A method and device for monitoring and early warning of neuroimmunological diseases

Abstract

The application discloses a kind of neuroimmunological disease monitoring early warning method and device, it is related to biomedical signal processing and artificial intelligence cross technical field, the method includes: from multichannel electroencephalogram, extract neuroinflammation related oscillation feature, immune-related functional connection feature and neuroimmunological state evolution feature, and construct neuroimmunological function state vector;The difference of the neuroimmunological function state vector of current period and individual baseline state vector is calculated as state offset;According to the neuroimmunological function state vector of historical period, the predicted value of the neuroimmunological function state vector of current period is predicted, and state prediction error is calculated;According to state offset and state prediction error, evolution instability is calculated;According to state offset and evolution instability, calculate comprehensive risk index.The application can realize the continuous, dynamic, prospective monitoring and early warning of neuroimmunological disease.

Description

Technical Field

[0001] This application relates to the interdisciplinary fields of biomedical signal processing and artificial intelligence, and in particular to a method and device for monitoring and early warning of neuroimmune diseases. Background Technology

[0002] Neuroimmune diseases, such as multiple sclerosis, neuromyelitis optica spectrum disorders, and autoimmune encephalitis, are a group of diseases caused by abnormal attacks on the nervous system by the immune system. Their course is complex, often with insidious onset and recurrent relapses. Accurate monitoring of disease activity and timely warning of relapse risk are crucial for improving patient prognosis and achieving individualized treatment.

[0003] Currently, clinical monitoring and evaluation of neuroimmunological diseases mainly rely on the following traditional methods: Medical imaging examinations, such as magnetic resonance imaging (MRI), can clearly show structural lesions of the central nervous system (such as demyelinating plaques), but the examinations are expensive and cannot be performed frequently, usually only once every few months or even years. Therefore, it is difficult to capture early, transient, or dynamic changes in disease activity.

[0004] Invasive fluid examinations, such as lumbar puncture to obtain cerebrospinal fluid for biochemical and cytological analysis, can provide direct evidence of immune inflammation. However, as an invasive procedure, it has low patient acceptance and carries certain risks, making it unsuitable as a routine long-term monitoring method.

[0005] Clinical symptoms and scale assessment. This method relies on patient complaints and physician examination, is highly subjective, and usually only identifies neurological deficits when they have become obvious, lacking objective and sensitive early warning capabilities.

[0006] In summary, existing clinical monitoring technologies suffer from several significant drawbacks, including low detection frequency, difficulty in achieving continuous dynamic monitoring, reliance on invasive or high-cost examinations, an emphasis on structural abnormalities rather than quantifying functional dynamic changes in neuroimmune interactions, and an inability to provide prospective early warnings before the onset of clinical symptoms. Summary of the Invention

[0007] The purpose of this application is to provide a method and device for monitoring and early warning of neuroimmune diseases, which can realize continuous, dynamic, and prospective monitoring and early warning of neuroimmune diseases.

[0008] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a method for monitoring and early warning of neuroimmune diseases, including: Acquire multi-channel EEG signals of the target object; the multi-channel EEG signals are acquired by a brain-computer interface device. From the multi-channel EEG signals, neuroinflammation-related oscillation features, immune-related functional connectivity features, and neuroimmune state evolution features are extracted, and a neuroimmune functional state vector is constructed. The difference between the current neuroimmune function state vector and the individual baseline state vector is calculated as the state offset; wherein, the individual baseline state vector is the mean of several neuroimmune function state vectors within the initial monitoring period; the initial monitoring period includes several consecutive stable monitoring cycles; Calculate the state prediction error between the predicted value and the neuroimmune function state vector of the current time period; the predicted value is the predicted value of the neuroimmune function state vector of the current time period after inputting the neuroimmune function state vector of the historical time period into the pre-trained time series prediction model. The evolutionary instability is calculated based on the state offset and the state prediction error; A comprehensive risk index is calculated based on the state offset and the evolution instability; the comprehensive risk index is used to determine early warning information.

[0009] Optionally, the formula for calculating the comprehensive risk index is: in, , These are the weighting coefficients. , For the normalized threshold, for The overall risk index for the period. for The magnitude of the state offset over a time period. for The instability of evolution over a period of time.

[0010] Optionally, neuroinflammation-related oscillation features, immune-related functional connectivity features, and neuroimmune state evolution features are extracted from the multi-channel EEG signals, and a neuroimmune functional state vector is constructed, specifically including: Neuroinflammatory-related oscillation features were extracted from the multi-channel EEG signals. Immune-related functional connectivity features were extracted from the multi-channel EEG signals. The evolutionary features of the neuroimmune state were extracted from the multi-channel electroencephalogram (EEG) signals. Based on the neuroinflammation-related oscillation characteristics, the immune-related functional connectivity characteristics, and the neuroimmune state evolution characteristics, a neuroimmune feature vector is constructed. A nonlinear dimensionality reduction method is used to map the neuroimmune feature vectors to a low-dimensional continuous state space to obtain the neuroimmune functional state vectors.

[0011] Optionally, the formula for calculating the evolutionary instability is: in, for Evolutionary instability over time; for The magnitude of the state offset during a time period; and These are the weighting coefficients; The standard deviation of the state prediction error sequence; for State prediction error for a given time period; for State prediction error for a given time period.

[0012] Optionally, The neuroinflammatory-related oscillation characteristics include: low-frequency energy ratio, Rhythm stability index and frequency drift coefficient; The immune-related functional connectivity features include: interbrain coherence, phase synchronization indices, and functional network topology parameters; The evolutionary characteristics of the neuroimmune state include: characteristic time drift rate, volatility instability index, and mutation detection index.

[0013] Optionally, The low-frequency energy ratio The calculation formula is: in, , , , These represent the total energy of EEG signals in the 0.5Hz-4Hz, 4Hz-8Hz, 8Hz-13Hz, and 13Hz-30Hz frequency bands, respectively. The Rhythm stability index The calculation formula is: ,in, The power spectral density in the 8Hz-13Hz frequency band. The coefficient of variation; The frequency drift coefficient The standard deviation of the main peak frequency in the 8Hz-13Hz band within a preset time window is calculated using the following formula: ,in, for time The main peak frequency of the frequency band. Standard deviation, For analysis of time windows; The coherence between brain regions The mean frequency domain coherence of the left frontal-left parietal region connection and the right frontal-right parietal region connection within the 4Hz-45Hz frequency band is calculated using the following formula: ,in, The frequency domain coherence function is calculated in the 4Hz-45Hz frequency band; F3-P3 is the connection between the left forehead and left top region; F4-P4 is the connection between the right forehead and right top region. The phase synchronization index The calculation formula is: ,in, Total number of brain regions; for Time-phase brain regions and brain regions The instantaneous phase difference between them; The functional network topology parameters The calculation formula is: in, The clustering coefficients of the actual brain network, The clustering coefficients are those of a random network with the same number of nodes as the actual brain network. The clustering coefficient is the number of nodes of a regular network that has the same number of nodes as the actual brain network. The actual brain network is a topology that reflects the real-time integration state of brain function, constructed with EEG acquisition channels as nodes and the synchronization of EEG signals between channels as connecting edges. The characteristic time drift rate The calculation formula is: ,in, The slope of the linear regression; The size of the time window; For a feature in continuous The average sequence over a period of time Indicates the first The average value of the characteristic over a period of time. Indicates the first The average value of a feature described in each time period. Indicates the first The average value of a feature described in each time period; The volatility instability index The calculation formula is: in, Let be the standard deviation. Let be the mean; The mutation detection index The calculation formula is: ,in, To detect mutation points using the cumulative sum algorithm, For the detection window, For a feature in continuous The average sequence over a period of time Indicates the first The average value of the characteristic over a period of time. Indicates the first The average value of the characteristic over a given time period.

[0014] Optionally, it also includes: Early warning information is generated based on the comprehensive risk index and early warning threshold.

[0015] Secondly, this application provides a neuroimmune disease monitoring and early warning device, comprising: Brain-computer interface devices are used to collect multi-channel electroencephalogram (EEG) signals from target subjects. A cloud server for executing any of the neuroimmune disease monitoring and early warning methods described above; The first terminal is connected to the cloud server to display first user data and record symptom logs input by the first user; the first user data includes multi-channel EEG signals and warning information.

[0016] Optionally, it also includes: The second terminal is used to allow a second user to remotely view the data of the first user and adjust the warning threshold.

[0017] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to implement the neuroimmune disease monitoring and early warning method described in any of the above claims.

[0018] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a method and device for monitoring and early warning of neuroimmune diseases. The method includes: acquiring multi-channel electroencephalogram (EEG) signals of a target object; the multi-channel EEG signals are acquired by a brain-computer interface device; extracting neuroinflammation-related oscillation features, immune-related functional connectivity features, and neuroimmune state evolution features from the multi-channel EEG signals, and constructing a neuroimmune functional state vector; calculating the difference between the neuroimmune functional state vector of the current time period and the individual baseline state vector as a state offset; wherein the individual baseline state vector is the mean of several neuroimmune functional state vectors within an initial monitoring period, and the initial monitoring period includes several consecutive stable monitoring cycles; calculating the state prediction error between the predicted value and the neuroimmune functional state vector of the current time period; the predicted value is the predicted value of the neuroimmune functional state vector of the current time period output after inputting the neuroimmune functional state vectors of historical time periods into a pre-trained time-series prediction model; calculating the evolutionary instability based on the state offset and the state prediction error; calculating a comprehensive risk index based on the state offset and the evolutionary instability; the comprehensive risk index is used to determine early warning information. This application utilizes a brain-computer interface device to conveniently, frequently, and even continuously acquire multi-channel EEG signals in various environments, including outside of hospitals. This completely overcomes the limitations of traditional MRI and cerebrospinal fluid examinations in terms of temporal resolution and accessibility, providing a dynamic data stream for disease management and achieving truly continuous, dynamic, and non-invasive monitoring. Furthermore, by extracting three types of quantitative features—neuroinflammatory oscillation characteristics, immune-related functional connectivity characteristics, and neuroimmune state evolution characteristics—this application directly captures the impact of immune activity on neural networks from multi-channel EEG signals, achieving a quantitative characterization of the neuroimmune functional state and overcoming the shortcomings of existing technologies that primarily focus on structural changes. This application uses a time-series prediction model to predict state evolution trends and calculates prediction errors and their volatility (standard deviation), enabling the identification of abnormal patterns that deviate from the normal evolutionary trajectory. These abnormalities often appear earlier than obvious clinical symptoms, thus achieving a truly forward-looking early warning of disease activity risks and gaining a valuable time window for early clinical intervention. Furthermore, by establishing an individualized baseline state vector and calculating the state offset of the current state relative to the baseline, the interference of basic physiological differences between different individuals is effectively overcome, improving accuracy. Therefore, this application can achieve continuous, dynamic, and forward-looking monitoring and early warning of neuroimmunological diseases. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is an application environment diagram of a neuroimmune disease monitoring and early warning method according to an embodiment of this application; Figure 2 A flowchart illustrating a method for monitoring and early warning of neuroimmune diseases provided in an embodiment of this application; Figure 3 A schematic diagram of the structure of a neuroimmune disease monitoring and early warning device provided in another embodiment of this application; Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] This application belongs to the interdisciplinary field of brain-computer interface technology, neural engineering, biomedical signal processing and artificial intelligence, and is applicable to long-term continuous monitoring and early warning of neuroimmune-related diseases such as multiple sclerosis, neuromyelitis optica spectrum disorders and autoimmune encephalitis.

[0023] Neuroimmune diseases are a group of diseases caused by abnormalities in the immune system, leading to dysfunction of the central or peripheral nervous system. Their course is typically characterized by insidious onset, fluctuating progression, and difficulty in timely identification of active phases. Current clinical diagnostic methods mainly rely on magnetic resonance imaging (MRI), cerebrospinal fluid examination, serological markers, or clinical symptom assessment, which have the following limitations: 1. Low detection frequency, making continuous monitoring difficult: Clinical examinations are usually conducted only once every few months or even years, making it impossible to capture early signs and dynamic changes in disease activity.

[0024] 2. Reliance on invasive or high-cost tests: For example, cerebrospinal fluid puncture is an invasive procedure, and MRI examinations are expensive and difficult to use as a routine monitoring method.

[0025] 3. Difficulty in reflecting dynamic changes in immune-related neural functions: Existing technologies mainly focus on structural changes and are not sensitive enough to functional changes, thus failing to reflect the dynamic process of neuroimmune interaction in real time.

[0026] 4. Inability to provide early warning of disease activity risk: Existing methods can only detect abnormalities after symptoms appear or worsen, lacking prospective early warning capabilities.

[0027] On the other hand, electroencephalogram (EEG) signals, as a direct electrophysiological reflection of the functional state of the nervous system, have advantages such as being non-invasive, having high temporal resolution (millisecond-level), and being suitable for long-term home monitoring. In recent years, brain-computer interface (BCI) technology has made significant progress in areas such as epilepsy monitoring, sleep analysis, and motor function rehabilitation. However, existing BCI technologies are mainly designed for specific neurological dysfunctions and lack dedicated modeling methods for the pathological mechanisms of neuroimmunological diseases, thus failing to effectively capture the influence of immune activity on neuroelectrophysiological characteristics.

[0028] Therefore, there is an urgent need for a brain-computer interface method and system that can reflect the evolution of neuroimmune function and provide early warning of disease activity risks, filling the technological gap in continuous monitoring and early warning of neuroimmune diseases.

[0029] This application constructs a neuroimmune functional state model to achieve continuous monitoring, state quantification, and multi-level risk warning of neuroimmune diseases, thereby solving the technical problems existing in the prior art, such as low detection frequency, difficulty in continuous monitoring, and lack of forward-looking early warning capabilities.

[0030] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0031] The neuroimmune disease monitoring and early warning method provided in this application embodiment can be applied to, for example... Figure 1In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on another server. Terminal 102 can send multi-channel EEG signals of the target object to be processed to server 104. After receiving the multi-channel EEG signals of the target object, server 104 extracts neuroinflammation-related oscillation features, immune-related functional connectivity features, and neuroimmune state evolution features from the multi-channel EEG signals, and constructs a neuroimmune function state vector; calculates the difference between the neuroimmune function state vector of the current time period and the individual baseline state vector as a state offset; wherein, the individual baseline state vector is the mean of several neuroimmune function state vectors within the initial monitoring period; the initial monitoring period includes several consecutive stable monitoring cycles; calculates the state prediction error between the predicted value and the neuroimmune function state vector of the current time period; the predicted value is the predicted value of the neuroimmune function state vector of the current time period after inputting the neuroimmune function state vectors of historical time periods into a pre-trained time series prediction model; calculates the evolutionary instability based on the state offset and the state prediction error; calculates the comprehensive risk index based on the state offset and the evolutionary instability; the comprehensive risk index is used to determine early warning information. Server 104 can feed back the obtained comprehensive risk index to terminal 102. In addition, in some embodiments, the neuroimmune disease monitoring and early warning method can also be implemented by server 104 or terminal 102 alone. For example, terminal 102 can directly process the multi-channel EEG signal of the target object to be processed, or server 104 can obtain the multi-channel EEG signal of the target object to be processed from the data storage system and process it accordingly.

[0032] The terminal 102 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. The server 104 can be implemented using a standalone server or a server cluster composed of multiple servers, or it can be a cloud server.

[0033] In one exemplary embodiment, such as Figure 2 As shown, a method for monitoring and early warning of neuroimmunological diseases is provided. This method is executed by a computer device, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1Taking server 104 as an example, the explanation includes the following steps S1 to S6. Wherein: S1. Acquire multi-channel EEG signals of the target object; the multi-channel EEG signals are acquired by a brain-computer interface device.

[0034] S2. Extract neuroinflammation-related oscillation features, immune-related functional connectivity features, and neuroimmune state evolution features from the multi-channel EEG signals, and construct a neuroimmune functional state vector.

[0035] S3. Calculate the difference between the current neuroimmune function state vector and the individual baseline state vector as the state offset; wherein, the individual baseline state vector is the average of several neuroimmune function state vectors within the initial monitoring period, and the initial monitoring period includes several consecutive stable monitoring cycles.

[0036] S4. Calculate the state prediction error between the predicted value and the neuroimmune function state vector of the current time period; the predicted value is the predicted value of the neuroimmune function state vector of the current time period after inputting the neuroimmune function state vector of the historical time period into the pre-trained time series prediction model.

[0037] S5. Calculate the evolution instability based on the state offset and the state prediction error.

[0038] S6. Calculate the comprehensive risk index based on the state offset and the evolution instability; the comprehensive risk index is used to determine early warning information.

[0039] In another exemplary embodiment of this application, when acquiring multi-channel EEG signals in step S1 above, the focus is on covering the frontal lobe, parietal lobe, and central region, collecting EEG signals in the resting state and during sleep.

[0040] In another exemplary embodiment of this application, step S2 is replaced by steps S21 to S25: S21. Extract neuroinflammatory-related oscillation features from the multi-channel EEG signals.

[0041] S22. Extract immune-related functional connectivity features from the multi-channel EEG signals.

[0042] S23. Extract the evolution characteristics of the neuroimmune state from the multi-channel EEG signals.

[0043] S24. Construct a neuroimmune feature vector based on the neuroinflammation-related oscillation features, the immune-related functional connectivity features, and the neuroimmune state evolution features.

[0044] S25. Using a nonlinear dimensionality reduction method, the neuroimmune feature vector is mapped to a low-dimensional continuous state space to obtain the neuroimmune function state vector.

[0045] In another exemplary embodiment of this application, the neuroinflammation-related oscillation features, immune-related functional connectivity features, and neuroimmune state evolution features extracted in steps S2 and S21 to S23 above are neuroimmune-specific features proposed in this embodiment.

[0046] The neuroinflammation-related oscillation characteristics include: low-frequency energy ratio, Rhythm stability indices and frequency drift coefficients; the immune-related functional connectivity features include: brain inter-brain coherence, phase synchronization indices, and functional network topology parameters; the neuroimmune state evolution features include: characteristic time drift rate, fluctuation instability index, and mutation detection indices.

[0047] Representing multichannel EEG signals as a time-series matrix: in: Number of brainwave channels ( ), This represents the total number of sampling points within the time window. Indicates the first The first channel in the Voltage values ​​at each sampling point (unit: ), .

[0048] The extraction process of the neuroinflammatory-related oscillation features includes: Low-frequency energy ratio: in, It represents the total energy of EEG signals in the 0.5Hz-4Hz frequency band, reflecting deep sleep and pathological slow-wave activity; The total energy in the 4Hz-8Hz frequency band reflects a mild neuroinflammatory state. The total energy in the 8Hz-13Hz frequency band reflects normal neural synchronous activity; The total energy in the 13Hz-30Hz frequency band reflects cortical excitability.

[0049] Physiological significance of low-frequency energy ratio: Neuroinflammation leads to an increase in low-frequency energy, and an increase in this ratio indicates abnormal neuroimmune function.

[0050] Rhythm stability index: ,in, The power spectral density in the 8Hz-13Hz frequency band. This is the coefficient of variation (standard deviation / mean).

[0051] Physiological significance of rhythm stability indices: Neuroinflammation reduces them. Rhythm stability, The smaller the value, the more unstable the neuroimmune function.

[0052] Frequency drift coefficient ;in, for time The main peak frequency of the frequency band (8Hz-13Hz) Standard deviation, The analysis time window is typically 5 minutes.

[0053] Physiological significance of frequency drift coefficient: When immune activity is abnormal, The peak frequency will exhibit irregular drift. Increased levels reflect a disorder in neuroimmune regulation.

[0054] Brain region coherence ;in: The frequency domain coherence function is calculated in the 4Hz-45Hz frequency band; F3-P3 is the left forehead-left top region connection; F4-P4 is the right forehead-right top region connection.

[0055] Physiological significance of interbrain region coherence: Neuroinflammation can disrupt frontoparietal functional connectivity. Reduces the impairment of information transmission between brain regions.

[0056] Phase synchronization index ,in: Total number of brain regions; for Time-phase brain regions and brain regions The instantaneous phase difference between them; the instantaneous phase is obtained through the Hilbert transform.

[0057] Physiological significance of phase synchronization indicators: Immune abnormalities can affect neural network synchronization. A decreased value indicates impaired neuroimmune coordination function.

[0058] Functional network topology parameters ,in The clustering coefficients of the actual brain network, The clustering coefficients are those of a random network with the same number of nodes as the actual brain network. The clustering coefficient is the number of nodes of a regular network that has the same number of nodes as the actual brain network. The actual brain network is a topology that reflects the real-time integration state of brain function, constructed with EEG acquisition channels as nodes and the synchronization (coherence, phase synchronization) of EEG signals between channels as connecting edges.

[0059] Physiological significance of functional network topology parameters: They reflect the small-world properties of brain networks; neuroinflammation can disrupt network structure. Deviations from the normal range suggest abnormal neuroimmune function.

[0060] Feature time drift rate ,in, The slope of the linear regression; The size of the time window (usually) (corresponding to 7 time periods) For a feature in continuous A series of average values ​​over a period of time. Indicates the first The average value of the characteristic over a period of time. Indicates the first The average value of a feature described over a period of time. Indicates the first The average value of a feature over a given time period; where "a feature" refers to a selected key feature, such as... .

[0061] The physiological significance of feature time drift rate: It reflects the trend of feature change over time. A significant increase indicates a rapid deterioration in the neuroimmune status.

[0062] Volatility Instability Index: in, Standard deviation, The mean; For time window (usually) (Divided into 7 time periods).

[0063] Physiological significance of the volatility index: It measures the degree of volatility of characteristic values. An increase in f8 reflects instability in neuroimmune regulation.

[0064] Mutation detection indicators ,in, To detect mutation points using the cumulative sum algorithm, For the detection window (usually) (corresponding to 14 time periods) For a feature in continuous A series of average values ​​over a period of time. Indicates the first The average value of the characteristic over a period of time. Indicates the first The average value of the characteristic over a given time period.

[0065] Physiological significance of mutation detection indicators: detecting mutation behavior of characteristic values. Exceeding the threshold indicates a significant change in the neuroimmune state.

[0066] (Note: In one specific embodiment of this application, the "time period" can be a time cycle in days, and the corresponding average value is the daily average value.) Finally, a 9-dimensional neuroimmune feature vector was constructed. ; in: The low-frequency energy ratio reflects the degree of nerve inflammation.

[0067] : Rhythm stability indicators reflect the stability of neural synchronization function.

[0068] Frequency drift coefficient, reflecting the stability of nerve oscillation frequency.

[0069] Brain region coherence reflects the strength of frontoparietal functional connectivity.

[0070] Phase synchronization index, reflecting the degree of synchronization of the whole brain's neural pathways.

[0071] Functional network topology parameters reflect the organizational structure of brain networks.

[0072] The feature time drift rate reflects the long-term trend of feature changes.

[0073] The volatility index reflects the degree of short-term volatility.

[0074] Mutation detection indicators reflect the risk of state mutations.

[0075] Furthermore, step S25 specifically includes: First, the constructed neuroimmune features for each dimension are individually standardized: in: The individual baseline period (also known as the initial monitoring period, in this example, the first 30 days of the monitoring period are selected) is the first 30 days of the monitoring period. The mean of each feature; For the individual baseline period The standard deviation of each feature; Output: Standardized neuroimmune feature vector .in, , and The first feature after standardization The second feature after standardization And the ninth feature after standardization .

[0076] Secondly, a nonlinear dimensionality reduction method is used to map the standardized neuroimmune feature vectors to a low-dimensional continuous state space to obtain the neuroimmune function state vectors.

[0077] This process specifically involves mapping the standardized neuroimmune feature vector to a neuroimmune functional state vector using a nonlinear mapping function, as shown in the formula: in: for Dimensional neuroimmune function state vector ( ); It is a nonlinear mapping function; For nonlinear activation functions (such as...) ); for Weight matrix, Here, the dimension of the input neural immune feature vector is... ; This is the bias vector.

[0078] In practice, the mapping model is trained using data from healthy control groups and patients in the stable phase, and nonlinear dimensionality reduction is performed using t-SNE or an autoencoder. After obtaining the low-dimensional vector, "healthy state region" and "alert region" are defined in the low-dimensional space. Individualized neuroimmune function state vectors are then output. The healthy zone and the alert zone can be defined by setting thresholds as needed.

[0079] In another exemplary embodiment of this application, step S4 described above can be replaced by the following steps: Based on the neuroimmune function state vectors of several historical periods, the predicted values ​​of the neuroimmune function state vectors of the current period are predicted by using a pre-trained time series prediction model, resulting in several predicted values. Calculate the state prediction error between several predicted values ​​and the neuroimmune function state vector of the current time period to obtain a state prediction error sequence, and calculate the standard deviation of the state prediction error sequence.

[0080] In another exemplary embodiment of this application, the formula for calculating evolutionary instability in step S5 above is: ; in, for Evolutionary instability over time; for The magnitude of the state offset during a time period; and These are the weighting coefficients; The standard deviation of the state prediction error sequence; for State prediction error for a given time period; for State prediction error for a given time period.

[0081] In another exemplary embodiment of this application, steps S4-S5 mainly involve time-series modeling of the neuroimmune function state vector, taking into account both current anomaly capture and future trend prediction functions.

[0082] Real-time status verification (used to calculate evolutionary deviations). Utilizing Based on historical window data from previous moments, we can extrapolate the theoretical health status that should be present at the current moment. and compared with the measured value Comparison: ; ; in: This is the theoretical state at the current moment (i.e., the predicted value of the neuroimmune function state vector for the current period) based on historical patterns. This represents a vector of neuroimmune function states over several historical time periods. The instantaneous state evolution deviation (i.e., state prediction error) reflects the degree to which the current physiological state deviates from the normal evolutionary trajectory.

[0083] The specific execution process can be selected from the following procedures: Individualized baseline construction: Calculate individual baseline state vectors during the initial monitoring phase (30 days): ; Temporal evolution modeling: Strategy 1 is used to calculate the theoretical state at the current moment (i.e., the predicted value of the neuroimmune function state vector for the current time period). and state prediction error .

[0084] Calculate the offset of the current neuroimmune function state vector relative to the individual's baseline state vector: .

[0085] Anomaly detection: Define the instability of the neuroimmune state evolution: U(t)=w1·||ΔS(t)||+w2·std{e(tk),...,e(t)}; in: The magnitude of the state offset. This represents the standard deviation of the recent prediction error sequence (reflecting the stability of the evolution process). , Weighting coefficients (e.g.) , ).

[0086] In another exemplary embodiment of this application, a dual temporal evolution strategy is employed, used to calculate real-time evolution instability and provide forward-looking trend references, respectively. The method for calculating real-time evolution instability has been described above; the following describes a process for providing forward-looking trend references: Forward-looking evolutionary predictions (for trend visualization). Based on current... Data at each moment, predicting the next moment. Status trend: in: As of now Data at any given time This represents a predicted state for future points in time, used to assess the direction of disease progression (trend warning), and is not included in subsequent risk index calculations. Note: For parameters LSTM network. This represents the size of the historical time window.

[0087] Calculate the trend forecast value for the next time step. .

[0088] In another exemplary embodiment of this application, step S6 described above can be replaced by the following steps: The warning level is determined by jointly considering state offset and evolutionary instability, warning information is generated, and a comprehensive risk index is defined. in, , Weighting coefficients ( ), , For the normalization threshold, for The overall risk index for the period. for The magnitude of the state offset during a time period. for The instability of evolution over a period of time.

[0089] according to Output different levels of warning information based on size: Level 0 (Normal): No warning; Level 1 (Follow): It is recommended to strengthen monitoring; Level 2 (Warning): Clinical consultation is recommended; Level 3 (High Risk): It is recommended to seek medical attention immediately.

[0090] The execution process can be selected from the following steps: Real-time calculation of comprehensive risk index ; The warning level is determined based on the warning threshold. A structured early warning report is generated, which includes risk level, key abnormal features, and recommended measures. The recommended measures correspond to the determined early warning level according to a pre-set mapping rule, specifically: Level 1 corresponds to enhanced monitoring, Level 2 corresponds to clinical consultation, and Level 3 corresponds to immediate medical attention. The key abnormal features are the feature dimensions with the largest standardized deviation from the individual baseline state vector when calculating the state offset.

[0091] It should be noted that the warning information output by the method and device provided in this application is only used as objective data processing and physiological state reference for the user's neuroimmune function status, and is not intended to directly obtain clinical medical diagnosis results. The final disease diagnosis still needs to be determined by a professional doctor in combination with clinical manifestations and other examination results.

[0092] Compared with the prior art, this embodiment has at least the following beneficial effects: 1. Enable continuous, non-invasive monitoring of neuroimmunological diseases: Through brain-computer interface devices, daily or even continuous monitoring can be achieved in a home environment, breaking through the time limitations of traditional clinical examinations and providing high-frequency data support for disease management.

[0093] 2. Ability to identify disease activity risk before symptoms appear: Dynamic evolution analysis based on neuroimmune function state vectors can identify disease activity risk in advance, gaining a valuable time window for clinical intervention.

[0094] 3. Possesses individualized adaptive capabilities: By constructing individual baseline state vectors and adaptive warning thresholds, it overcomes physiological differences between individuals and improves the accuracy and specificity of monitoring.

[0095] 4. Applicable to unified monitoring of various neuroimmune diseases: The core features and modeling methods of this embodiment target the common mechanisms of neuroimmune interaction and can be applied to various neuroimmune diseases such as multiple sclerosis, neuromyelitis optica spectrum disorders, and autoimmune encephalitis.

[0096] 5. Provide multi-level early warning and clinical decision support: By quantifying risk levels and providing recommended measures, it provides clear clinical decision-making basis for patients and doctors, reducing the risk of unnecessary medical visits or delayed treatment.

[0097] 6. Reduce healthcare costs and improve quality of life: Early warning and precise monitoring can reduce unnecessary MRI scans and hospitalizations, while helping patients better manage their diseases and improve their quality of life.

[0098] Clinical applicability: This embodiment can be widely applied in the fields of clinical monitoring, drug efficacy evaluation, and disease prognosis prediction for neuroimmunological diseases. Specific applications include: 1. Home health monitoring: Patients can conduct daily monitoring at home, reducing the number of hospital visits.

[0099] 2. Clinical trial evaluation: Provide objective and quantitative efficacy evaluation indicators for clinical trials of new drugs for neuroimmunological diseases.

[0100] 3. Personalized treatment: The immunotherapy regimen is dynamically adjusted based on monitoring results to achieve precision medicine.

[0101] 4. Preventive intervention: Identify high-risk patients in the subclinical stage and provide preventive treatment.

[0102] 5. Medical resource optimization: By risk stratification, optimize the allocation of medical resources and improve medical efficiency.

[0103] Based on the same inventive concept, such as Figure 3 As shown in the illustration, this application also provides a neuroimmune disease monitoring and early warning device, comprising: Brain-computer interface devices are used to collect multi-channel electroencephalogram (EEG) signals from target subjects.

[0104] A cloud server is used to execute the aforementioned neuroimmune disease monitoring and early warning methods.

[0105] The first terminal is used to display the first user's data and record the symptom log entered by the first user; the first user's data includes multi-channel EEG signals and warning information.

[0106] In another exemplary embodiment of this application, the neuroimmune disease monitoring and early warning device further includes: The second terminal is used to allow a second user to remotely view the data of the first user and adjust the warning threshold.

[0107] In another exemplary embodiment of this application, the above modules are connected via a data bus to form a closed-loop monitoring and early warning system.

[0108] The brain-computer interface device is responsible for collecting, preprocessing and transmitting multi-channel EEG signals. It uses existing mature brain-computer interface devices (non-invasive EEG acquisition devices) and includes electrode caps, signal amplifiers, analog-to-digital converters and wireless transmission units.

[0109] The technical parameters are: sampling rate ≥256Hz, number of channels ≥16, bandwidth 0.5-100Hz, and common-mode rejection ratio >100dB.

[0110] Execution process: First, the electrode caps are placed according to the international 10-20 system, focusing on covering sites such as F3, F4, Fz, C3, C4, Cz, P3, P4, and Pz.

[0111] Then, multichannel EEG signals were collected during resting state (5 minutes with eyes closed and 5 minutes with eyes open) and during sleep (throughout the night).

[0112] Next, preprocessing is performed, including power frequency interference filtering, baseline drift correction, and artifact removal.

[0113] In another exemplary embodiment of this application, the neuroimmune disease monitoring and early warning device is a user interaction and cloud management module, used to provide a user interface, perform data storage, support doctor access, and system management; the user interaction and cloud management module includes: The mobile interactive terminal (first terminal) is used to display monitoring data to the user, receive and display the early warning information output by the disease risk assessment and early warning module, and record the symptom log entered by the user. A cloud server is used to store the raw data acquired by the EEG signal acquisition module, the processing results of the neuroimmune feature extraction module and the state evolution analysis module, and the early warning records output by the disease risk assessment and early warning module. The doctor's workstation (second terminal) allows doctors to remotely access and view patient data, adjust warning thresholds in the disease risk assessment and warning module, and provide medical advice. The system management backend is used for user account management, access control, and system data backup.

[0114] In another exemplary embodiment of this application, multiple virtual modules can be set up on the cloud server, including: an EEG signal acquisition module, a neuroimmune feature extraction module, a neuroimmune state modeling module, a state evolution analysis module, a disease risk assessment and early warning module, and a user interaction and cloud management module.

[0115] The EEG signal acquisition module acquires processed multi-channel EEG signals and transmits them to the feature extraction module via Bluetooth or WiFi.

[0116] The neuroimmune feature extraction module is used to extract neuroinflammation-related oscillation features, immune-related functional connectivity features, and neuroimmune state evolution features from the multi-channel EEG signals, and construct a neuroimmune feature vector.

[0117] Data flow: The feature vector F' output by this module will be transmitted to the neuroimmune state modeling module as input data for constructing the neuroimmune functional state space.

[0118] The neuroimmune state modeling module is used to map multidimensional features to a low-dimensional continuous state space to form a neuroimmune functional state vector.

[0119] The state evolution analysis module is used for: The difference between the current neuroimmune function state vector and the individual baseline state vector is calculated as the state offset; the individual baseline state vector is the average of several neuroimmune function state vectors within the initial monitoring period. Based on the neuroimmune function state vectors of historical periods, a pre-trained time-series prediction model is used to predict the neuroimmune function state vector of the current period, and the state prediction error between the predicted value and the actual constructed neuroimmune function state vector of the current period is calculated. The evolutionary instability is calculated based on the state offset and the state prediction error.

[0120] Data Relationship: Receives the state vector S from the neuroimmune state modeling module as the basic data for analysis.

[0121] The disease risk assessment and early warning module is used to calculate and output a health risk early warning index based on the state offset and the evolutionary instability; the early warning index is used to determine early warning information.

[0122] The overall data flow is as follows: EEG signal acquisition module → neuroimmune feature extraction module; neuroimmune feature extraction module → neuroimmune state modeling module; neuroimmune state modeling module → state evolution analysis module; state evolution analysis module → disease risk assessment and early warning module; disease risk assessment and early warning module → user interaction and cloud management module; user interaction and cloud management module → various modules (parameter feedback, system configuration).

[0123] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 4 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and databases. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media to run. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for monitoring and early warning of neuroimmunological diseases.

[0124] Those skilled in the art will understand that Figure 4 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0125] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0126] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0127] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use and processing of the relevant data are carried out in compliance with the relevant data protection laws and policies of the country where the location is located, and with the authorization granted by the owner of the corresponding device.

[0128] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0129] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0130] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0131] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A monitoring and early warning device for neuroimmune diseases, characterized in that, include: Brain-computer interface devices are used to collect multi-channel electroencephalogram (EEG) signals from target subjects. A cloud server is connected to the brain-computer interface device; A first terminal, connected to the cloud server, is used to display first user data and record symptom logs input by the first user; the first user data includes multi-channel EEG signals and warning information. The cloud server is specifically used for: Acquire multi-channel EEG signals from the target subject; From the multi-channel EEG signals, neuroinflammation-related oscillation features, immune-related functional connectivity features, and neuroimmune state evolution features are extracted, and a neuroimmune functional state vector is constructed. The difference between the current neuroimmune function state vector and the individual baseline state vector is calculated as the state offset; wherein, the individual baseline state vector is the mean of several neuroimmune function state vectors within the initial monitoring period; the initial monitoring period includes several consecutive stable monitoring cycles; Calculate the state prediction error between the predicted value and the neuroimmune function state vector of the current time period; the predicted value is the predicted value of the neuroimmune function state vector of the current time period after inputting the neuroimmune function state vector of the historical time period into the pre-trained time series prediction model. The evolutionary instability is calculated based on the state offset and the state prediction error; A comprehensive risk index is calculated based on the state offset and the evolutionary instability; the comprehensive risk index is used to determine early warning information. The formula for calculating the comprehensive risk index is as follows: in, , These are the weighting coefficients. , For the normalized threshold, for The overall risk index for the period. for The magnitude of the state offset over a time period. for Evolutionary instability over time; The formula for calculating the evolutionary instability is: in, for Evolutionary instability over time; for The magnitude of the state offset during a time period; and These are the weighting coefficients; The standard deviation of the state prediction error sequence; for State prediction error for a given time period; for State prediction error for a given time period; The neuroinflammatory-related oscillation characteristics include: low-frequency energy ratio, Rhythm stability index and frequency drift coefficient; The immune-related functional connectivity features include: interbrain coherence, phase synchronization indices, and functional network topology parameters; The evolutionary characteristics of the neuroimmune state include: characteristic time drift rate, volatility instability index, and mutation detection index.

2. The neuroimmune disease monitoring and early warning device according to claim 1, characterized in that, From the multi-channel EEG signals, neuroinflammation-related oscillation features, immune-related functional connectivity features, and neuroimmune state evolution features are extracted, and a neuroimmune functional state vector is constructed, specifically including: Neuroinflammatory-related oscillation features were extracted from the multi-channel EEG signals. Immune-related functional connectivity features were extracted from the multi-channel EEG signals. The evolutionary features of the neuroimmune state were extracted from the multi-channel electroencephalogram (EEG) signals. Based on the neuroinflammation-related oscillation characteristics, the immune-related functional connectivity characteristics, and the neuroimmune state evolution characteristics, a neuroimmune feature vector is constructed. A nonlinear dimensionality reduction method is used to map the neuroimmune feature vectors to a low-dimensional continuous state space to obtain the neuroimmune functional state vectors.

3. The neuroimmune disease monitoring and early warning device according to claim 1, characterized in that, The low-frequency energy ratio The calculation formula is: in, , , , These represent the total energy of EEG signals in the 0.5Hz-4Hz, 4Hz-8Hz, 8Hz-13Hz, and 13Hz-30Hz frequency bands, respectively. The Rhythm stability index The calculation formula is: ,in, The power spectral density in the 8Hz-13Hz frequency band. The coefficient of variation; The frequency drift coefficient The standard deviation of the main peak frequency in the 8Hz-13Hz band within a preset time window is calculated using the following formula: ,in, for time The main peak frequency of the frequency band. Standard deviation, For analysis of time windows; The coherence between brain regions The mean frequency domain coherence of the left frontal-left parietal region connection and the right frontal-right parietal region connection within the 4Hz-45Hz frequency band is calculated using the following formula: ,in, The frequency domain coherence function is calculated in the 4Hz-45Hz frequency band; F3-P3 is the connection between the left forehead and left top region; F4-P4 is the connection between the right forehead and right top region. The phase synchronization index The calculation formula is: ,in, Total number of brain regions; for Time-phase brain regions and brain regions The instantaneous phase difference between them; The functional network topology parameters The calculation formula is: in, The clustering coefficients of the actual brain network, The clustering coefficients are those of a random network with the same number of nodes as the actual brain network. The clustering coefficient is the number of nodes of a regular network that has the same number of nodes as the actual brain network. The actual brain network is a topology that reflects the real-time integration state of brain function, constructed with EEG acquisition channels as nodes and the synchronization of EEG signals between channels as connecting edges. The characteristic time drift rate The calculation formula is: ,in, The slope of the linear regression; The size of the time window; For a feature in continuous A series of average values ​​over a period of time. Indicates the first The average value of the characteristic over a period of time. Indicates the first The average value of a feature described over a period of time. Indicates the first The average value of a feature described in each time period; The volatility instability index The calculation formula is: in, Let be the standard deviation. Let be the mean; The mutation detection index The calculation formula is: ,in, To detect mutation points using the cumulative sum algorithm, For the detection window, For a feature in continuous A series of average values ​​over a period of time. Indicates the first The average value of the characteristic over a period of time. Indicates the first The average value of the characteristic over a given time period.

4. The neuroimmune disease monitoring and early warning device according to claim 1, characterized in that, Also includes: Early warning information is generated based on the comprehensive risk index and early warning threshold.

5. The neuroimmune disease monitoring and early warning device according to claim 1, characterized in that, Also includes: The second terminal is used to allow a second user to remotely view the data of the first user and adjust the warning threshold.

Images