Signal processing methods, implantable closed-loop neurostimulation systems and storage media

Abstract

This application relates to the field of medical device technology, and discloses a signal processing method, an implantable closed-loop neurostimulation system, and a storage medium. The method includes: calculating feature change indices in at least two dimensions based on a target EEG signal; determining a current processing strategy based on the feature change indices in the at least two dimensions; generating a first control command when the current processing strategy is the first strategy; the first control command instructing the output of a first electrical stimulation signal according to a fixed first set of stimulation parameters; generating a second control command when the current processing strategy is the second strategy; the second control command instructing the output of a second electrical stimulation signal according to a dynamically adjusted second set of stimulation parameters; wherein each stimulation parameter in the dynamically adjusted second set of stimulation parameters is calculated based on the modulation of the target EEG signal by the previous stimulation parameter.

Description

Technical Field

[0001] This application relates to the field of medical device technology, and more specifically, to a signal processing method, an implantable closed-loop neurostimulation system, and a storage medium. Background Technology

[0002] Implantable closed-loop neurostimulation systems, as advanced medical devices, have great development prospects. Currently, the operation of implantable closed-loop neurostimulation systems typically involves outputting electrical stimulation according to preset stimulation parameters when a target condition is detected.

[0003] However, while the above-mentioned working method of the implantable closed-loop neurostimulation system is logically simple and relatively safe, it has another drawback: the working method of the implantable closed-loop neurostimulation system is not flexible enough. Summary of the Invention

[0004] This application provides a signal processing method, an implantable closed-loop neurostimulation system, and a storage medium to address the problem of poor flexibility in the operation of existing implantable closed-loop neurostimulation systems.

[0005] To address the aforementioned problems, this application discloses a signal processing method applied to a signal processing system, the signal processing method comprising:

[0006] Based on the target EEG signal, at least two dimensions of feature change indices are calculated; wherein, the target EEG signal is an EEG signal conforming to the target signal pattern, and the feature change indices include: a first feature index based on the change of signal frequency domain energy in the target frequency band, a second feature index based on the change of signal complexity, a third feature index based on the change of signal amplitude, a fourth feature index based on the change of the propagation speed of the target EEG signal, a fifth feature index based on the change of the propagation range of the target EEG signal, a sixth feature index based on the change of the signal dominant frequency, and a seventh feature index based on the change of the signal spectral centroid;

[0007] Based on the feature change indicators of at least two dimensions, the current processing strategy is determined; wherein, the processing strategy includes a first strategy and a second strategy;

[0008] When the current processing strategy is the first strategy, a first control instruction is generated; wherein, the first control instruction is used to instruct the output of a first electrical stimulation signal according to a fixed first set of stimulation parameters;

[0009] When the current processing strategy is the second strategy, a second control instruction is generated; wherein the second control instruction is used to instruct the output of a second electrical stimulation signal according to a dynamically adjusted second set of stimulation parameters; wherein, in the dynamically adjusted second set of stimulation parameters, each stimulation parameter is calculated based on the modulation of the target EEG signal by the previous stimulation parameter.

[0010] This application also discloses an implantable closed-loop neurostimulation system, which includes:

[0011] The feature index module is used to calculate feature change indices in at least two dimensions based on the target EEG signal; wherein the target EEG signal is an EEG signal that conforms to the target signal pattern, and the feature change indices include multiple of the following: a first feature index based on the change of signal frequency domain energy in the target frequency band, a second feature index based on the change of signal complexity, a third feature index based on the change of signal amplitude, a fourth feature index based on the change of the propagation speed of the target EEG signal, a fifth feature index based on the change of the propagation range of the target EEG signal, a sixth feature index based on the change of the signal dominant frequency, and a seventh feature index based on the change of the signal spectral centroid;

[0012] A processing strategy module is used to determine the current processing strategy based on the feature change indicators of the at least two dimensions; wherein the processing strategy includes a first strategy and a second strategy.

[0013] The first processing module is configured to generate a first control instruction when the current processing strategy is the first strategy; wherein the first control instruction is configured to instruct the output of a first electrical stimulation signal according to a fixed first set of stimulation parameters.

[0014] The second processing module is used to generate a second control instruction when the current processing strategy is the second strategy; wherein the second control instruction is used to instruct the output of a second electrical stimulation signal according to a dynamically adjusted second set of stimulation parameters; wherein each stimulation parameter in the dynamically adjusted second set of stimulation parameters is calculated based on the modulation of the target EEG signal by the previous stimulation parameter.

[0015] This application also discloses a signal processing system, including the implantable closed-loop neurostimulation system described above.

[0016] This application also discloses a computer-readable storage medium storing a computer program that, when executed by a processor, implements one or more of the methods described in this application.

[0017] This application also discloses a computer program product, including a computer program that, when executed by a processor, implements one or more of the methods described in this application.

[0018] The beneficial effects of the technical solutions provided in this application are:

[0019] In this embodiment, before outputting the stimulation signal, a multi-dimensional in-depth analysis is first performed on the EEG signal conforming to the target signal pattern. This involves calculating characteristic change indicators across multiple dimensions, such as frequency domain energy change, signal complexity change, signal amplitude change, propagation characteristics, dominant frequency, and spectral centroid. This allows for a more refined and comprehensive analysis and evaluation of the EEG signal state, leading to intelligent decision-making on the current processing strategy. A first strategy based on fixed stimulation parameters can be used to output the electrical stimulation signal, ensuring processing efficiency and reliability. Alternatively, a second strategy based on dynamic stimulation parameters can be employed. In the second strategy, each stimulation parameter is calculated based on the actual modulation of the target signal by the previous stimulation parameter, forming a real-time, adaptive parameter optimization loop. This application effectively overcomes the technical problems of existing implantable closed-loop neurostimulation systems, such as their single working mode and poor flexibility, by establishing an intelligent prediction and strategy switching mechanism based on multi-dimensional signal feature analysis. Attached Figure Description

[0020] 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:

[0021] Figure 1 A flowchart of the signal processing method provided in the embodiments of this application;

[0022] Figure 2 A schematic diagram of the structure of the implantable closed-loop neurostimulation system provided in an embodiment of this application;

[0023] Figure 3 This is a schematic diagram of the signal processing system provided in an embodiment of this application. Detailed Implementation

[0024] 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.

[0025] Those skilled in the art will understand that, unless otherwise stated, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the terms “comprising” and “including” as used in the embodiments of this application mean that the corresponding feature can be implemented as the presented feature, information, data, step, operation, element, and / or component, but do not exclude implementation as other features, information, data, step, operation, element, component, and / or combinations thereof supported by the art. It should be understood that when we say that an element is “connected” or “coupled” to another element, the one element can be directly connected or coupled to the other element, or it can mean that the one element and the other element establish a connection relationship through an intermediate element. Furthermore, “connected” or “coupled” as used herein can include wireless connection or wireless coupling. The term “multiple” refers to two or more; therefore, in the embodiments of this application, “multiple” can also be understood as “at least two.” The term "and / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / ", unless otherwise specified, generally indicates that the related objects before and after it are in an "or" relationship.

[0026] 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.

[0027] To facilitate understanding of the technical solution of this application, the following terms will be introduced.

[0028] An electroencephalogram (EEG) is formed by summing the postsynaptic potentials that occur synchronously among a large number of neurons during brain activity. It records the changes in electrical waves during brain activity and is a comprehensive reflection of the electrophysiological activity of brain nerve cells on the surface of the cerebral cortex or scalp. It can also be called an electroencephalogram or brainwave.

[0029] An implantable closed-loop neurostimulation system can collect electroencephalogram (EEG) signals through electrodes placed near the epileptogenic focus, perform real-time analysis, and predict or monitor epileptic seizures. When abnormalities in the patient's EEG signals are detected, electrical stimulation is automatically applied to the cortex or target brain region via electrodes to inhibit excessive synchronized firing of brain neurons, thereby suppressing epileptic seizures. This electrical stimulation, also known as an electrical stimulation signal, is the electrical signal used to stimulate the brain.

[0030] Currently, existing implantable closed-loop neurostimulation systems typically employ a pre-set set of fixed stimulation parameters. When electrical stimulation is required, a signal is output according to these parameters. For example, in implantable closed-loop neurostimulation systems used to treat epilepsy, ensuring these parameters achieve optimal therapeutic effects requires physicians or medical experts to carefully formulate the parameters based on the patient's overall condition and clinical presentation, combined with their extensive clinical experience. This process is time-consuming, labor-intensive, and highly dependent on the physician. However, patients' conditions vary greatly, and even for the same patient, their condition may change unpredictably over time. Therefore, existing implantable closed-loop neurostimulation systems still have significant shortcomings in their stimulation strategies, relying on fixed parameters for electrical stimulation, resulting in poor flexibility.

[0031] Therefore, this application proposes a signal processing method applied to a signal processing system. In some embodiments, the signal processing system includes an implantable closed-loop neurostimulation system, which can be used for neuroscience research, brain-computer interface research, and treatment of targeted diseases. For ease of understanding, the following uses an implantable closed-loop neurostimulation system as an example to illustrate the signal processing method. Figure 1 As shown, the signal processing method includes:

[0032] Step 101: Calculate feature change indicators in at least two dimensions based on the target EEG signal.

[0033] In this step, the target EEG signal is an EEG signal that conforms to a target signal pattern. The implantable closed-loop neurostimulation system can monitor EEG signals and identify their characteristics. For example, it can identify whether the EEG signal conforms to the target signal pattern. The target signal pattern can be a predefined signal pattern or signal characteristic. Taking the treatment of epilepsy with an implantable closed-loop neurostimulation system as an example, the target signal pattern can be the characteristics of epilepsy onset, and the EEG signal conforming to the target signal pattern can be an EEG signal that conforms to the characteristics of epilepsy onset, i.e., the EEG signal during an epileptic seizure.

[0034] In an exemplary scenario of treating a patient using an implantable closed-loop neurostimulation system, the system continuously monitors the patient's electroencephalogram (EEG) signals via implanted electrodes. The system is pre-programmed with algorithms for identifying pathological EEG signals, such as those based on time-frequency analysis or machine learning models, to detect EEG signals (e.g., spikes, sharp waves, or paroxysmal rhythm changes) that match the pathological characteristics of the target condition in real time.

[0035] In this step, the characteristic change indicators include: a first characteristic indicator based on the change of signal frequency domain energy in the target frequency band, a second characteristic indicator based on the change of signal complexity, a third characteristic indicator based on the change of signal amplitude, a fourth characteristic indicator based on the change of the propagation speed of the target EEG signal, a fifth characteristic indicator based on the change of the propagation range of the target EEG signal, a sixth characteristic indicator based on the change of the signal dominant frequency, and a seventh characteristic indicator based on the change of the signal spectrum centroid.

[0036] The power of a signal can be used to represent its spectral energy. In some embodiments, the target frequency band is a predefined band, such as a high-frequency oscillation (HFO) signal. In some embodiments, a first characteristic index can characterize the power change of the HFO signal. Taking the treatment of epilepsy with an implantable circular nerve stimulation system as an example, the HFO signal can refer to the EEG signal in the 80-500Hz frequency band, which is closely related to epileptic activity. For example, HFOs can be considered as an extreme manifestation of abnormal synchronous discharge of neuronal groups. In the normal state (seizure-free epilepsy), the power of the HFO signal is usually low, while during an epileptic seizure, the power of the HFO signal is usually high. In some embodiments, the first characteristic index is defined as the rate of change of the HFO signal power after stimulation relative to the baseline power before stimulation (such as the power of the HFO signal in the epileptic EEG signal). For example, a power decrease of more than 50% may indicate good therapeutic effect. This first characteristic index reflects the improvement of the current epileptic seizure.

[0037] Signal complexity can be represented by the sample entropy of a signal. In some embodiments, a second characteristic index can characterize the change in sample entropy of the EEG signal. Sample entropy can be used to measure the complexity of a time series. In this embodiment, sample entropy is used to quantify the complexity of the EEG response signal. Continuing with the example of treating epilepsy using an in-circuit neurostimulation system, in a normal state (non-seizure epilepsy), the EEG signal is usually complex and irregular, and its sample entropy is usually high. During an epileptic seizure, the EEG signal becomes synchronous and monotonous, and its sample entropy is usually low. Therefore, in some embodiments, the amount or rate of change of sample entropy can be determined through the second characteristic index to reflect the recovery of brain function.

[0038] Signal amplitude can be represented by signal strength. In some embodiments, changes in the amplitude of the EEG signal can directly reflect the intensity of voltage fluctuations in the EEG signal. Continuing with the example of treating epilepsy with an implanted closed-loop neurostimulation system, in some embodiments, the implanted closed-loop neurostimulation system calculates the amplitude characteristics of the EEG signal during a certain time window, such as peak amplitude, average amplitude, or root mean square value, and compares it with the baseline amplitude to obtain the amplitude change value. During an epileptic seizure, the amplitude of the EEG signal usually increases significantly.

[0039] The propagation speed of the target EEG signal can be understood as the speed at which the target EEG signal propagates within a brain region. Continuing with the example of using an implantable closed-loop neurostimulation system to treat epilepsy, the propagation speed of the target EEG signal can be seen as the speed at which the abnormal discharge causing epilepsy spreads within the patient's brain region. In some embodiments, the implantable closed-loop neurostimulation system identifies the starting point of the target EEG signal by analyzing signals recorded by multi-channel electrodes. Then, by calculating the time delay of the target EEG signal propagating to adjacent channels and combining this with the physical distance between the electrodes, the propagation speed is estimated. The faster the propagation speed, the more aggressive the epileptic seizure and the higher the risk.

[0040] The propagation range of the target EEG signal is similar to the propagation speed of the target EEG signal mentioned above, and will not be described in detail here.

[0041] The dominant frequency of a signal includes the dominant frequency of the target EEG signal, which is the frequency component with the highest energy concentration in the power spectrum of the target EEG signal. Continuing with the example of using an implanted closed-loop neurostimulation system to treat epilepsy, the dominant frequency of the target EEG signal / the EEG signal presenting as the seizure typically changes during an epileptic seizure. The implanted closed-loop neurostimulation system calculates the power spectrum of the current EEG signal presenting as the seizure through Fast Fourier Transform (FFT), identifies the dominant frequency, and calculates the difference between it and the baseline dominant frequency as the change value.

[0042] The centroid of the signal spectrum includes the frequency at which the "center of gravity" of the power spectrum of the target EEG signal is located; it is an indicator of the overall distribution characteristics of the signal spectrum. Continuing with the example of an implanted closed-loop neurostimulation system for treating epilepsy, the centroid of the signal spectrum can be seen as shifting towards lower frequencies (i.e., a decrease in centroid value) during an epileptic seizure due to a significant increase in the energy of low-frequency components. The implanted closed-loop neurostimulation system calculates the current signal's centroid and compares it with the baseline centroid to determine the change.

[0043] In some embodiments, the characteristic change index further includes an eighth characteristic index based on the correlation coefficient between EEG signals from different channels and the EEG signal at the onset of the disease. An increased correlation coefficient (e.g., approaching 1) indicates enhanced synchronicity of electrical activity in different brain regions, suggesting that abnormal discharges are propagating or coupling between multiple brain regions. In some embodiments, the average correlation coefficient for all channel pairs may also be calculated.

[0044] Step 102: Determine the current processing strategy based on feature change indicators in at least two dimensions; wherein the processing strategy includes a first strategy and a second strategy.

[0045] In this step, the first strategy and the second strategy are two different strategies used to achieve the signal processing objective. For example, in an exemplary scenario of treating epilepsy, the first strategy and the second strategy can be two different treatment strategies or two different seizure levels. At least two dimensions of feature change indicators can be used as seizure risk assessment parameters to determine the current seizure level. In some embodiments, high and low thresholds can be set. If any dimension's seizure risk assessment parameter exceeds its high threshold, it is determined to be a first seizure level. If all dimension parameters are below the high threshold, but at least one seizure risk assessment parameter exceeds the low threshold, it is determined to be a second seizure level. In some embodiments, if all seizure risk assessment parameters are below the low threshold, it may be determined to be a seizure-free or mild seizure, without triggering stimulation. In some embodiments, a machine learning model can be used to fuse and classify the seizure risk assessment parameters to output the seizure level. In some embodiments, the seizure risk assessment parameters can also be weighted and summed to obtain an assessment score, which is then used to determine the seizure level.

[0046] Step 103: If the current processing strategy is the first strategy, generate a first control instruction; wherein the first control instruction is used to instruct the output of a first electrical stimulation signal according to a fixed first set of stimulation parameters.

[0047] In this step, upon execution of the first control command, the implanted closed-loop neurostimulation system will output a first electrical stimulation signal according to a fixed first set of stimulation parameters. For example, it will output an electrical stimulation signal according to a fixed set of stimulation parameters.

[0048] The following example uses the intracorporeal nerve stimulation system to treat epilepsy.

[0049] If the system determines that the current treatment strategy is the first strategy or the current seizure severity is the first severity level, it indicates that the epileptic seizure is severe and requires rapid and intensive intervention. Therefore, for the sake of caution, a fixed-parameter stimulation strategy will be adopted. In some embodiments, the fixed stimulation parameters include stimulation amplitude or intensity (e.g., 1 mA), stimulation frequency (e.g., 10-150 Hz), pulse width (e.g., 100-500 μs), and stimulation duration (e.g., 100-500 ms). In some embodiments, the fixed-parameter strategy does not require real-time adjustment of the stimulation parameters, thereby ensuring that abnormal discharges are suppressed within the shortest possible delay and preventing seizure escalation.

[0050] Step 104: When the current processing strategy is the second strategy, generate a second control instruction; wherein the second control instruction is used to instruct the output of a second electrical stimulation signal according to the dynamically adjusted second set of stimulation parameters; wherein each stimulation parameter in the dynamically adjusted second set of stimulation parameters is calculated based on the modulation of the target EEG signal by the previous stimulation parameter.

[0051] In this step, when the second control command is executed, the implanted closed-loop neurostimulation system will output a second electrical stimulation signal according to the dynamically adjusted second set of stimulation parameters. For example, the stimulation parameters are adjusted cyclically, and electrical stimulation is released based on the adjusted stimulation parameters until the target conditions are met and then the release of electrical stimulation is stopped.

[0052] In this embodiment, before outputting the stimulation signal, a multi-dimensional in-depth analysis is first performed on the EEG signal conforming to the target signal pattern. This involves calculating characteristic change indicators across multiple dimensions, such as frequency domain energy change, signal complexity change, signal amplitude change, propagation characteristics, dominant frequency, and spectral centroid. This allows for a more refined and comprehensive analysis and evaluation of the EEG signal state, leading to intelligent decision-making on the current processing strategy. A first strategy based on fixed stimulation parameters can be used to output the electrical stimulation signal, ensuring processing efficiency and reliability. Alternatively, a second strategy based on dynamic stimulation parameters can be employed. In the second strategy, each stimulation parameter is calculated based on the actual modulation of the target signal by the previous stimulation parameter, forming a real-time, adaptive parameter optimization loop. This application effectively overcomes the technical problems of existing implantable closed-loop neurostimulation systems, such as their single working mode and poor flexibility, by establishing an intelligent prediction and strategy switching mechanism based on multi-dimensional signal feature analysis.

[0053] In some embodiments, before calculating feature change indices in at least two dimensions based on the target EEG signal, the method further includes:

[0054] The first detection algorithm is used to detect the EEG signal to be detected, and the first detection result is obtained.

[0055] If the first detection result indicates that the signal pattern matches the target signal pattern, the EEG signal to be detected is detected again using a second detection algorithm. The sensitivity of the first detection algorithm is higher than that of the second detection algorithm, and the accuracy of the second detection algorithm is higher than that of the first detection algorithm.

[0056] If the second detection result indicates the target signal pattern, the EEG signal to be detected is identified as the target EEG signal.

[0057] It should be noted that the first detection algorithm performs real-time analysis and detection on the continuously monitored raw EEG signals (the EEG signals to be detected) to obtain the first detection result. The first detection algorithm is characterized by its high sensitivity, meaning its primary goal is to capture as many possible events or features as possible, minimizing omissions. In some embodiments, the first detection algorithm may employ rules with low computational complexity and rapid response. For example, time-domain threshold detection: continuously monitoring the amplitude or energy of the EEG signal; when the signal amplitude exceeds a relatively low threshold for a short duration, a preliminary alarm is triggered. Another example is frequency-domain energy detection: calculating power in a specific frequency band (such as the Gamma band common in epileptic seizures or lower); if the power rises rapidly and exceeds a lenient threshold, it is deemed suspicious.

[0058] The first detection result can be a binary output, indicating whether it conforms to the target signal pattern or not.

[0059] Then, after the initial screening is passed, a second detection algorithm is activated for confirmation. For example, the system will only activate the second detection algorithm to re-detect the same or most recently input EEG signal if and only if the first detection result indicates that it matches the target signal pattern.

[0060] The characteristics of the second detection algorithm: This algorithm is designed to have high accuracy (i.e., high specificity). It verifies the preliminary results of the first detection algorithm and filters out false alarms. The accuracy of the second detection algorithm is higher than that of the first detection algorithm, while its sensitivity is generally lower. In some embodiments, the second detection algorithm may employ a more complex, computationally intensive, but more discriminative model. For example, a multi-feature fusion model: simultaneously extracting multiple features of the signal, such as time domain, frequency domain, and nonlinear dynamics, to construct a more comprehensive discriminative model. Another example is a machine learning / deep learning model: using a pre-trained classifier to analyze EEG signal segments.

[0061] The second detection result is also a binary output, indicating whether it matches or does not match the target signal pattern. If the second detection result indicates that it matches the target signal pattern, then the EEG signal matching the target signal pattern has been identified. At this point, the double verification is complete.

[0062] In this embodiment, by constructing a tandem detection pipeline of "high-sensitivity screening + high-accuracy confirmation", the reliability of EEG signal detection is greatly improved without significantly increasing the average power consumption and latency of the system.

[0063] In some embodiments, the method further includes:

[0064] Given that the current processing strategy is the second strategy and the fourth characteristic index is greater than the target velocity threshold, a third control command is generated; wherein, the third control command is used to instruct the output of a third electrical stimulation signal in a manner that increases the stimulation frequency.

[0065] It should be noted that the fourth characteristic index can represent the change in the propagation speed of the target EEG signal. If this characteristic index is too high, it indicates that the propagation speed is very fast, requiring rapid and forceful intervention. For example, in the exemplary scenario of treating epilepsy using an in-circuit neurostimulation system, a forceful intervention method can be designed around the fourth characteristic index. For example, when a second strategy needs to be implemented, the propagation speed of the target EEG signal is compared to a target speed threshold. This target speed threshold can be an important clinical parameter; for example, it can be set by the physician before surgery or through an external device based on the patient's individual historical seizure data, epilepsy type, and brain anatomy. When the monitored propagation speed is greater than this target speed threshold, it indicates that the epilepsy is spreading very strongly in the brain network, generating a third control command. In response to the third control command, targeted parameter adjustments are made, i.e., increasing the input frequency of the electrical stimulation signal.

[0066] In some embodiments, the adjustment or increase of the stimulation frequency or output frequency can be stepwise, such as directly increasing by a fixed step (e.g., 20 Hz), or it can be increased by a preset ratio, such as increasing to 1.5 times the current frequency.

[0067] In this embodiment, the system can identify scenarios requiring strong intervention by using the key dynamic indicator of propagation speed, and promptly take measures to increase the frequency of stimulation for strong intervention.

[0068] In some embodiments, the method further includes:

[0069] When the current processing strategy is the second strategy and the fifth feature index is greater than the target range threshold, a fourth control instruction is generated; wherein, the fourth control instruction is used to indicate switching to a multi-point stimulation mode that covers the origin point of the target EEG signal.

[0070] It should be noted that the fifth characteristic indicator can represent changes in the propagation range of the target EEG signal. If this characteristic indicator is too large, it indicates an excessive impact, requiring prompt and forceful intervention. For example, in the exemplary scenario of treating epilepsy using an implanted closed-loop neurostimulation system, a forceful intervention approach can be designed around the fifth characteristic indicator. For instance, when a second strategy needs to be implemented, the propagation range of the target EEG signal is compared to the target range threshold. This target range threshold can be an important clinical parameter; for example, it can be set by the physician before surgery or via an external device based on the patient's individual historical seizure data, epilepsy type, and brain anatomy. When the monitored propagation range exceeds this target range threshold, it indicates that the abnormal electrical activity is no longer confined to a small focal point but has spread to a relatively wide brain region. At this point, a fourth control command is generated, and in response to the fourth control command, the implanted closed-loop neurostimulation system is switched to a multi-point stimulation mode covering the origin point of the target EEG signal.

[0071] In some embodiments, the process in response to the fourth control command includes:

[0072] By combining historical and real-time EEG data, the initial origin channel or origin point of the current seizure can be quickly located. Then, with this origin point as the center, multiple adjacent electrode channels around it are selected as electrode contacts for outputting electrical stimulation signals, and multiple electrodes are controlled to output electrical stimulation signals simultaneously or in a specific time sequence.

[0073] In some embodiments, when controlling multiple electrodes to output electrical stimulation signals, all selected electrodes can be controlled to output electrical pulses with the same or different parameters at the same time, forming a powerful electric field with a wider coverage, so as to suppress abnormal activity in multiple regions at the same time.

[0074] In some embodiments, when controlling multiple electrodes to output electrical stimulation signals, electrical stimulation signals can be output sequentially on different electrode pairs according to a preset sequence, effectively interrupting abnormal synchronization connections between different nodes.

[0075] In this embodiment, the system can identify scenarios requiring strong intervention by using the key dynamic indicator of propagation range, and promptly switch to multi-point stimulation mode for strong intervention.

[0076] In some embodiments, after generating the second control command, the method further includes:

[0077] In response to the second control command, the stimulation parameters are cyclically adjusted to generate a dynamically changing second set of stimulation parameters, and a second electrical stimulation signal is output according to the second set of stimulation parameters until the target condition is met and the output of the second electrical stimulation signal is stopped. The following steps are performed in each cycle:

[0078] Collect the EEG response signal after the most recent electrical stimulation signal output ends;

[0079] Calculate feature change indicators in at least two dimensions based on EEG response signals;

[0080] The adjustment amount of the stimulus parameters is determined based on the characteristic change indicators in at least two dimensions;

[0081] The stimulation parameters are adjusted according to the adjustment amount, and a second electrical stimulation signal is output based on the adjusted stimulation parameters.

[0082] It should be noted that after each electrical stimulation signal output, the implantable closed-loop neurostimulation system enters the data acquisition phase to ensure that the neural response after stimulation is captured. It can be understood that the EEG response signal can be the EEG signal acquired after the electrical stimulation signal output has ended. Since the EEG signal acquired at this time is the response of the nerve after receiving stimulation, it can also be called the EEG response signal.

[0083] Characteristic change indices can represent the state of a nerve after receiving stimulation. Therefore, appropriate adjustments to stimulation parameters can be determined based on these indices. For example, in an exemplary scenario of treating epilepsy, during a cycle, after each output of an electrical stimulation signal, the acquired electroencephalogram (EEG) response signal can determine the patient's state or the treatment effect after this electrical stimulation. In this embodiment, the treatment effect is measured using characteristic change indices across multiple dimensions. Ultimately, based on the treatment effect, the appropriate stimulation parameter values ​​or adjustment amounts are determined for the next output of an electrical stimulation signal.

[0084] In this embodiment, multi-dimensional information can be used to comprehensively and accurately assess the neural state after stimulation; and then the amount of stimulation parameter adjustment can be determined based on this comprehensive assessment to avoid overstimulation.

[0085] In some embodiments, the first feature index includes: the power change of the high-frequency oscillating FOS signal of the EEG response signal compared to the baseline EEG signal; the second feature index includes: the change of the sample entropy of the EEG response signal compared to the baseline EEG signal; the baseline EEG signal is the EEG signal before the most recent electrical stimulation output.

[0086] Based on characteristic change indicators in at least two dimensions, determine the adjustment amount of the stimulus parameters, including:

[0087] The evaluation score is obtained by weighted summing of the normalized values ​​of the first feature index and the normalized values ​​of the second feature index.

[0088] Based on the assessment scores, the adjustment amount of the stimulus parameters is determined; the adjustment amount of the stimulus parameters is negatively correlated with the assessment scores.

[0089] It should be noted that after electrical stimulation, the first characteristic indicator typically decreases, while the second characteristic indicator typically increases. Therefore, in some embodiments, the first characteristic indicator includes the power decrease of the HFOs signal compared to the symptom-related EEG signal, and the second characteristic indicator includes the increase of the sample entropy of the EEG response signal compared to the symptom-related EEG signal.

[0090] For details regarding HFOs signals, please refer to the description in the above embodiments, which will not be repeated here.

[0091] The following example uses the implantable circular nerve stimulation system to treat epilepsy.

[0092] The power drop of the HFOs signal can indicate the therapeutic effect or the inhibitory effect on epilepsy after the most recent output electrical stimulation signal. For example, a large power drop indicates a good therapeutic effect. The power drop of the HFOs signal can be calculated in the following way:

[0093] First, baseline power calculation: After detecting the onset of EEG signals and before outputting the first electrical stimulation signal, a segment of the onset EEG signal (e.g., lasting 500 milliseconds) is extracted, then the HFOs signal components are extracted, and their average power is calculated, denoted as . .

[0094] Second, response power calculation: Within the first time window after the first electrical stimulation signal output ends, a segment of the EEG response signal is extracted, and the average power of its HFOs component is calculated using the same signal processing method, denoted as... .

[0095] Third, the calculation of the decrease value: the first characteristic indicator. This is the decrease in power, calculated using the following formula: . It is a positive value; the larger the value, the better the effect of the first electrical stimulation signal on suppressing abnormal high-frequency oscillations.

[0096] Similarly, the relevant content regarding sample entropy can be found in the description of the above embodiments, and will not be repeated here. An increase in sample entropy can indicate the therapeutic effect or the inhibitory effect on epilepsy after the most recent output electrical stimulation signal. For example, a large increase in sample entropy indicates a very good therapeutic effect.

[0097] The evaluation score is a parameter for measuring treatment effectiveness, calculated by comprehensively considering the first and second characteristic indicators. In some embodiments, different weights may be pre-assigned to the first and second characteristic indicators, and then the evaluation score may be calculated using a weighted summation method.

[0098] In this embodiment, the first feature index and the second feature index can be normalized first, and then the evaluation score can be calculated based on the normalized value.

[0099] In some embodiments, the normalized value of the first feature index The following formula is used for calculation:

[0100] Formula 1: ;in, Indicates the first characteristic index, This represents the baseline power mentioned above.

[0101] Similarly, the calculation process for the normalized value of the second characteristic indicator is similar to that for the first characteristic indicator, and will not be repeated here.

[0102] In some embodiments, the evaluation score can be calculated using the following formula two.

[0103] Formula 2: ;

[0104] in, and The preset weighting coefficients are used, and they satisfy the following conditions: . This is the normalized value of the first characteristic index. This is the normalized value of the second characteristic indicator. The weighting coefficient can be adjusted according to different epilepsy types or patient specificity. For example, for temporal lobe epilepsy where HFOs are known to be highly correlated with seizures, a weighting can be set... For patients with more significant complex changes in EEG during an attack, a more targeted approach can be adopted. .

[0105] "Negative correlation" means that the lower the current assessment score, the less ideal the outcome. For example, in the epilepsy treatment scenario, a lower current assessment score indicates a worse therapeutic effect. Correspondingly, the magnitude of the parameter adjustment (adjustment amount) will be larger. Conversely, a higher current assessment score indicates a more ideal outcome or that the therapeutic effect has improved significantly and is close to the target, so the adjustment magnitude will be smaller.

[0106] For example, when adjusting the stimulation parameters in each cycle, the system does not directly use the preset fixed step size. Instead, it first uses the evaluation score updated after the output of the electrical stimulation signal in the current cycle as a key feedback signal to dynamically calculate the parameter adjustment amount for this cycle.

[0107] In some embodiments, when the evaluation score after an output electrical stimulation signal is extremely low, it indicates that the current stimulation parameters are completely ineffective or severely insufficient. In this case, a larger adjustment step size (such as significantly increasing the stimulation intensity) can be adopted to quickly jump out of the ineffective parameter region and avoid wasting time and energy in the ineffective region.

[0108] In some embodiments, when the evaluation score after a certain output electrical stimulation signal is already at a high level (e.g., close to but not reaching the threshold for stopping treatment), it indicates that the current stimulation parameters are very close to the optimal value. At this time, automatically switching to the small step fine-tuning mode can effectively avoid overstimulation or energy waste that may be caused by excessively large step sizes.

[0109] In this embodiment of the application, this intelligent adjustment strategy can quickly jump out of the invalid parameter area, avoiding wasting time and energy in the invalid area; it can also effectively avoid overstimulation or energy waste that may be caused by excessive step size.

[0110] In some embodiments, an artifact removal time window precedes the first time window for acquiring EEG response signals during each cycle.

[0111] Accordingly, in each cycle, after outputting a second electrical stimulation signal based on the adjusted stimulation parameters, the method further includes:

[0112] Determine the mapping relationship between stimulus parameter intervals and time windows; wherein the average value of the stimulus parameter interval is positively correlated with the duration of the time window;

[0113] The time window corresponding to the target stimulus parameter range is selected as the artifact removal time window, where the target stimulus parameter range includes the adjusted stimulus parameters.

[0114] It should be noted that when the implanted closed-loop neurostimulation system outputs an electrical stimulation signal, the EEG signal immediately following the end of the stimulation is overwhelmed by significant stimulation artifacts. This portion of the signal cannot accurately reflect the physiological response of the neural tissue. If EEG signals are collected at this stage to calculate characteristic change indicators, the characteristic change indicators will be distorted to some extent.

[0115] To this end, this embodiment introduces and defines a key time interval—the artifact elimination time window. After the electrical stimulation signal output ends, the implantable closed-loop neurostimulation system waits for a specific period of time until the stimulation artifacts have sufficiently decayed to a negligible level before opening the first time window to collect pure and reliable EEG response signals.

[0116] In some embodiments, the process of determining the mapping relationship between stimulus parameter ranges and time windows may include:

[0117] An implantable closed-loop neurostimulation system pre-stores or dynamically maintains a mapping table or mapping function. This mapping table defines the correspondence between stimulation parameter ranges and time window durations. The stimulation parameter range refers to several consecutive sub-ranges formed by dividing the numerical range of one or more stimulation parameters. For example, the stimulation intensity range (e.g., 0.5mA to 10mA) can be divided into three ranges: [0.5mA, 3mA], [3mA, 6mA], and [6mA, 10mA]. Alternatively, a range within a multi-dimensional parameter space can be defined based on multiple parameters (e.g., intensity and frequency).

[0118] The duration of the time window, i.e., the specific duration of the artifact elimination time window, can be measured in milliseconds (ms). The principle for establishing the mapping relationship is that the average value of the stimulation parameter range is positively correlated with the duration of the time window. It can be understood that the stimulation parameters collectively determine the total charge injected into the tissue in a single stimulus. The higher the stimulation energy, the larger the amplitude of the generated electrical stimulation artifact, and the longer it takes for it to decay to the baseline level. Therefore, a longer artifact elimination time window needs to be allocated to a parameter range representing a higher energy level (with a larger average value).

[0119] The system selects the time window corresponding to the target stimulation parameter range as the artifact removal time window. After each output of an electrical stimulation signal, the system immediately performs this selection operation: comparing the currently used stimulation parameter with a pre-stored stimulation parameter range to determine the specific range to which the stimulation parameter belongs. The specific range containing the stimulation parameter is defined as the target stimulation parameter range. The system then queries the duration of the artifact removal time window uniquely corresponding to the target stimulation parameter range. Using this queryed duration, a timer is started at the end of the stimulation. Before this timer expires, the data acquisition channel remains closed or the acquired data is ignored. Once the timer expires, the system immediately opens the first time window and begins acquiring EEG response signals.

[0120] In this embodiment, an intelligent and personalized artifact removal strategy is achieved by establishing a dynamic mapping between stimulus parameters and artifact removal windows. For high-intensity stimuli, a sufficiently long cooling time is provided to ensure that artifacts do not affect the evaluation; for low-intensity stimuli, unnecessary excessive waiting is avoided.

[0121] In some embodiments, the target condition includes at least one of the following:

[0122] The number of electrical stimulation signals output within the first duration reaches the threshold.

[0123] The total amount of charge injected within the second time period reaches the charge threshold;

[0124] The time elapsed since the first output of the electrical stimulation signal has reached the duration threshold.

[0125] It should be noted that the first duration can refer to a preset time window. For example, it can be set to 60 seconds or 180 seconds. Similarly, the second duration is the same as the first duration; the second duration can be the same as or different from the first duration.

[0126] The number of stimulations threshold refers to the maximum total number of electrical stimulation signals allowed to be output within a first duration. For example, the threshold could be set to 5 times. The charge threshold refers to the safe upper limit of the total amount of charge allowed to be injected into brain tissue via electrical stimulation within a second duration.

[0127] The threshold for the number of stimuli can serve as a safety boundary to achieve safety control and prevent excessive stimulation from causing nerve tissue fatigue or potential damage. For example, in the context of epilepsy treatment, limiting the density of stimulation in a single epileptic seizure event can prevent the patient's brain from receiving excessive stimulation in a short period of time due to cyclical logic failure or persistent poor treatment efficacy.

[0128] The charge threshold condition can serve as a safety boundary to achieve safety control and avoid excessive stimulation that could lead to nerve tissue fatigue or potential damage.

[0129] The duration threshold refers to the maximum duration of the output electrical stimulation signal, starting from the first output signal. For example, the duration threshold could be set to 3 minutes. This condition can serve as a final safety condition. For instance, in epilepsy treatment, regardless of efficacy, the cycle would be forcibly terminated once the total treatment time reached this threshold. This ensures that treatment will not continue indefinitely, even in cases of abnormal counting or calculation, and also addresses epileptic seizures with abnormally prolonged durations.

[0130] In some embodiments, the target conditions are monitored using a logical "OR" relationship. That is, as soon as any one condition is met, the system immediately exits the loop and stops outputting subsequent electrical stimulation signals.

[0131] Based on the same principle as the method provided in the embodiments of this application, the embodiments of this application also provide an implantable closed-loop neurostimulation system, such as... Figure 2 As shown, it includes:

[0132] The feature index module 201 is used to calculate feature change indices in at least two dimensions based on the target EEG signal; wherein the target EEG signal is an EEG signal that conforms to the target signal pattern, and the feature change indices include multiple of the following: a first feature index based on the change of signal frequency domain energy in the target frequency band, a second feature index based on the change of signal complexity, a third feature index based on the change of signal amplitude, a fourth feature index based on the change of the propagation speed of the target EEG signal, a fifth feature index based on the change of the propagation range of the target EEG signal, a sixth feature index based on the change of the signal dominant frequency, and a seventh feature index based on the change of the signal spectral centroid;

[0133] The processing strategy module 202 is used to determine the current processing strategy based on feature change indicators in at least two dimensions; wherein the processing strategy includes a first strategy and a second strategy.

[0134] The first processing module 203 is used to generate a first control instruction when the current processing strategy is the first strategy; wherein the first control instruction is used to instruct the output of a first electrical stimulation signal according to a fixed first set of stimulation parameters;

[0135] The second processing module 204 is used to generate a second control instruction when the current processing strategy is the second strategy; wherein the second control instruction is used to instruct the output of a second electrical stimulation signal according to the dynamically adjusted second set of stimulation parameters; wherein each stimulation parameter in the dynamically adjusted second set of stimulation parameters is calculated based on the modulation of the target EEG signal by the previous stimulation parameter.

[0136] In some embodiments, the implantable closed-loop neurostimulation system further includes: a detection module, used for:

[0137] The first detection algorithm is used to detect the EEG signal to be detected, and the first detection result is obtained.

[0138] If the first detection result indicates that the signal pattern matches the target signal pattern, the EEG signal to be detected is detected again using a second detection algorithm. The sensitivity of the first detection algorithm is higher than that of the second detection algorithm, and the accuracy of the second detection algorithm is higher than that of the first detection algorithm.

[0139] If the second detection result indicates the target signal pattern, the EEG signal to be detected is identified as the target EEG signal.

[0140] In some embodiments, the implantable closed-loop neurostimulation system further includes:

[0141] The third processing module is used to generate a third control command when the current processing strategy is the second strategy and the fourth characteristic index is greater than the target speed threshold; wherein the third control command is used to instruct the output of a third electrical stimulation signal in a manner that increases the stimulation frequency.

[0142] In some embodiments, the implantable closed-loop neurostimulation system further includes:

[0143] The fourth processing module is used to generate a fourth control instruction when the current processing strategy is the second strategy and the fifth feature index is greater than the target range threshold; wherein, the fourth control instruction is used to instruct switching to a multi-point stimulation mode that covers the origin point of the target EEG signal.

[0144] In some embodiments, the implantable closed-loop neurostimulation system further includes:

[0145] The response module, in response to the second control command, cyclically adjusts the stimulation parameters to generate a dynamically changing second set of stimulation parameters, and outputs a second electrical stimulation signal according to the second set of stimulation parameters until the target condition is met and the output of the second electrical stimulation signal is stopped. The following steps are executed in each cycle:

[0146] Collect the EEG response signal after the most recent electrical stimulation signal output ends;

[0147] Calculate feature change indicators in at least two dimensions based on EEG response signals;

[0148] The adjustment amount of the stimulus parameters is determined based on the characteristic change indicators in at least two dimensions;

[0149] The stimulation parameters are adjusted according to the adjustment amount, and a second electrical stimulation signal is output based on the adjusted stimulation parameters.

[0150] In some embodiments, the first feature index includes: the power change of the high-frequency oscillating FOS signal of the EEG response signal compared to the baseline EEG signal; the second feature index includes: the change of the sample entropy of the EEG response signal compared to the baseline EEG signal; the baseline EEG signal is the EEG signal before the most recent electrical stimulation output.

[0151] The response module is specifically used for:

[0152] The evaluation score is obtained by weighted summing of the normalized values ​​of the first feature index and the normalized values ​​of the second feature index.

[0153] Based on the assessment scores, the adjustment amount of the stimulus parameters is determined; the adjustment amount of the stimulus parameters is negatively correlated with the assessment scores.

[0154] In some embodiments, an artifact removal time window precedes the first time window for acquiring EEG response signals during each cycle.

[0155] In some embodiments, the implantable closed-loop neurostimulation system further includes:

[0156] The mapping module is used to determine the mapping relationship between stimulus parameter intervals and time windows; wherein, the average value of the stimulus parameter interval is positively correlated with the duration of the time window;

[0157] The mapping selection module is used to select the time window corresponding to the target stimulus parameter range as the artifact elimination time window, wherein the target stimulus parameter range includes the adjusted stimulus parameters.

[0158] In some embodiments, the target condition includes at least one of the following:

[0159] The number of electrical stimulation signals output within the first duration reaches the threshold.

[0160] The total amount of charge injected within the second time period reaches the charge threshold;

[0161] The time elapsed since the first output of the electrical stimulation signal has reached the duration threshold.

[0162] The implantable closed-loop neurostimulation system provided in this application embodiment can achieve… Figure 1 The various processes implemented in the method embodiments are not described in detail here to avoid repetition.

[0163] The implantable closed-loop neurostimulation system provided in this application first performs multi-dimensional in-depth analysis on the EEG signal conforming to the target signal pattern before outputting the stimulation signal. This analysis calculates characteristic change indicators in multiple dimensions, including frequency domain energy changes, signal complexity changes, signal amplitude changes, propagation characteristics, dominant frequency, and spectral centroid. This allows for a more refined and comprehensive analysis and evaluation of the EEG signal state, and then intelligently decides on the current processing strategy. A first strategy based on fixed stimulation parameters can be used to output the electrical stimulation signal, ensuring processing efficiency and reliability. A second strategy based on dynamic stimulation parameters can also be used. In the second strategy, each stimulation parameter is calculated and generated based on the actual modulation of the target signal by the previous stimulation parameter, forming a real-time, adaptive parameter optimization loop. This application effectively overcomes the technical problems of existing implantable closed-loop neurostimulation systems, such as their single working mode and poor flexibility, by establishing an intelligent prediction and strategy switching mechanism based on multi-dimensional signal feature analysis.

[0164] The implantable closed-loop neurostimulation system of this application embodiment can execute the signal processing method provided in the embodiment of this application. The implementation principle is similar. The actions performed by each module and unit in the implantable closed-loop neurostimulation system in each embodiment of this application correspond to the steps in the signal processing method in each embodiment of this application. For detailed functional descriptions of each module of the implantable closed-loop neurostimulation system, please refer to the descriptions in the corresponding signal processing methods shown above. They will not be repeated here.

[0165] Based on the same principles as the methods shown in the embodiments of this application, the embodiments of this application also provide a signal processing system, which includes the implantable closed-loop neurostimulation system provided in the above embodiments.

[0166] In an alternative embodiment, a signal processing system, such as Figure 3 As shown, Figure 3 The signal processing system 3000 shown includes a processor 3001 and a memory 3003. The processor 3001 and the memory 3003 are connected, for example, via a bus 3002. Optionally, the signal processing system 3000 may further include a transceiver 3004, which can be used for data interaction between the signal processing system and other electronic devices, such as data transmission and / or data reception. It should be noted that in practical applications, the transceiver 3004 is not limited to one type, and the structure of this signal processing system 3000 does not constitute a limitation on the embodiments of this application.

[0167] Processor 3001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 3001 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0168] Bus 3002 may include a pathway for transmitting information between the aforementioned components. Bus 3002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 3002 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0169] The memory 3003 may be ROM (Read Only Memory) or other types of static storage devices capable of storing static information and instructions, RAM (Random Access Memory) or other types of dynamic storage devices capable of storing information and instructions, or EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media, other magnetic storage devices, or any other medium capable of carrying or storing computer programs and capable of being read by a computer, without limitation herein.

[0170] The memory 3003 is used to store computer programs that execute the embodiments of this application, and the execution is controlled by the processor 3001. The processor 3001 is used to execute the computer programs stored in the memory 3003 to implement the steps shown in the foregoing method embodiments.

[0171] This application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it can implement the steps and corresponding content of the aforementioned method embodiments.

[0172] This application also provides a computer program product, including a computer program that, when executed by a processor, can implement the steps and corresponding content of the aforementioned method embodiments.

[0173] The terms "first," "second," "third," "fourth," "1," "2," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in a sequence other than that shown in the illustrations or text descriptions.

[0174] It should be understood that although arrows indicate various operation steps in the flowcharts of this application's embodiments, the order in which these steps are implemented is not limited to the order indicated by the arrows. Unless explicitly stated herein, in some implementation scenarios of this application's embodiments, the implementation steps in each flowchart can be executed in other orders as required. Furthermore, some or all steps in each flowchart, based on the actual implementation scenario, may include multiple sub-steps or multiple stages. Some or all of these sub-steps or stages can be executed at the same time, and each sub-step or stage can also be executed at different times. In scenarios where execution times differ, the execution order of these sub-steps or stages can be flexibly configured according to requirements, and this application's embodiments do not limit this.

[0175] The above description is only an optional implementation method for some implementation scenarios 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, The implantable closed-loop neurostimulation system includes: The feature index module is used to calculate feature change indices in at least two dimensions based on the target EEG signal; wherein the target EEG signal is an EEG signal that conforms to the target signal pattern, and the feature change indices include multiple of the following: a first feature index based on the change of signal frequency domain energy in the target frequency band, a second feature index based on the change of signal complexity, a third feature index based on the change of signal amplitude, a fourth feature index based on the change of the propagation speed of the target EEG signal, a fifth feature index based on the change of the propagation range of the target EEG signal, a sixth feature index based on the change of the signal dominant frequency, and a seventh feature index based on the change of the signal spectral centroid; A processing strategy module is used to determine the current processing strategy based on the feature change indicators of the at least two dimensions; wherein the processing strategy includes a first strategy and a second strategy. The first processing module is configured to generate a first control instruction when the current processing strategy is the first strategy; wherein the first control instruction is configured to instruct the output of a first electrical stimulation signal according to a fixed first set of stimulation parameters. The second processing module is used to generate a second control instruction when the current processing strategy is the second strategy; wherein the second control instruction is used to instruct the output of a second electrical stimulation signal according to a dynamically adjusted second set of stimulation parameters; wherein each stimulation parameter in the dynamically adjusted second set of stimulation parameters is calculated based on the modulation of the target EEG signal by the previous stimulation parameter; The response module is configured to respond to the second control command by cyclically adjusting the stimulation parameters to generate a dynamically changing second set of stimulation parameters, and outputting a second electrical stimulation signal according to the second set of stimulation parameters until the target condition is met and the output of the second electrical stimulation signal is stopped. The following steps are performed in each cycle: Collect the EEG response signal after the most recent electrical stimulation signal output ends; Based on the EEG response signal, calculate the feature change indicators in at least two dimensions; The adjustment amount of the stimulus parameters is determined based on the characteristic change indicators of at least two dimensions; The stimulation parameters are adjusted according to the stated adjustment amount, and a second electrical stimulation signal is output based on the adjusted stimulation parameters.

2. The implantable closed-loop neurostimulation system according to claim 1, characterized in that, The implantable closed-loop neurostimulation system further includes: a detection module, used for: The first detection algorithm is used to detect the EEG signal to be detected, and the first detection result is obtained. If the first detection result indicates that the target signal pattern is met, the EEG signal to be detected is detected again using a second detection algorithm to obtain a second detection result; the sensitivity of the first detection algorithm is higher than that of the second detection algorithm, and the accuracy of the second detection algorithm is higher than that of the first detection algorithm. If the second detection result indicates a target signal pattern, the EEG signal to be detected is determined to be the target EEG signal.

3. The implantable closed-loop neurostimulation system according to claim 1, characterized in that, The implantable closed-loop neurostimulation system also includes: The third processing module is used to generate a third control command when the current processing strategy is the second strategy and the fourth characteristic index is greater than the target speed threshold; wherein the third control command is used to instruct the output of a third electrical stimulation signal in a manner that increases the stimulation frequency.

4. The implantable closed-loop neurostimulation system according to claim 1, characterized in that, The implantable closed-loop neurostimulation system also includes: The fourth processing module is used to generate a fourth control instruction when the current processing strategy is the second strategy and the fifth feature index is greater than the target range threshold; wherein the fourth control instruction is used to instruct switching to a multi-point stimulation mode covering the origin point of the target EEG signal.

5. The implantable closed-loop neurostimulation system according to claim 1, characterized in that, The first feature index includes: the power change of the high-frequency oscillating (HFOs) signal of the EEG response signal compared to the baseline EEG signal; the second feature index includes: the change of the sample entropy of the EEG response signal compared to the baseline EEG signal; the baseline EEG signal is the EEG signal before the most recent electrical stimulation output. The response module is specifically used for: The evaluation score is obtained by weighted summing of the normalized values ​​of the first feature index and the normalized values ​​of the second feature index. Based on the assessment score, the adjustment amount of the stimulus parameter is determined; wherein the adjustment amount of the stimulus parameter is negatively correlated with the assessment score.

6. The implantable closed-loop neurostimulation system according to claim 1, characterized in that, In each cycle, there is an artifact removal time window before the first time window for acquiring the EEG response signal.

7. The implantable closed-loop neurostimulation system according to claim 6, characterized in that, The implantable closed-loop neurostimulation system also includes: The mapping module is used to determine the mapping relationship between stimulus parameter intervals and time windows; wherein the average value of the stimulus parameter interval is positively correlated with the duration of the time window; The mapping selection module is used to select the time window corresponding to the target stimulus parameter range as the artifact elimination time window, wherein the target stimulus parameter range includes the adjusted stimulus parameters.

8. The implantable closed-loop neurostimulation system according to claim 1, characterized in that, The target condition includes at least one of the following: The number of electrical stimulation signals output within the first duration reaches the threshold. The total amount of charge injected within the second time period reaches the charge threshold; The time elapsed since the first output of the electrical stimulation signal has reached the duration threshold.

9. A signal processing system, characterized in that, Including the implantable closed-loop neurostimulation system as described in any one of claims 1 to 8.

Images