Implantable closed-loop nerve stimulation system and data storage method thereof

Abstract

The invention provides an implantable closed-loop nerve stimulation system and a data storage method thereof, and the data storage method comprises the steps: carrying out the decomposition processing of continuously collected multi-channel electroencephalogram signals, and obtaining signal components of a plurality of frequency bands; determining a first value of a five-dimensional feature based on the plurality of signal components; when the occurrence of a neural anomaly event is determined based on the first value of the five-dimensional feature, determining a first level corresponding to the neural anomaly event based on the first value of the five-dimensional feature; based on a first level corresponding to the neural anomaly event, related data of the neural anomaly event is stored in a structured storage partition, and the structured storage partition comprises a cache region and an event region; wherein the cache region is used for caching an original electroencephalogram signal and a five-dimensional feature value of a neural abnormal event within a preset duration; the event area comprises a plurality of sub-areas, and the plurality of sub-areas are used for storing related data of different levels of neural anomaly events.

Description

Technical Field

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

[0002] Implantable closed-loop neurostimulation systems acquire and analyze multi-channel EEG signals, automatically applying electrical stimulation to intervene when abnormal neural activity is detected. However, multi-channel EEG signals generate a large amount of data every second, and long-term storage not only consumes significant storage resources but also increases the complexity of subsequent analysis.

[0003] In addition, existing systems typically employ a simple first-in-first-out (FIFO) overlay strategy, storing only raw EEG signals or simple features, lacking the ability to classify and store EEG signals from complex neurological abnormalities.

[0004] Therefore, there is an urgent need for an implantable closed-loop neurostimulation system that can efficiently process, intelligently classify, and optimize the storage of multi-channel EEG signals and related data. Summary of the Invention

[0005] This application addresses the shortcomings of existing methods by proposing an implantable closed-loop neurostimulation system and its data storage method, aiming to solve the problem that existing implantable systems cannot efficiently process, intelligently classify, and optimize the storage of multi-channel EEG signals and related data.

[0006] In a first aspect, embodiments of this application provide a data storage method, the method comprising: The continuously acquired multi-channel EEG signals are decomposed and processed to obtain signal components in multiple frequency bands; Based on multiple signal components, a first value of a five-dimensional feature is determined, the five-dimensional feature including: time-frequency coherence, spatial gradient, temporal gradient, propagation consistency, and duration; When a neural abnormality event is determined to occur based on the first value of the five-dimensional features, the first level corresponding to the neural abnormality event is determined based on the first value of the five-dimensional features. Based on the first level corresponding to the neural abnormal event, the relevant data of the neural abnormal event is stored in a structured storage partition, which includes: a cache area and an event area; The buffer area is used to cache the raw EEG signals and five-dimensional feature values ​​of neural abnormal events within a preset time period, and the buffer area adopts a circular coverage strategy; the event area includes multiple sub-areas, which are used to store relevant data of neural abnormal events of different levels; the event area adopts a classification coverage strategy, which includes: for data of the same level, a first-in-first-out coverage strategy is used for storage; for data of different levels, higher-level data cannot be covered by lower-level data.

[0007] Secondly, embodiments of this application provide an implantable closed-loop neurostimulation system, comprising: The signal acquisition module is used to acquire multi-channel EEG signals; The processor is used to decompose and process continuously acquired multi-channel EEG signals to obtain signal components in multiple frequency bands, and determine a first value of a five-dimensional feature based on the multiple signal components. When a neurological abnormality event is determined based on the first value of the five-dimensional feature, a first level corresponding to the neurological abnormality event is determined based on the first value of the five-dimensional feature. The five-dimensional feature includes: time-frequency coherence, spatial gradient, temporal gradient, propagation consistency, and duration. The controller is used to store the relevant data of the neural abnormal event to a structured storage partition based on the first level corresponding to the neural abnormal event. The structured storage partition includes a cache area and an event area. The buffer area is used to cache the raw EEG signals and five-dimensional feature values ​​of neural abnormal events within a preset time period, and the buffer area adopts a circular coverage strategy; the event area includes multiple sub-areas, which are used to store relevant data of neural abnormal events of different levels; the event area adopts a classification coverage strategy, which includes: for data of the same level, a first-in-first-out coverage strategy is used for storage; for data of different levels, higher-level data cannot be covered by lower-level data.

[0008] Thirdly, embodiments of this application also disclose a signal processing system, including the implantable closed-loop neurostimulation system as described in the second aspect.

[0009] Fourthly, embodiments of this application also disclose a computer-readable storage medium storing a computer program that, when executed by a processor, implements one or more of the data storage methods described in the embodiments of the first aspect of this application.

[0010] Fifthly, embodiments of this application also disclose a computer program product, including a computer program that, when executed by a processor, implements one or more data storage methods as described in the embodiments of the first aspect of this application.

[0011] The beneficial technical effects of the technical solutions provided in this application include: The solution in this application decomposes and processes continuously acquired multi-channel EEG signals to obtain signal components in multiple frequency bands, and calculates five-dimensional features based on these signal components, thereby achieving a comprehensive characterization of neurological abnormal events and improving the accuracy of neurological abnormal event detection.

[0012] By determining the level of neural abnormal events based on the first value of five-dimensional features, it is possible to distinguish neural abnormal events of different severity, achieve intelligent classification, and then adopt different storage strategies to optimize the existing single storage strategy.

[0013] In addition, the cache area employs a circular overlay strategy: it caches the raw EEG signals and five-dimensional feature values ​​of neurological abnormal events within a preset time period, ensuring that important data in the short term is not lost, while avoiding unlimited growth in storage requirements. The event area uses a categorized overlay strategy: different overlay rules are applied to data of different levels (high-level data cannot be overwritten by low-level data), ensuring that data from important events is prioritized for preservation, thereby increasing the value density of the data. This hierarchical storage strategy not only saves storage resources but also improves the efficiency of data retrieval and analysis.

[0014] 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

[0015] 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 flowchart illustrating a data storage method for an implantable closed-loop neurostimulation system. Figure 2 This is a schematic diagram of the structure of an implantable closed-loop neurostimulation system provided in an embodiment of this application; Figure 3 This is a schematic diagram of the signal processing 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.

[0016] 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."

[0017] 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.

[0018] To facilitate understanding of the technical solution of this application, the following terms will be introduced.

[0019] 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.

[0020] 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.

[0021] Implantable closed-loop neurostimulation systems acquire and analyze multi-channel EEG signals, automatically applying electrical stimulation to intervene when abnormal neural activity is detected. However, multi-channel EEG signals generate a large amount of data every second, and long-term storage not only consumes significant storage resources but also increases the complexity of subsequent analysis.

[0022] In addition, existing systems often treat all abnormal events the same, failing to distinguish between mild abnormalities and severe outbreaks, which may lead to important information being ignored or covered up. Therefore, they lack an intelligent classification mechanism.

[0023] Furthermore, existing systems typically employ a simple first-in-first-out (FIFO) coverage strategy, storing not only strategy units but also raw EEG signals or simple features. Therefore, they lack the ability to classify and store EEG signals for complex neurological abnormalities.

[0024] Therefore, this application proposes an implantable closed-loop neurostimulation system capable of efficiently processing, intelligently classifying, and optimizing the storage of multi-channel EEG signals and related data. It aims to solve the following technical problems: 1. How to accurately determine the existence and severity of neurological abnormalities based on features extracted from continuously acquired multi-channel EEG signals.

[0025] 2. How to achieve effective storage of neurological abnormalities of varying severity with limited storage resources.

[0026] 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.

[0027] The implantable closed-loop neurostimulation system and its data storage method proposed in this application will be described in detail below with reference to the accompanying drawings.

[0028] In some embodiments, a data storage method for an implantable closed-loop neurostimulation system is provided. For example... Figure 1 As shown, the method includes: S1. The continuously acquired multi-channel EEG signals are decomposed and processed to obtain signal components of multiple frequency bands.

[0029] In some embodiments, the EEG signal of each channel can be decomposed into five sub-bands (e.g., 5 layers, yielding δ (1~4 Hz), θ (4~8 Hz), α (8~13 Hz), β (13~30 Hz), and γ (30~80 Hz)) to obtain the signal components of each frequency band. .

[0030] In some embodiments, after obtaining the signal components of each frequency band, the instantaneous energy can also be calculated for each frequency band signal, for example: energy. Among them, the energy in the γ band (30–80 Hz) is highly correlated with high-frequency oscillations (HFOs). It can also help determine key time points of events. t peak Optionally, calculate the time curve of total brain energy in the gamma band:

[0031] And the moment of the global maximum value is taken as the key time point of the event: This point in time represents the most active period of the event and is used for subsequent extraction of spatial and temporal features. It should be understood that total brain energy refers to the energy available at a given moment. t The scalar is obtained by adding the instantaneous energy values ​​of all effective EEG channels (or cortical contacts) in a specific frequency band (usually the gamma band, 30–80 Hz).

[0032] S2. Determine the first value of the five-dimensional feature based on multiple signal components.

[0033] In some embodiments, the five-dimensional features include: time-frequency coherence, spatial gradient, temporal gradient, propagation consistency, and duration.

[0034] It should be understood that in the embodiments of this application, the five-dimensional features are a set of event-level feature values ​​generated for each candidate neural abnormal event window after the collaborative analysis of multi-channel and multi-band signal components. These features are used to characterize the synchronicity, focality, dynamics, propagation patterns, and persistence of the neural abnormal event.

[0035] Optionally, time-frequency coherence is used to characterize the intensity of synchronous oscillations of multiple channels within a specific time-frequency region.

[0036] Optionally, spatial gradients are used to characterize the spatial focality of abnormal neural discharges on the scalp or cortex (e.g., the larger the gradient, the more focal the origin).

[0037] Optionally, the time gradient is used to characterize the steepness of the rise or fall of abnormal neural discharge energy (e.g., a rapid rise indicates strong abnormality).

[0038] Optionally, propagation consistency is used to determine whether an abnormal rhythm is repeatedly propagated along a fixed path.

[0039] Optionally, the duration is the duration of the abnormal neural discharge.

[0040] In some embodiments, time-frequency coherence can be obtained through the following process: 1. Calculate all channel pairs (i, j) in the frequency band using the following formula. f Complex wavelet coherence values ​​on: , in, W i ( t , f ) is a channel i Continuous wavelet transform, tIndicates time average.

[0041] 2. Normalize the largest complex wavelet coherence value among all channel pairs to obtain the time-frequency coherence. Ctf .

[0042] It should be understood that high Ctf Characterizes synchronous discharge in multiple regions (or multiple channels).

[0043] In some embodiments, the spatial gradient can be obtained through the following process: 1. In t peak At each moment, the energy of the γ band of each channel is collected: .

[0044] 2. For each channel i Calculate the maximum energy difference between it and the neighboring channels: , in, N ( i ) is a predefined spatial adjacency set.

[0045] 3. Take the global maximum value and normalize it: .

[0046] It should be understood that high G The s characterization indicates that energy is concentrated in a few channels, which suggests that abnormal neural events are highly focal.

[0047] In some embodiments, the time gradient can be obtained through the following process: 1. In t peak Backtracking to the previous time (e.g., 2 seconds), calculate the rising slope of the total energy of the whole brain in the γ band: , 2. Normalize the slope of the rising edge. Optionally, calculate the ratio of this rising edge slope to the baseline standard deviation during non-ictal periods, and take the smaller of this ratio and the value in step 1. Gt .

[0048] Optional, high Gt The sudden surge in energy indicates that the onset of abnormal neural events is rapid.

[0049] In some embodiments, the propagation consistency index can be obtained through the following process: 1. Use the cross-correlation delay method to estimate the activation timing between each pair of channels; Optionally, for any two channels i and j Extract signals within the event window x i( t ), x j ( t (e.g., in the γ band), calculate the cross-correlation function: , Find Rij ( τ Maximum delay τ ij .if τ ij >0, then channel j Lagging behind the channel i The direction of propagation is from the channel i To the passage j ;if τ ij If the value is less than 0, the propagation direction is from the channel. j To the passage i .

[0050] 2. Construct a propagation direction map based on the estimated activation timing.

[0051] Optionally, each channel can be treated as a node in a graph, and for each pair of channels, according to τ ij Add a directed edge to the symbol (e.g.) i → j Where, the edge weight = | τ ij The reciprocal of |.

[0052] 3. Calculate the path consistency index P c (e.g., most paths point to the same region), output normalized value P c∈[0,1].

[0053] Optionally, the event window can be divided into three segments: before, during, and after. A propagation graph can be constructed for each segment, and the overlap rate of the starting points of the three graphs can be calculated as... P c.

[0054] Optional, P c = 1 indicates highly stable propagation.

[0055] In some embodiments, the duration D can be obtained through the following process: 1. Calculate the duration: , in, As the starting point of the event: when any three-dimensional feature (e.g., : Ctf , Gs , Gt At the same time, they exceed their respective baseline values ​​+2σ. The event termination point is defined as the time when all features fall back to their respective baseline values ​​+1σ and remain below this value for more than 500 ms.

[0056] 2. Normalization process: , The 5-second time is a normalized reference value.

[0057] S3. When a neural abnormality event is determined to have occurred based on the first value of the five-dimensional features, the first level corresponding to the neural abnormality event is determined based on the first value of the five-dimensional features.

[0058] In some embodiments, a weighted arithmetic mean can be taken from the first values ​​of the five-dimensional features to calculate the first level corresponding to the neurological abnormality event.

[0059] Optional, grade S= w 1× Ctf + w 2× Gs + w 3× Gt + w 4× Pc + w 5× D .

[0060] in, Ctf , Gs , Gt , Pc , D These are the normalized time-frequency coherence, spatial gradient, temporal gradient, propagation consistency, and duration, respectively.

[0061] It should be understood that, in the embodiments of this application, weights w 1 to w 5 represents the dynamic weights, with initial values ​​of 0.25, 0.20, 0.15, 0.25, and 0.15, respectively, and satisfying the following conditions: w 1+ w 2 + w 3+ w 4+ w 5=1. When the propagation consistency is greater than the first dynamic threshold, w 4. Increased to 0.30; when the spatial focus of the origin heatmap is strong, w 2. Increased to 0.25; when the duration exceeds 5 seconds, w The value was increased from 5 to 0.20. Optionally, strong spatial focus of the epilepsy origin heatmap means that the area with a thermal value exceeding a preset threshold is limited to a single anatomical brain region, and the area of ​​that region does not exceed 15% of the total brain surface area.

[0062] It should also be understood that in the embodiments of this application, S∈[0,1], and the higher the value of S, the more typical and severe the abnormal neural event is, and the higher the level.

[0063] In some embodiments, prior to step S3, the method may further include: The first value of the five-dimensional features is used to determine whether a neurological abnormality has occurred.

[0064] In some embodiments, the occurrence of a neurological abnormality event can be determined through the following process: Determine whether the first values ​​for propagation consistency and duration are both greater than their respective thresholds; If so, then determine whether the first values ​​of time-frequency coherence, spatial gradient, and temporal gradient are all greater than their respective thresholds; If so, then a neurological event has been confirmed.

[0065] In the above embodiments, based on five-dimensional features, a two-level judgment logic including cascaded preconditions and parallel joint conditions is used to identify neural abnormal events. Only when both the cascaded preconditions and the parallel joint conditions are met is it determined to be a neural abnormal event, which can trigger subsequent data storage operations.

[0066] Optionally, the cascaded preconditions include: determining whether the propagation consistency is greater than the first dynamic threshold and whether the duration is greater than the second dynamic threshold; the parallel joint determination includes: if the cascaded preconditions are met, further determining whether the time-frequency coherence, spatial gradient and temporal gradient are all greater than their respective corresponding dynamic thresholds adaptively updated based on historical data during non-epidemic periods.

[0067] S4. Based on the first level corresponding to the neurological abnormal event, store the relevant data of the neurological abnormal event in the structured storage partition.

[0068] Optionally, structured storage partitions include: cache area and event area.

[0069] In some embodiments, the cache area is used to cache the raw EEG signals and five-dimensional feature values ​​of neurological abnormal events within a preset time period. Optionally, the cache area adopts a circular coverage strategy. In some embodiments, the event area includes multiple sub-areas, which are used to store relevant data of neurological abnormal events at different levels. Optionally, the event area adopts a classification coverage strategy. Optionally, the classification coverage strategy includes: for data of the same level, a first-in-first-out (FIFO) coverage strategy is used for storage; for data of different levels, higher-level data cannot be overwritten by lower-level data.

[0070] In some embodiments, when the total storage occupancy of the event area exceeds a preset upper limit, the data stored in the sub-area corresponding to the lower level is automatically compressed or metadata processed.

[0071] Optionally, metadata processing refers to deleting the original EEG signals and derived data (visualization) and generating structured metadata records containing timestamps, five-dimensional feature values, event levels, key channel identifiers, and judgment markers to replace the original event data for long-term storage.

[0072] In some embodiments, the plurality of sub-regions includes: a first sub-region, a second sub-region, and a third sub-region.

[0073] In some embodiments, the first sub-region is used to store the raw EEG signals and derived data of all channels of the first neurological abnormality event. Optionally, the first neurological abnormality event is a neurological abnormality event with a level greater than or equal to a first level threshold.

[0074] In some embodiments, the derived data includes: time-frequency coherence heatmaps, spatial gradient matrices, propagation path maps, and heatmaps of the origins of neural abnormalities.

[0075] Optionally, the origin heatmap of neural abnormalities is determined by spatial frequency statistics of the EEG channels or cortical locations with the largest spatial gradient amplitudes in historical neural abnormality events, and / or the starting locations of the propagation paths.

[0076] In some embodiments, the second sub-region is used to store the raw EEG signals and derived data of the key channels of the second neurological abnormality event. Optionally, the second neurological abnormality event is a neurological abnormality event with a level greater than or equal to a second level threshold and less than a first level threshold.

[0077] It should be understood that, in this embodiment, the key channel can be understood as the EEG channel with the largest spatial gradient in the said neurological abnormal event, or the channel with the highest energy peak in the high-frequency band, or the channel identified as the earliest activated in multi-channel propagation timing analysis.

[0078] In some embodiments, the third sub-region is used to store metadata of the third neural abnormality event. Optionally, the metadata includes: values ​​of the five-dimensional features, the timestamp of the third neural abnormality event, and its level. Optionally, the third neural abnormality event is a neural abnormality event with a level lower than a second level threshold.

[0079] For example, in Example 1, the first sub-interval can store the first neural abnormality event with S≥0.7, and save the original EEG signals and derived data of all channels; the second sub-interval can store the second neural abnormality event with 0.5≤S<0.7, and save only the original EEG signals and derived data of the key channels; the third sub-interval can store the third neural abnormality event with S<0.5, and save only the event metadata (including timestamp, five-dimensional feature value and S).

[0080] In some embodiments, the above method may further include: Based on the first and / or second locations, the stimulation targets for neural modulation are determined.

[0081] Optionally, the first position is the EEG channel or cortical location with the largest spatial gradient amplitude in the current neurological abnormality event, and the largest spatial gradient amplitude exceeds a first threshold.

[0082] Optionally, the second location includes one or more locations in the heat map of the origin of the neural abnormality where the thermal value exceeds the second threshold.

[0083] In some embodiments, the intensity of the stimulation parameters of the neural modulation is positively correlated with the first level of the neural abnormality. Optionally, the higher the first level, the higher the intensity of the stimulation current.

[0084] For example, following Example 1 above, in Example 2, if S≥0.7, the stimulation current intensity can be 1.6 mA; if 0.5≤S<0.7, the stimulation current intensity can be 1.2 mA; if 0.3≤S<0.5, the stimulation current intensity can be 0.8 mA; otherwise, no stimulation is triggered.

[0085] In some embodiments, the intensity of the neurally modulated stimulation parameters is adjusted based on the current value of the temporal gradient and / or time-frequency coherence. Optionally, the intensity of the neurally modulated stimulation parameters may be adjusted proportionally to the current value of the temporal gradient and / or time-frequency coherence.

[0086] In some embodiments, the intensity of the stimulus parameter is further adjusted proportionally to the current value of the temporal gradient and / or time-frequency coherence, based on an initial value determined according to a first level.

[0087] Optional, final stimulation current intensity I It can be represented as: I = I 0×(1+ 0.3× Gt + 0.2× Ctf ),in, I 0 is the initial value determined based on the first level. Gt For time gradient, Ctf This refers to time-frequency coherence.

[0088] It should be noted that the final stimulation current intensity does not exceed the hardware safety limit, for example, 1.6 mA. For instance, following Example 2 above, in Example 3, S = 0.65, then I 0 = 1.2mA, if Gt = 0.9, Ctf = 0.8, then I=1.2×(1+0.3×0.9+0.2×0.8)=1.716mA, and the final output stimulation current intensity should be 1.6 mA.

[0089] In some embodiments, the above method may further include: Electrical stimulation pulses are applied to a defined stimulation target based on the intensity of the stimulation parameters. Within a preset time window after the application of the electrical stimulation pulse, the second value of the five-dimensional feature is determined, and the first level is updated based on the second value of the five-dimensional feature; If the updated first level is less than the difference between the original first level and the preset threshold, the stimulus is determined to be effective; otherwise, the stimulus is determined to be ineffective, and the target or intensity of the next stimulus is adaptively adjusted based on the updated first level and the second value of the five-dimensional features.

[0090] In some embodiments, if S′ ≥ S – Δ, the stimulus is ineffective, indicating that the current target or intensity has failed to inhibit the abnormal nerve, and the target or intensity of the next stimulus needs to be adaptively adjusted. Here, S′ is the updated first level, S is the original first level, and Δ is a preset threshold, for example: Δ = 0.2.

[0091] Optionally, in this embodiment, the preset time window can be 2 to 5 seconds after stimulation.

[0092] In this embodiment, by introducing a verification scheme to determine the effectiveness of the stimulus, a complete closed loop is formed. This not only avoids the blind repetition of ineffective stimuli and improves safety, but also enables adaptive and precise optimization of stimulus parameters. For example, if the spatial gradient... Gs If the target does not decrease, it indicates that the target may have shifted and needs to be adjusted; if the time gradient... Gt If it remains high, it indicates that the current intensity is insufficient and the current needs to be increased; if the time-frequency coherence is... Ctf An increase in frequency indicates that the stimulus may induce synchronization, and the frequency needs to be reduced.

[0093] In some embodiments, the structured storage partition also includes an archive area for storing initial data on neurological abnormalities confirmed by a user (e.g., a physician), and / or key events tagged by an implanted closed-loop neurostimulation system. Optionally, the archive area employs a non-automatic overwrite strategy. It should be understood that the data stored in the archive area is data actively migrated by the user.

[0094] Optionally, the first data includes: raw EEG signals, values ​​of five-dimensional features, derived data, and neuromodulation stimulation records. Optionally, the neuromodulation stimulation records may include, but are not limited to: stimulation time, target, stimulation parameters, and post-stimulation response characteristics.

[0095] Optional, key events include: neurological events that are successfully interrupted by neural modulation, neurological events that are unresponsive to stimulation but have typical characteristics, and neurological events that are the first or new pattern of seizures.

[0096] It should be understood that, in this embodiment, a neural abnormality event that is successfully interrupted by neural modulation can be understood as the neural abnormality event corresponding to the determination that the stimulus is effective.

[0097] Optionally, a typical neurological abnormal event can be defined as a neurological abnormal event that meets the following conditions: propagation consistency greater than 0.5, duration greater than 2 seconds, time-frequency coherence exceeding the baseline standard deviation of the non-ictal period, spatial gradient normalization value greater than 0.4, and a five-dimensional weighted rank not less than 0.5.

[0098] Optionally, novel seizure pattern events can be determined based on any of the following methods: Method 1: Check if the minimum Euclidean distance between the five-dimensional feature vector of the event and historical neurological abnormal events is greater than 0.5. If so, the event is a novel seizure pattern event.

[0099] Method 2: Whether the origin of the event is located in a low-probability area of ​​the historical origin heatmap (e.g., heat value is less than a preset threshold). If so, the event is a new type of outbreak pattern event.

[0100] In the above embodiments, the implantable closed-loop neurostimulation system can be used to optimize stimulation strategies by marking key events.

[0101] In summary, the embodiments of this application provide a data storage method for efficiently processing, intelligently classifying, and optimizing the storage of multi-channel EEG signals and related data, which can achieve the following technical effects: By decomposing and processing continuously acquired multi-channel EEG signals, signal components of multiple frequency bands are obtained, and five-dimensional features are calculated based on these signal components, thus achieving a comprehensive characterization of neurological abnormal events and improving the accuracy of neurological abnormal event detection.

[0102] By determining the level of neural abnormal events based on the first value of five-dimensional features, it is possible to distinguish neural abnormal events of different severity, achieve intelligent classification, and then adopt different storage strategies to optimize the existing single storage strategy.

[0103] In addition, the cache area employs a circular overlay strategy: it caches the raw EEG signals and five-dimensional feature values ​​of neurological abnormal events within a preset time period, ensuring that important data in the short term is not lost, while avoiding unlimited growth in storage requirements. The event area uses a categorized overlay strategy: different overlay rules are applied to data of different levels (high-level data cannot be overwritten by low-level data), ensuring that data from important events is prioritized for preservation, thereby increasing the value density of the data. This hierarchical storage strategy not only saves storage resources but also improves the efficiency of data retrieval and analysis.

[0104] Based on the same inventive concept, embodiments of this application provide an implantable closed-loop neurostimulation system, such as... Figure 2 As shown, the implantable closed-loop neurostimulation system 10 includes: a data acquisition module 11, a processor 12, and a controller 13 connected in sequence.

[0105] In some embodiments, the acquisition module 11 is used to acquire multi-channel EEG signals; the processor 12 is used to decompose and process the continuously acquired multi-channel EEG signals to obtain signal components of multiple frequency bands, and determine a first value of the five-dimensional feature based on the multiple signal components, and when a neurological abnormality event is determined based on the first value of the five-dimensional feature, a first level corresponding to the neurological abnormality event is determined based on the first value of the five-dimensional feature; the controller 13 is used to store the relevant data of the neurological abnormality event to a structured storage partition based on the first level corresponding to the neurological abnormality event.

[0106] Optional, the five-dimensional features include: time-frequency coherence, spatial gradient, temporal gradient, propagation consistency, and duration.

[0107] Optionally, structured storage partitions include: cache area and event area.

[0108] In some embodiments, the cache area is used to cache the raw EEG signals and five-dimensional feature values ​​of neurological abnormal events within a preset time period. Optionally, the cache area adopts a circular coverage strategy. In some embodiments, the event area includes multiple sub-areas, which are used to store relevant data of neurological abnormal events at different levels. Optionally, the event area adopts a classification coverage strategy.

[0109] In some embodiments, the classification coverage strategy includes: for data of the same level, a first-in-first-out coverage strategy is used for storage; for data of different levels, higher levels cannot be covered by lower levels.

[0110] In some embodiments, when the total storage occupancy of the event area exceeds a preset upper limit, the data stored in the sub-area corresponding to the lower level is automatically compressed or metadata processed.

[0111] In some embodiments, the plurality of sub-regions includes: a first sub-region, a second sub-region, and a third sub-region.

[0112] In some embodiments, the first sub-region is used to store the raw EEG signals and derived data of all channels of the first neurological abnormality event. Optionally, the first neurological abnormality event is a neurological abnormality event with a level greater than or equal to a first level threshold.

[0113] In some embodiments, the derived data includes: time-frequency coherence heatmaps, spatial gradient matrices, propagation path maps, and heatmaps of the origins of neural abnormalities.

[0114] Optionally, the origin heatmap of neural abnormalities is determined by spatial frequency statistics of the EEG channels or cortical locations with the largest spatial gradient amplitudes in historical neural abnormality events, and / or the starting locations of the propagation paths.

[0115] In some embodiments, the second sub-region is used to store the raw EEG signals and derived data of the key channels of the second neurological abnormality event. Optionally, the second neurological abnormality event is a neurological abnormality event with a level greater than or equal to a second level threshold and less than a first level threshold.

[0116] It should be understood that, in this embodiment, the key channel can be understood as the EEG channel with the largest spatial gradient in the said neurological abnormal event, or the channel with the highest energy peak in the high-frequency band, or the channel identified as the earliest activated in multi-channel propagation timing analysis.

[0117] In some embodiments, the third sub-region is used to store metadata of the third neural abnormality event. Optionally, the metadata includes: values ​​of the five-dimensional features, the timestamp of the third neural abnormality event, and its level. Optionally, the third neural abnormality event is a neural abnormality event with a level lower than a second level threshold.

[0118] In some embodiments, the processor 12 is further configured to: determine the stimulation target of neural modulation based on a first location and / or a second location.

[0119] Optionally, the first position is the EEG channel or cortical location with the largest spatial gradient amplitude in the current neurological abnormality event, and the largest spatial gradient amplitude exceeds a first threshold.

[0120] Optionally, the second location includes one or more locations in the heat map of the origin of the neural abnormality where the thermal value exceeds the second threshold.

[0121] In some embodiments, the intensity of the stimulation parameters of the neural modulation is positively correlated with the first level of the neural abnormality.

[0122] In some embodiments, the intensity of the neural modulation stimulation parameters is adjusted based on the temporal gradient and / or the current value of time-frequency coherence. In some embodiments, the controller 13 is further configured to: apply an electrical stimulation pulse to a determined stimulation target based on the intensity of the stimulation parameters; the processor 12 is further configured to: determine a second value of the five-dimensional feature within a preset time window after the application of the electrical stimulation pulse, update the first level based on the second value of the five-dimensional feature, and if the updated first level is less than the difference between the first level before the update and a preset threshold, determine that the stimulation is effective; otherwise, determine that the stimulation is ineffective, and adaptively adjust the target or intensity of the next stimulation based on the updated first level and the second value of the five-dimensional feature.

[0123] In some embodiments, the structured storage partition further includes an archive area for storing initial data on user-confirmed neural abnormalities, and / or key events tagged by an implanted closed-loop neurostimulation system. Optionally, the archive area employs a non-automatic overwrite strategy.

[0124] Optionally, the first data includes: raw EEG signals, values ​​of five-dimensional features, derived data, and records of neuromodulation stimulation.

[0125] Optional, key events include: neurological events that are successfully interrupted by neural modulation, neurological events that are unresponsive to stimulation but have typical characteristics, and neurological events that are the first or new pattern of seizures.

[0126] In some embodiments, the processor 12 is further configured to perform the following operations: Determine whether the first values ​​for propagation consistency and duration are both greater than their respective thresholds; If so, then determine whether the first values ​​of time-frequency coherence, spatial gradient, and temporal gradient are all greater than their respective thresholds; If so, then a neurological event has been confirmed.

[0127] The implantable closed-loop neurostimulation system provided in this application embodiment can achieve the above-mentioned... Figure 1 The various processes implemented in the method embodiments shown will not be described again here to avoid repetition.

[0128] The implantable closed-loop neurostimulation system provided in this application decomposes and processes continuously acquired multi-channel EEG signals to obtain signal components in multiple frequency bands, and calculates five-dimensional features based on these signal components, thereby achieving a comprehensive characterization of abnormal neurological events and improving the accuracy of abnormal neurological event detection.

[0129] By determining the level of neural abnormal events based on the first value of five-dimensional features, it is possible to distinguish neural abnormal events of different severity, achieve intelligent classification, and then adopt different storage strategies to optimize the existing single storage strategy.

[0130] In addition, the cache area employs a circular overlay strategy: it caches the raw EEG signals and five-dimensional feature values ​​of neurological abnormal events within a preset time period, ensuring that important data in the short term is not lost, while avoiding unlimited growth in storage requirements. The event area uses a categorized overlay strategy: different overlay rules are applied to data of different levels (high-level data cannot be overwritten by low-level data), ensuring that data from important events is prioritized for preservation, thereby increasing the value density of the data. This hierarchical storage strategy not only saves storage resources but also improves the efficiency of data retrieval and analysis.

[0131] The implantable closed-loop neurostimulation system of this application embodiment can execute the data storage method provided in this application embodiment. 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 data storage 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 data storage methods shown above. They will not be repeated here.

[0132] 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.

[0133] In an alternative embodiment, a signal processing system, such as Figure 3 As shown, Figure 3 The signal processing system 20 shown includes a processor 21 and a memory 23. The processor 21 is communicatively connected to the memory 23, for example, via a bus 22.

[0134] Processor 21 may be a CPU (Central Processing Unit), general-purpose processor, DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), or other programmable logic device, transistor logic device, hardware component, or any combination thereof, including the chip or implantable closed-loop neurostimulation system described in any of the above embodiments. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 21 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of DSP and microprocessor, etc.

[0135] Bus 22 may include a pathway for transmitting information between the aforementioned components. Bus 22 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 22 may 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.

[0136] The memory 23 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 disc 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 not limited thereto.

[0137] The memory 23 is used to store computer programs that execute the embodiments of this application, and the execution is controlled by the processor 21. The processor 21 is used to execute the computer programs stored in the memory 23 to implement the steps shown in the foregoing method embodiments.

[0138] 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.

[0139] 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.

[0140] 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.

[0141] 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.

[0142] 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.

[0143] 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.

[0144] 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. A data storage method for an implantable closed-loop neurostimulation system, characterized in that, The method includes: The continuously acquired multi-channel EEG signals are decomposed and processed to obtain signal components in multiple frequency bands; Based on multiple signal components, a first value of a five-dimensional feature is determined, the five-dimensional feature including: time-frequency coherence, spatial gradient, temporal gradient, propagation consistency, and duration; When a neural abnormality event is determined to occur based on the first value of the five-dimensional features, the first level corresponding to the neural abnormality event is determined based on the first value of the five-dimensional features. Based on the first level corresponding to the neural abnormal event, the relevant data of the neural abnormal event is stored in a structured storage partition, which includes: a cache area and an event area; The buffer area is used to cache the raw EEG signals and five-dimensional feature values ​​of neural abnormal events within a preset time period, and the buffer area adopts a circular coverage strategy; the event area includes multiple sub-areas, which are used to store relevant data of neural abnormal events of different levels; the event area adopts a classification coverage strategy, which includes: for data of the same level, a first-in-first-out coverage strategy is used for storage; for data of different levels, higher-level data cannot be covered by lower-level data.

2. The data storage method according to claim 1, characterized in that, When the total storage occupancy rate of the event area exceeds the preset upper limit, the data stored in the sub-area corresponding to the lower level will be automatically compressed or metadata processed.

3. The data storage method according to claim 1, characterized in that, The plurality of sub-regions includes: a first sub-region, a second sub-region, and a third sub-region, wherein, The first sub-region is used to store the raw EEG signals and derived data of all channels of the first neurological abnormal event, which is a neurological abnormal event with a level greater than or equal to the first level threshold. The second sub-region is used to store the raw EEG signals and derived data of the key channels of the second neurological abnormal event. The first neurological abnormal event is a neurological abnormal event with a level greater than or equal to the second level threshold and less than the first level threshold. The third sub-region is used to store metadata of the third neural abnormal event. The metadata includes: the value of the five-dimensional feature, the timestamp and level of the third neural abnormal event, and the third neural abnormal event is a neural abnormal event with a level lower than the second level threshold.

4. The data storage method according to claim 3, characterized in that, The derived data includes: time-frequency coherence heatmaps, spatial gradient matrices, propagation path maps, and heatmaps of the origins of neural abnormalities; The aforementioned heatmap of the origin of neural abnormalities is determined by spatial frequency statistics of the EEG channels or cortical locations with the largest spatial gradient amplitudes in historical neural abnormality events, and / or the starting locations of the propagation paths.

5. The data storage method according to claim 4, characterized in that, The method further includes: Based on the first and / or second location, determine the stimulation target for neural modulation; The first position is the EEG channel or cortical location with the largest spatial gradient amplitude in the current neurological abnormality event, and the largest spatial gradient amplitude exceeds a first threshold. The second position includes one or more locations in the heat map of the origin of the neurological abnormality where the thermal value exceeds a second threshold.

6. The data storage method according to claim 5, characterized in that, The intensity of the stimulation parameters of the neural modulation is positively correlated with the first level of the neural abnormality event.

7. The data storage method according to claim 6, characterized in that, The intensity of the stimulation parameters of the neural modulation is adjusted based on the time gradient and / or the current value of the time-frequency coherence.

8. The data storage method according to claim 6 or 7, characterized in that, The method further includes: An electrical stimulation pulse is applied to a determined stimulation target based on the intensity of the stimulation parameters. Within a preset time window after the application of the electrical stimulation pulse, a second value of the five-dimensional feature is determined, and the first level is updated based on the second value of the five-dimensional feature; If the updated first level is less than the difference between the original first level and the preset threshold, the stimulus is determined to be effective; otherwise, the stimulus is determined to be ineffective, and the target or intensity of the next stimulus is adaptively adjusted based on the updated first level and the second value of the five-dimensional feature.

9. The data storage method according to any one of claims 1-7, characterized in that, The structured storage partition also includes an archive area for storing first data of neural abnormal events confirmed by the user, and / or key events marked by the implanted closed-loop neural stimulation system. The archive area adopts a non-automatic overwrite strategy. The first data includes: raw EEG signals, values ​​of five-dimensional features, derived data, and records of neuromodulation stimulation; The key events include: neurological abnormalities that are successfully interrupted by neural modulation, neurological abnormalities that are ineffective in response to stimulation but have typical characteristics, and neurological abnormalities that are the first or a new pattern of seizures.

10. The data storage method according to any one of claims 1-7, characterized in that, The method further includes: Determine whether the first values ​​of the propagation consistency and the duration are both greater than their respective thresholds; If so, determine whether the first values ​​of the time-frequency coherence, the spatial gradient, and the temporal gradient are all greater than their respective thresholds; If so, then a neurological event has been confirmed.

11. An implantable closed-loop neurostimulation system, characterized in that, include: The signal acquisition module is used to acquire multi-channel EEG signals; The processor is used to decompose and process continuously acquired multi-channel EEG signals to obtain signal components in multiple frequency bands, and determine a first value of a five-dimensional feature based on the multiple signal components. When a neurological abnormality event is determined based on the first value of the five-dimensional feature, a first level corresponding to the neurological abnormality event is determined based on the first value of the five-dimensional feature. The five-dimensional feature includes: time-frequency coherence, spatial gradient, temporal gradient, propagation consistency, and duration. The controller is used to store the relevant data of the neural abnormal event to a structured storage partition based on the first level corresponding to the neural abnormal event. The structured storage partition includes a cache area and an event area. The buffer area is used to cache the raw EEG signals and five-dimensional feature values ​​of neural abnormal events within a preset time period, and the buffer area adopts a circular coverage strategy; the event area includes multiple sub-areas, which are used to store relevant data of neural abnormal events of different levels; the event area adopts a classification coverage strategy, which includes: for data of the same level, a first-in-first-out coverage strategy is used for storage; for data of different levels, higher-level data cannot be covered by lower-level data.

12. The implantable closed-loop neurostimulation system according to claim 11, characterized in that, When the total storage occupancy rate of the event area exceeds the preset upper limit, the data stored in the sub-area corresponding to the lower level will be automatically compressed or metadata processed.

13. The implantable closed-loop neurostimulation system according to claim 11, characterized in that, The plurality of sub-regions includes: a first sub-region, a second sub-region, and a third sub-region, wherein, The first sub-region is used to store the raw EEG signals and derived data of all channels of the first neurological abnormal event, which is a neurological abnormal event with a level greater than or equal to the first level threshold. The second sub-region is used to store the raw EEG signals and derived data of the key channels of the second neurological abnormal event, which is a neurological abnormal event with a level greater than or equal to the second level threshold and less than the first level threshold. The third sub-region is used to store metadata of the third neural abnormal event. The metadata includes: the value of the five-dimensional feature, the timestamp and level of the third neural abnormal event, and the third neural abnormal event is a neural abnormal event with a level lower than the second level threshold.

14. The implantable closed-loop neurostimulation system according to claim 13, characterized in that, The derived data includes: time-frequency coherence heatmaps, spatial gradient matrices, propagation path maps, and heatmaps of the origins of neural abnormalities; The aforementioned heatmap of the origin of neural abnormalities is determined by spatial frequency statistics of the EEG channels or cortical locations with the largest spatial gradient amplitudes in historical neural abnormality events, and / or the starting locations of the propagation paths.

15. The implantable closed-loop neurostimulation system according to claim 14, characterized in that, The processor is further configured to: determine the stimulation target of neural modulation based on a first location and / or a second location; wherein the first location is the EEG channel or cortical location with the largest spatial gradient amplitude in the current neural abnormality event, and the largest spatial gradient amplitude exceeds a first threshold, and the second location includes one or more locations in the thermal map of the origin of the neural abnormality that have thermal values ​​exceeding a second threshold.

16. The implantable closed-loop neurostimulation system according to claim 15, characterized in that, The intensity of the stimulation parameters of the neural modulation is positively correlated with the first level of the neural abnormality event.

17. The implantable closed-loop neurostimulation system according to claim 16, characterized in that, The intensity of the stimulation parameters of the neural modulation is adjusted based on the time gradient and / or the current value of the time-frequency coherence.

18. The implantable closed-loop neurostimulation system according to claim 16 or 17, characterized in that, The controller is also configured to: apply an electrical stimulation pulse to a determined stimulation target based on the intensity of the stimulation parameters; The processor is further configured to: within a preset time window after the application of an electrical stimulation pulse, determine a second value of the five-dimensional feature, update the first level based on the second value of the five-dimensional feature, and if the updated first level is less than the difference between the first level before the update and a preset threshold, determine that the stimulation is effective; otherwise, determine that the stimulation is ineffective, and adaptively adjust the target point or intensity of the next stimulation based on the updated first level and the second value of the five-dimensional feature.

19. The implantable closed-loop neurostimulation system according to any one of claims 11-17, characterized in that, The structured storage partition also includes an archive area for storing first data of neural abnormal events confirmed by the user, and / or key events marked by the implanted closed-loop neural stimulation system. The archive area adopts a non-automatic overwrite strategy. The first data includes: raw EEG signals, values ​​of five-dimensional features, derived data, and records of neuromodulation stimulation; The key events include: neurological abnormalities that are successfully interrupted by neural modulation, neurological abnormalities that are ineffective in response to stimulation but have typical characteristics, and neurological abnormalities that are the first or a new pattern of seizures.

20. The implantable closed-loop neurostimulation system according to any one of claims 11-17, characterized in that, The processor is also used to perform the following operations: Determine whether the first values ​​of the propagation consistency and the duration are both greater than their respective thresholds; If so, determine whether the first values ​​of the time-frequency coherence, the spatial gradient, and the temporal gradient are all greater than their respective thresholds; If so, then a neurological event has been confirmed.

21. A signal processing system, characterized in that, Including the implantable closed-loop neurostimulation system as described in any one of claims 11 to 17.

22. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the data storage method according to any one of claims 1 to 10.

23. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the data storage method according to any one of claims 1 to 10.

Images