is configured to fit a Randle equivalent circuit model based on electrochemical impedance spectroscopy data, extract charge transfer resistance and double-layer capacitance, and send a predictive maintenance alert to the doctor's terminal when the charge transfer resistance of any electrode shows a monotonically increasing trend over 21 consecutive days (which may correspond to a third preset duration, but is not limited to this).

[0086] Optionally, the complex impedance Z can be calculated using the following formula: Z = R ct / (1+j×2π×f×R ct ×C dl ).

[0087] Where j=-1, f is the frequency of the excitation signal (Hz), and R ct The charge transfer resistance reflects the electrochemical activity of the electrode surface, C dl This is a double-layer capacitance, reflecting the effective area of ​​the interface. When R is present... ct Monotonically increasing and Cdl When the synergistic trend continues to decline, it can be determined that the electrode coating has degraded.

[0088] In some embodiments, communication between the local optimization control device 12 and the cloud server 13 is encrypted using an encryption protocol.

[0089] In some embodiments, such as Figure 2 As shown, the local optimization control device 12 includes a hardware security module. Optionally, the hardware security module is configured to: digitally verify the second control policy downloaded from the cloud server 13, and allow loading after successful verification.

[0090] In summary, the embodiments of this application provide a safe, intelligent, and personalized implantable closed-loop neurostimulation system, which can achieve the following technical effects: 1. This invention provides a highly efficient, collaborative closed-loop neuromodulation system supporting multiple time granularities. Specifically, the local real-time control device detects abnormal signals and outputs stimulation within the first time window (millisecond level); the local optimization control device dynamically adjusts the first control strategy based on patient feedback signals and physiological data within the second time window (second level); and the cloud server analyzes long-term data within the third time window (day / week / month level) to generate a second control strategy for optimizing local decisions. These three components work together efficiently to form a closed-loop system of "fast response + slow optimization."

[0091] 2. A neuromodulation system that supports adaptive switching of working modes is provided. It can automatically switch between multiple working modes such as routine monitoring, high sensitivity, caution, feedback-driven, and safety pause based on signal confidence, impedance change rate, and patient feedback events. It can dynamically adjust the length of the first time window and / or the second time window, thereby improving the detection rate of abnormal events and reducing the risk of false stimulation.

[0092] 3. The stimulation parameters are adjusted for humanized comfort. By structuring the timestamps and event types of patient feedback events and generating a second instruction based on their frequency, the system can make personalized fine-tuning of the stimulation parameters within safe limits, achieving a treatment experience that becomes more comfortable the more you use it.

[0093] 4. The second control policy (such as mapping table, SCI threshold rule) generated by the cloud server is encrypted and then distributed. It is then loaded after being verified by the hardware security module of the local optimized control device, ensuring that the policy source is trustworthy and the execution is controllable, thus ensuring the security of the control policy.

[0094] 5. It can actively monitor the state of the electrodes. By periodically calculating the rate of change of electrode impedance, the implantable closed-loop neurostimulation system can identify abnormalities such as electrode coating peeling and tissue encapsulation in the early stage. Combined with patient feedback, it can trigger a safety pause mechanism to effectively prevent tissue damage or equipment failure.

[0095] It should be understood that the devices in the above embodiments (e.g., local real-time control device 11, local optimization control device 12, and cloud server 13, as well as patient feedback device 14 and doctor terminal) may include processors. Processors may include CPUs (Central Processing Units), general-purpose processors, DSPs (Digital Signal Processors), ASICs (Application Specific Integrated Circuits), FPGAs (Field-Programmable Gate Arrays), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. They can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processors may also be combinations that implement computational functions, such as combinations of one or more microprocessors, combinations of DSPs and microprocessors, etc.

[0096] It should also be understood that the devices in the above embodiments may further include a memory. The memory may be a ROM (Read-Only Memory) or other type of static storage device capable of storing static information and instructions, RAM (Random Access Memory) or other type of dynamic storage device capable of storing information and instructions, or it may be an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read-Only Memory) or other optical disc storage, optical disk storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media, or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0097] It should also be understood that the devices in the above embodiments may further include a communication unit. The communication unit can be used for receiving and transmitting signals. The communication unit can allow wireless or wired communication between the devices to exchange data. It should be noted that in practical applications, the number of communication units is not limited to one.

[0098] It should also be understood that the devices in the above embodiments may further include an input unit. The input unit can be used to receive input numbers, characters, images, and / or sound information, or to generate key signal inputs related to user settings and function control of each device. The input unit may include, but is not limited to, one or more of the following: a touchscreen, a physical keyboard, function keys (such as volume control buttons, power buttons, etc.), a trackball, a mouse, a joystick, a camera, a microphone, etc.

[0099] It should also be understood that each device in the above embodiments may further include an output unit. The output unit can be used to output or display information processed by the processor of each device. The output unit may include, but is not limited to, one or more of a display device, a speaker, a vibration device, etc.

[0100] Based on the same inventive concept, this application provides a control method for an implantable closed-loop neurostimulation system, which may include the implantable closed-loop neurostimulation system described in the above embodiments. Figure 3 As shown, the method includes: S1. The implanted intracranial local real-time control device calculates the signal confidence of multiple channels of EEG signals in real time within the first time window to determine whether there is an abnormal first EEG signal, and triggers pulse stimulation when an abnormal first EEG signal is determined, and periodically calculates the rate of change of electrode impedance.

[0101] Optionally, the signal confidence level is calculated by the local real-time control device based on the signal-to-noise ratio of the EEG signal, multiple channel consistency parameters, and parameters characterizing physiological artifacts.

[0102] S2. The local real-time control device sends the signal confidence level and the electrode impedance change rate to the local optimization control device.

[0103] S3. The local optimization control device is an extracranial portable device. It receives the electrode impedance change rate and the signal confidence level sent by the local real-time control device, processes the trigger signal from the patient feedback device within a second time window that is greater than the first time window to obtain a patient feedback event, and determines a first control strategy based on at least one of the electrode impedance change rate, the signal confidence level, and the patient feedback event.

[0104] Optionally, the first control strategy includes: whether to switch the current working mode, and when it is determined to switch the current working mode, sending the first instruction to the local real-time control device.

[0105] Optionally, different working modes correspond to different lengths of the first time window, and / or different lengths of the second time window.

[0106] Optionally, the operating modes include: normal monitoring mode, high sensitivity mode, cautious mode, feedback-driven mode, and safety pause mode.

[0107] Optionally, the switching trigger condition for the high-sensitivity mode is: the signal confidence level is greater than or equal to 85, and it lasts for a fourth preset duration.

[0108] Optionally, the switching trigger condition for the cautious mode is: the signal confidence level is less than or equal to 35, and no electrode impedance change rate is detected to continuously exceed a preset threshold for a second preset duration.

[0109] Optionally, the switching trigger condition for the feedback-driven mode is: the existence of a patient feedback event of any event type.

[0110] Optionally, the switching trigger condition for the safe pause mode is: the rate of change of electrode impedance is continuously detected to exceed a preset threshold within a fifth preset time period, and the event type of the patient feedback event is the first event type.

[0111] Optionally, the patient feedback event includes the timestamp and event type corresponding to the trigger signal.

[0112] S4. The local real-time control device adjusts the length of the first time window based on the first instruction sent by the local optimized control device.

[0113] S5. The cloud server analyzes the first data and generates an alarm message and / or a second control strategy within a third time window that is longer than the second time window.

[0114] Optionally, the first data includes: the patient's physiological data and treatment-related data within the first time unit, as well as a collection of patient feedback-related data.

[0115] Optionally, the first time unit is at least one of week, month, and year.

[0116] Optionally, the second control strategy includes a set of rules for optimizing the local optimization control device to determine the first control strategy.

[0117] Optionally, the second control strategy is generated by the cloud server based on the first data.

[0118] Optionally, the second control strategy includes at least one of the following: a threshold for characterizing signal confidence abnormality, a priority for switching working modes, and a mapping relationship between patient feedback events and stimulation parameters.

[0119] In some embodiments, the local real-time control device includes: multiple electrodes, a processor, and a controller, wherein the multiple electrodes acquire EEG signals from multiple channels; the processor determines whether an abnormal first EEG signal exists in the EEG signals of the multiple channels based on the signal confidence level, and when the existence of the abnormal first EEG signal is determined, sends abnormal event information to the controller, the abnormal event information including the channel identifier corresponding to the first EEG signal; when the controller receives the abnormal event information, it matches initial stimulation parameters in a preset stimulation strategy based on the signal confidence level, performs a safety limit on the initial stimulation parameters based on the current impedance of the target electrode group, adjusts the limited stimulation parameters based on the second instruction, applies pulse stimulation to the target electrode group, and monitors the impedance of the target electrode group in real time.

[0120] Optionally, the controller can calculate the maximum permissible output current based on real-time measured electrode impedance values, ensuring that the stimulation voltage does not exceed a preset safe voltage limit (e.g., 10V). If the current corresponding to the initial stimulation parameters exceeds this maximum permissible value, it is forcibly limited to the aforementioned maximum permissible current. This safety limiting mechanism is independent of patient feedback events and takes precedence over any comfort adjustments to ensure that electrical stimulation is always within an electrochemical safety window.

[0121] Optionally, the initial pulse parameters include at least one of stimulation current, pulse width, stimulation duration, and frequency.

[0122] Optionally, the target electrode group is determined based on the channel identifier.

[0123] Optionally, the second instruction is generated by the local optimization control device based on patient feedback events of a first event type within a first preset time period.

[0124] In some embodiments, the controller includes an impedance monitoring circuit that, when the impedance monitoring circuit detects an impedance abnormality in the electrode, stops the output of pulse stimulation by controlling a hardware enable signal to go low.

[0125] In some embodiments, the patient feedback device may include a first button and a second button, wherein the priority of a first event type corresponding to the first button is higher than the priority of a second event type corresponding to the second button.

[0126] In some embodiments, the control method described above may further include: when the rate of change of electrode impedance continuously exceeds a preset threshold for a second preset duration, the local real-time control device sends an electrode abnormality signal to the local optimization control device. When the local optimization control device receives the electrode abnormality signal and the patient feedback event is of the first event type, the following operations are performed: A third instruction is sent to the local real-time control device to instruct the output of pulse stimulation to be disabled; Adjust the length of the second time window to the lower limit of the second time window length; An encrypted event is sent to the cloud server, triggering the cloud server to send a notification message to the doctor's terminal.

[0127] In some embodiments, the treatment-related data includes impedance spectroscopy data, which is uploaded to the cloud server via the local optimization control device. In this embodiment, the control method may further include: the cloud server, based on the charge transfer resistance and double-layer capacitance included in the impedance spectroscopy data, determining when the impedance of any electrode shows an increasing trend within a third preset time period, and then sending the alarm message to the doctor's terminal.

[0128] In some embodiments, the communication between the local optimization control device and the cloud server is encrypted using an encryption protocol. If the local optimization control device includes a hardware security module, the control method may further include: the hardware security module performing digital signature verification on the second control policy downloaded from the cloud server, and allowing loading after successful verification.

[0129] Optionally, the second control strategy is generated by the cloud server based on the first data.

[0130] Optionally, the second control strategy includes at least one of the following: a threshold for characterizing signal confidence abnormality, a priority for switching working modes, and a mapping relationship between patient feedback events and stimulation parameters.

[0131] In the description of this application, the terms "center," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate directions or positional relationships based on the exemplary directions or positional relationships shown in the accompanying drawings. They are used to facilitate the description or simplification of the embodiments of this application and are not intended to indicate or imply that the device or component referred to must have a specific orientation or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0132] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0133] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0134] In the description of this specification, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0135] The above description is only a partial implementation of this application. It should be noted that for those skilled in the art, other similar implementation methods based on the technical concept of this application, without departing from the technical concept of this application, also fall within the protection scope of the embodiments of this application.

Claims

1. An implantable closed-loop neurostimulation system, characterized in that, include: A local real-time control device, a local optimized control device, and a cloud server are sequentially connected via communication, along with a patient feedback device connected to the local optimized control device. The local real-time control device is implanted in the cranium and is configured to: calculate the signal confidence of the EEG signals in multiple channels collected in real time within a first time window to determine whether there is an abnormal first EEG signal, and trigger pulse stimulation when an abnormal first EEG signal is determined to exist; it is also configured to: periodically calculate the electrode impedance change rate, send the signal confidence and the electrode impedance change rate to the local optimization control device, and adjust the length of the first time window based on the first instruction sent by the local optimization control device. The local optimization control device is an extracranial portable device, configured to: process trigger signals from the patient feedback device to obtain patient feedback events within a second time window greater than the first time window, and further configured to: determine a first control strategy based on at least one of the electrode impedance change rate, the signal confidence level, and the patient feedback event, the first control strategy including: whether to switch the current working mode, and when it is determined that the current working mode should be switched, send the first instruction to the local real-time control device; The cloud server is configured to analyze first data and generate alarm messages and / or second control strategies within a third time window that is longer than the second time window. The first data includes: physiological data and treatment-related data of the patient within the first time unit, as well as a set of patient feedback-related data.

2. The implantable closed-loop neurostimulation system according to claim 1, characterized in that, The local real-time control device includes: multiple electrodes, a processor, and a controller, wherein, The multiple electrodes are used to acquire the electroencephalogram (EEG) signals from multiple channels; The processor is configured to determine, based on the signal confidence level, whether the abnormal first EEG signal exists in the EEG signals of the plurality of channels, and when it is determined that the abnormal first EEG signal exists, to send abnormal event information to the controller, the abnormal event information including the channel identifier corresponding to the first EEG signal; The controller is used to, upon receiving the abnormal event information, match the initial stimulation parameters in a preset stimulation strategy based on the signal confidence level, and perform a safety limit on the initial stimulation parameters based on the current impedance of the target electrode group, then adjust the limited stimulation parameters based on the second instruction, apply pulse stimulation to the target electrode group, and monitor the impedance of the target electrode group in real time. The initial pulse parameters include at least one of stimulation current, pulse width, stimulation duration, and frequency; the target electrode group is determined based on the channel identifier; and the second instruction is generated by the local optimization control device based on patient feedback events of a first event type within a first preset duration.

3. The implantable closed-loop neurostimulation system according to claim 2, characterized in that, The controller includes an impedance monitoring circuit, which is configured to stop the output of pulse stimulation by controlling a hardware enable signal to go low when an impedance abnormality of the electrode is detected.

4. The implantable closed-loop neurostimulation system according to any one of claims 1-3, characterized in that, The patient feedback event includes the timestamp and event type corresponding to the trigger signal; The patient feedback device includes a first button and a second button, wherein the priority of a first event type corresponding to the first button is higher than the priority of a second event type corresponding to the second button.

5. The implantable closed-loop neurostimulation system according to any one of claims 1-3, characterized in that, The local real-time control device is also configured to send an electrode abnormality signal to the local optimization control device when the electrode impedance change rate continuously exceeds a preset threshold for a second preset duration. The local optimization control device is further configured to perform the following operations when it receives the electrode abnormality signal and the event type of the patient feedback event is a first event type: A third instruction is sent to the local real-time control device to instruct the output of pulse stimulation to be disabled; Adjust the length of the second time window to the lower limit of the second time window length; An encrypted event is sent to the cloud server, triggering the cloud server to send a notification message to the doctor's terminal.

6. The implantable closed-loop neurostimulation system according to any one of claims 1-3, characterized in that, The treatment-related data includes impedance spectroscopy data, which is uploaded to the cloud server via the local optimization control device. The cloud server is configured to send an alarm message to the doctor's terminal when it determines that the impedance of any electrode shows an increasing trend within a third preset time period, based on the charge transfer resistance and double-layer capacitance included in the impedance spectrum data.

7. The implantable closed-loop neurostimulation system according to any one of claims 1-3, characterized in that, The operating modes include: normal monitoring mode, high sensitivity mode, cautious mode, feedback-driven mode, and safety pause mode; The switching trigger condition for the high-sensitivity mode is: the signal confidence level is greater than or equal to 85, and it lasts for a fourth preset duration; The switching trigger condition for the cautious mode is: the signal confidence level is less than or equal to 35, and no electrode impedance change rate is detected to continuously exceed a preset threshold for a second preset duration. The switching trigger condition for the feedback-driven mode is: the existence of a patient feedback event of any event type; The switching trigger condition for the safe pause mode is: the rate of change of electrode impedance is continuously detected to exceed a preset threshold within a fifth preset time period, and the event type of the patient feedback event is the first event type.

8. The implantable closed-loop neurostimulation system according to claim 7, characterized in that, Different working modes correspond to different lengths of the first time window, and / or different lengths of the second time window.

9. The implantable closed-loop neurostimulation system according to any one of claims 1-3, characterized in that, The signal confidence level is calculated by the local real-time control device based on the signal-to-noise ratio of the EEG signal, multiple channel consistency parameters, and parameters characterizing physiological artifacts.

10. The implantable closed-loop neurostimulation system according to any one of claims 1-3, characterized in that, The second control strategy includes a set of rules for optimizing the local optimization control device to determine the first control strategy.

11. The implantable closed-loop neurostimulation system according to claim 10, characterized in that, The communication between the local optimization control device and the cloud server is encrypted using an encryption protocol. The local optimization control device includes a hardware security module, which is configured to: perform digital signature verification on the second control policy downloaded from the cloud server, and allow loading after successful verification. The second control strategy is generated by the cloud server based on the first data, and the second control strategy includes at least one of the following: a threshold for representing abnormal confidence of the signal, a priority for switching working modes, and a mapping relationship between patient feedback events and stimulation parameters.

12. A control method, characterized in that, The control method is applied to the implantable closed-loop neurostimulation system as described in any one of claims 1-11, and the control method includes: The implanted intracranial local real-time control device calculates the signal confidence of multiple channels of EEG signals in real time within the first time window to determine whether there is an abnormal first EEG signal, and triggers pulse stimulation when an abnormal first EEG signal is determined; periodically calculates the electrode impedance change rate, and sends the signal confidence and the electrode impedance change rate to the local optimization control device, and adjusts the length of the first time window based on the first instruction sent by the local optimization control device; The local optimization control device is an extracranial portable device. It receives the electrode impedance change rate and the signal confidence level sent by the local real-time control device. Within a second time window that is greater than the first time window, it processes the trigger signal from the patient feedback device to obtain a patient feedback event. Based on at least one of the electrode impedance change rate, the signal confidence level, and the patient feedback event, it determines a first control strategy. The first control strategy includes: whether to switch the current working mode, and when it is determined that the current working mode should be switched, sending the first instruction to the local real-time control device. The cloud server analyzes the first data and generates alarm messages and / or a second control strategy within a third time window that is longer than the second time window. The first data includes: the patient's physiological data and treatment-related data within the first time unit, as well as a collection of patient feedback-related data.

13. The control method according to claim 12, characterized in that, The local real-time control device includes: multiple electrodes, a processor, and a controller; the method further includes: The multiple electrodes acquire the electroencephalogram (EEG) signals from multiple channels; The local real-time control device calculates the signal confidence of the acquired EEG signals from multiple channels in real time within a first time window to determine whether there is an abnormal first EEG signal, and triggers pulse stimulation when an abnormal first EEG signal is determined to exist, including: The processor determines whether the abnormal first EEG signal exists in the EEG signals of the multiple channels based on the signal confidence level, and sends abnormal event information to the controller when it determines that the abnormal first EEG signal exists. The abnormal event information includes the channel identifier corresponding to the first EEG signal. When the controller receives the abnormal event information, it matches the initial stimulation parameters in the preset stimulation strategy based on the signal confidence, and performs a safety limit on the initial stimulation parameters based on the current impedance of the target electrode group. Then, it adjusts the limited stimulation parameters based on the second instruction, applies pulse stimulation to the target electrode group, and monitors the impedance of the target electrode group in real time. The initial pulse parameters include at least one of stimulation current, pulse width, stimulation duration, and frequency; the target electrode group is determined based on the channel identifier; and the second instruction is generated by the local optimization control device based on patient feedback events of a first event type within a first preset duration.

14. The control method according to claim 13, characterized in that, The controller includes an impedance monitoring circuit. When the impedance monitoring circuit detects an impedance abnormality of the electrode, it stops the output of pulse stimulation by controlling the hardware enable signal to go low.

15. The control method according to any one of claims 11-13, characterized in that, The patient feedback event includes the timestamp and event type corresponding to the trigger signal; The patient feedback device includes a first button and a second button, wherein the priority of a first event type corresponding to the first button is higher than the priority of a second event type corresponding to the second button.

16. The control method according to any one of claims 11-13, characterized in that, The control method further includes: When the rate of change of the electrode impedance continuously exceeds a preset threshold for a second preset duration, the local real-time control device sends an electrode abnormality signal to the local optimization control device. When the local optimization control device receives the electrode abnormality signal and the patient feedback event is of the first event type, the following operations are performed: A third instruction is sent to the local real-time control device to instruct the output of pulse stimulation to be disabled; Adjust the length of the second time window to the lower limit of the second time window length; An encrypted event is sent to the cloud server, triggering the cloud server to send a notification message to the doctor's terminal.

17. The control method according to any one of claims 11-13, characterized in that, The treatment-related data includes impedance spectroscopy data, which is uploaded to the cloud server via the local optimization control device. The control method further includes: Based on the impedance spectrum data, including charge transfer resistance and double-layer capacitance, the cloud server determines when the impedance of any electrode shows an increasing trend within a third preset time period, and then sends the alarm message to the doctor's terminal.

18. The control method according to any one of claims 11-13, characterized in that, The operating modes include: normal monitoring mode, high sensitivity mode, cautious mode, feedback-driven mode, and safety pause mode; The switching trigger condition for the high-sensitivity mode is: the signal confidence level is greater than or equal to 85, and it lasts for a fourth preset duration; The switching trigger condition for the cautious mode is: the signal confidence level is less than or equal to 35, and no electrode impedance change rate is detected to continuously exceed a preset threshold for a second preset duration. The switching trigger condition for the feedback-driven mode is: the existence of a patient feedback event of any event type; The switching trigger condition for the safe pause mode is: the rate of change of electrode impedance is continuously detected to exceed a preset threshold within a fifth preset time period, and the event type of the patient feedback event is the first event type.

19. The control method according to claim 18, characterized in that, Different working modes correspond to different lengths of the first time window, and / or different lengths of the second time window.

20. The control method according to any one of claims 11-13, characterized in that, The signal confidence level is calculated by the local real-time control device based on the signal-to-noise ratio of the EEG signal, multiple channel consistency parameters, and parameters characterizing physiological artifacts.

21. The control method according to any one of claims 11-13, characterized in that, The second control strategy includes a set of rules for optimizing the local optimization control device to determine the first control strategy.

22. The control method according to claim 21, characterized in that, The communication between the local optimization control device and the cloud server is encrypted using an encryption protocol. The local optimization control device includes a hardware security module. The control method further includes: The hardware security module performs digital signature verification on the second control policy downloaded from the cloud server, and allows loading after successful verification; The second control strategy is generated by the cloud server based on the first data, and the second control strategy includes at least one of the following: a threshold for representing abnormal confidence of the signal, a priority for switching working modes, and a mapping relationship between patient feedback events and stimulation parameters.

Images