Brain-computer interface motor rehabilitation system parameter adaptive adjustment system and method

Abstract

The application discloses a brain-computer interface motion rehabilitation system parameter self-adaptive adjustment system and method, a data receiving and preprocessing module is configured to receive and preprocess electroencephalogram signals of a patient; an index calculation module is configured to determine a cognitive state index of the patient based on the preprocessed electroencephalogram signals, determine a deviation value of the cognitive state compared with a corresponding baseline, and if the deviation value exceeds a preset standard deviation, generate state change information and send the state change information to a training parameter adjustment module; the training parameter adjustment module is configured to retrieve a pre-associated adjustment strategy according to the state change information, and self-adaptively adjust the motion rehabilitation system parameters according to the adjustment strategy. The application can make the rehabilitation system parameters self-adaptively adjusted according to the rehabilitation process of the patient by monitoring the cognitive state index in real time, reduce the situation of rehabilitation interruption or invalid training caused by unsuitable parameters, and thus improve the continuity and efficiency of rehabilitation training.

Description

Technical Field

[0001] This application belongs to the field of brain-computer interface technology, specifically relating to a brain-computer interface-based adaptive adjustment system and method for motor rehabilitation system parameters. Background Technology

[0002] Brain-computer interface (BCI) technology aims to establish an information and control pathway between the human brain and the external environment using computers, independent of peripheral nerves and muscle tissue. It plays a crucial role in the rehabilitation of upper limb motor dysfunction after stroke. BCI technology based on motor imagery (MI) electroencephalogram (EEG) signals has been widely applied clinically and proven to effectively promote neural function remodeling. However, existing MI-BCI rehabilitation systems are mostly based on fixed training procedures and cannot adaptively adjust training parameters based on the patient's real-time cognitive state. Summary of the Invention

[0003] This application addresses the technical problem that current brain-computer interface motor rehabilitation systems cannot adaptively adjust training parameters based on the patient's real-time cognitive state, and provides a brain-computer interface motor rehabilitation system parameter adaptive adjustment system and method.

[0004] To achieve the above technical objectives, this application adopts the following technical solution.

[0005] In a first aspect, embodiments of this application provide a brain-computer interface motor rehabilitation system parameter adaptive adjustment system, comprising:

[0006] The data receiving and preprocessing module is configured to receive and preprocess the patient's electroencephalogram (EEG) signals.

[0007] The index calculation module is configured to determine the patient's cognitive state index based on the preprocessed EEG signal, determine the deviation value of the cognitive state from the corresponding baseline, and if the deviation value exceeds the preset standard deviation, generate state change information and send the state change information to the training parameter adjustment module.

[0008] The training parameter adjustment module is configured to retrieve a pre-associated adjustment strategy based on the state change information, and to adaptively adjust the parameters of the sports rehabilitation system according to the adjustment strategy.

[0009] Furthermore, the cognitive state index includes a mental fatigue index, and the index calculation module is configured to determine the expression for the mental fatigue index as follows:

[0010]

[0011] Among them, EI is an indicator of mental fatigue, E β E represents the proportion of beta wave power in the total power of the preselected frequency band in the EEG signal. α E represents the proportion of alpha wave power in the total power of the preselected frequency band in the electroencephalogram (EEG) signal. θ The theta wave power in the EEG signal accounts for the total power of the preselected frequency band, which includes the beta wave band, alpha wave band, and theta wave band.

[0012] Furthermore, the index calculation module is specifically configured to: compare the mental fatigue index with the mental fatigue index baseline to obtain the mental fatigue index deviation value; if the mental fatigue index deviation value exceeds the preset standard deviation, generate first state change information and send the first state change information to the training parameter adjustment module.

[0013] The training parameter adjustment module is specifically configured to: retrieve a first adjustment strategy associated with the first state change information, and adjust the first set of exercise rehabilitation system parameters according to the first adjustment strategy;

[0014] Furthermore, the cognitive state indicators include psychological load indicators;

[0015] The data receiving and preprocessing module is configured to: perform data preprocessing on the electroencephalogram (EEG) signal, extract non-overlapping EEG signal segments according to a preset time length, perform high-pass filtering and band-pass filtering on each segment of the EEG signal in sequence, and divide it into four frequency bands: θ, α, β, and γ.

[0016] The indicator calculation module is configured to determine the psychological load indicator, including:

[0017] The power spectral density of each frequency band is determined based on the preprocessed EEG signal, and the total power in each frequency band is determined according to the power spectral density to obtain the PSD feature matrix.

[0018] The PSD feature matrix is ​​input into the trained regression model, and the mapping value is obtained using the regression model. The mapping value is then used as the psychological load index.

[0019] Furthermore, the indicator calculation module is also configured as follows:

[0020] The psychological load index is compared with the psychological load index baseline to obtain the psychological load index deviation value. If the psychological load index deviation value exceeds the preset standard deviation, a second state change information is generated and the second state change information is sent to the training parameter adjustment module.

[0021] The training parameter adjustment module is specifically configured to: retrieve a second adjustment strategy associated with the second state change information, and adjust the parameters of the second exercise rehabilitation system according to the second adjustment strategy.

[0022] Furthermore, the training method for the regression model includes:

[0023] Historical EEG signals were acquired during the calibration phase;

[0024] The historical EEG signals are preprocessed, and non-overlapping EEG signal segments are extracted according to a preset time window. High-pass filtering and band-pass filtering are then performed on each segment of the EEG signal in sequence.

[0025] The power spectral density is determined based on the filtered EEG signal, and the total power in each of the four divided frequency bands (θ, α, β, γ) is determined based on the power spectral density to obtain the historical PSD feature matrix.

[0026] Label the load category of the historical EEG signals described in each segment;

[0027] The historical PSD feature matrix and the corresponding load category are input into the regression model to train the regression model. The regression model includes an LDA module and a Softmax layer. The LDA module obtains the mapping value based on the input historical PSD feature matrix, and the Softmax layer determines the corresponding category based on the mapping value.

[0028] Furthermore, the cognitive state index includes a mental effort index, and the index calculation module is configured to determine the mental effort index by: dividing the electroencephalogram signal into specific frequency bands and determining the sum of power within the set frequency bands as the mental effort index.

[0029] Furthermore, the cognitive state indicators also include stress perception indicators, and the indicator calculation module is configured to determine the stress perception indicators, including:

[0030] For the EEG signals within each time window, the correlation between the power of beta waves and delta waves was assessed using the Pearson correlation coefficient.

[0031] Determine the difference in average alpha wave power between two pre-selected EEG channels;

[0032] Determine the fuzzy entropy difference in energy on the two pre-selected EEG channels;

[0033] Principal component analysis is used to map the correlation, alpha wave average power difference, and fuzzy entropy difference to a single-dimensional feature space to obtain mapping features, which are then used as the pressure perception index.

[0034] Furthermore, the indicator calculation module is specifically configured to: compare the mental effort indicator with the mental effort indicator baseline to obtain a mental effort indicator deviation value, or compare the stress perception indicator with the stress perception indicator baseline to obtain a stress perception indicator deviation value; if the mental effort indicator deviation value exceeds a preset standard deviation, generate third state change information; if the stress perception indicator deviation value is less than a preset standard deviation, generate fourth state change information; and send the third state change information and the fourth state change information to the training parameter adjustment module.

[0035] The training parameter adjustment module is specifically configured to: retrieve the associated third adjustment strategy based on the state change information and the fourth state change information, and adjust the third exercise rehabilitation system parameters according to the third adjustment strategy.

[0036] Secondly, embodiments of this application provide a method for adaptive adjustment of parameters of a brain-computer interface motor rehabilitation system, comprising: receiving an electroencephalogram (EEG) signal from a patient and performing preprocessing;

[0037] Based on the preprocessed EEG signal, the patient's cognitive state index is determined, and the deviation value of the cognitive state from the corresponding baseline is determined. If the deviation value exceeds the preset standard deviation, state change information is generated and sent to the training parameter adjustment module.

[0038] Based on the state change information, a pre-associated adjustment strategy is retrieved, and the parameters of the sports rehabilitation system are adaptively adjusted according to the adjustment strategy.

[0039] Compared with existing technologies, the brain-computer interface motor rehabilitation system and method provided in this application, by collecting patients' electroencephalogram (EEG) signals to determine cognitive state indicators, can gain a deeper understanding of each patient's real-time psychological and cognitive status during the rehabilitation process. As rehabilitation training progresses, the patient's cognitive state and physical functions continuously change. This application monitors cognitive state indicators in real time, enabling the rehabilitation system parameters to adaptively adjust according to the patient's rehabilitation progress. This reduces rehabilitation interruptions or ineffective training due to parameter inappropriateness, thereby improving the continuity and efficiency of rehabilitation training. Attached Figure Description

[0040] The accompanying drawings described herein are for illustrative purposes only and are not intended to limit the scope of this application in any way. Furthermore, the shapes and scales of the components in the drawings are merely illustrative to aid in understanding this application and do not specifically limit the shapes and scales of the components. Those skilled in the art, guided by the teachings of this application, can select various possible shapes and scales to implement this application according to specific circumstances. In the drawings:

[0041] Figure 1This is a schematic diagram of the adaptive adjustment system for parameters of the brain-computer interface motor rehabilitation system provided in the embodiment.

[0042] Figure 2 This is a schematic diagram illustrating the principle of the indicator calculation module in the embodiment;

[0043] Figure 3 A schematic diagram of the upper limb rehabilitation training process provided in the embodiment of the brain-computer interface motor rehabilitation system parameter adaptive adjustment system;

[0044] Figure 4 This is a schematic diagram of the adaptive adjustment method for parameters of a brain-computer interface motor rehabilitation system provided in an embodiment. Detailed Implementation

[0045] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in 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 should fall within the scope of protection of this application.

[0046] In the description of this application, 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. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more features.

[0047] To address the current limitations of brain-computer interface (BCI) motor rehabilitation systems in adaptively adjusting training parameters based on the patient's real-time cognitive state, this application provides a BCI motor rehabilitation system parameter adaptive adjustment system. This system can identify motor intentions based on the electroencephalogram (EEG) signals generated when the patient performs motor imagery, determine the patient's cognitive state, and adaptively adjust the motor rehabilitation system parameters according to the identification results and the patient's real-time cognitive state.

[0048] like Figure 1 As shown, the system provided in this embodiment includes a data receiving and preprocessing module, an index calculation module, and a training parameter adjustment module.

[0049] The data receiving and preprocessing module is configured to receive EEG data transmitted from the EEG acquisition module in real time and perform preprocessing (which may include filtering, dividing into time windows, or dividing into different frequency bands according to frequency).

[0050] The index calculation module is configured to determine the patient's cognitive state index based on the preprocessed EEG signal, determine the deviation value of the cognitive state from the corresponding baseline, and if the deviation value exceeds the preset standard deviation, generate state change information and send the state change information to the training parameter adjustment module.

[0051] The training parameter adjustment module is configured to retrieve pre-associated adjustment strategies based on state change information and adaptively adjust the parameters of the motor rehabilitation system according to these strategies. The module can achieve parameter adjustment based on motion intention recognition and / or gamified interaction paradigms.

[0052] In some embodiments, the brain-computer interface motor rehabilitation system parameter adaptive adjustment system further includes a data storage and report generation module, which can generate report data and send it to the gamified interaction paradigm after training is completed.

[0053] like Figure 2 As shown, the cognitive state indicators determined by the indicator calculation module may include mental fatigue calculation, psychological load calculation, mental effort calculation, and stress perception calculation modules. The core objective of the indicator calculation module is to map EEG signals into cognitive state curves and compare these curves with corresponding baselines to obtain state change information. Specifically, the data receiving and preprocessing module extracts EEG signal segments of a specific time window length at fixed time intervals Δt. The indicator calculation module uses the mental fatigue calculation, psychological load calculation, mental effort calculation, and stress perception calculation modules to calculate the indicator values ​​of these EEG segments in the four dimensions of mental fatigue, psychological load, mental effort, and stress perception, respectively. Each indicator value is then compared with its corresponding value to obtain a deviation value. If the deviation value exceeds a preset standard deviation, state change information is sent to the training parameter adjustment module.

[0054] This application utilizes cognitive state indicators calculated from electroencephalogram (EEG) signals and can capture real-time changes in these states. Therefore, it can provide personalized interventions for each individual, helping to improve training effectiveness or work efficiency.

[0055] In some embodiments, cognitive state indicators include mental fatigue (MF) indicators.

[0056] Mental fatigue arises when a task requires a high level of attention and focus. Research indicates that there are four main categories of factors that trigger mental fatigue: 1) drowsiness, referring to a decline in cognitive attention accompanied by a desire to sleep; 2) the transition phase, defined as the shift from wakefulness to sleep; 3) prolonged tasks, where human performance declines and mental fatigue begins to appear as time is spent on a task; and 4) task engagement, referring to an active and excited state influenced by workload, task duration, motivation, and emotions. Specifically, in the context of rehabilitation training, this study will focus on mental fatigue induced by task engagement.

[0057] As an example, the calculation steps for the Mental Attention (MF) index are as follows:

[0058] 1) Acquire a signal segment containing the four electrode positions O3, O4, F3, and F4 within a time window (e.g., a time window of 20 seconds);

[0059] 2) Use the Hanning window to smooth signal segments;

[0060] 3) Calculate the power spectrum using FFT, and combine the band power to calculate the total power across the three bandwidths (θ: 4-7Hz, α: 8-12Hz, β: 13-30Hz); 4) Divide the bandwidth power by the total power to produce the percentage power E. θ E α and E β 5) The Participation Index (EI) is calculated using the following formula:

[0061]

[0062] In practical application embodiments, the index calculation module uses the EI curve calculated from the EEG signals collected during the calibration phase of the brain-computer interface motor rehabilitation system parameter adaptive adjustment system as the baseline, and continuously measures the deviation of the EI value from the baseline during formal training. For every increase or decrease of 0.2 standard deviations, a first state change information is generated. This first state change information is sent to the training parameter adjustment module as the basis for parameter adjustment.

[0063] The training parameter adjustment module can be configured to: retrieve a first adjustment strategy associated with the first state change information, and adjust the parameters of the first motor rehabilitation system according to the first adjustment strategy; the first adjustment strategy includes: if the mental fatigue index is greater than the preset standard deviation according to the first state change information, then increase the rest time setting ratio of the motor imagery task interval; if the mental fatigue index is less than the preset standard deviation according to the first state change information, then decrease the rest time setting ratio of the motor imagery task interval.

[0064] In some embodiments, cognitive state indicators include mental workload (MW) indicators.

[0065] Under high-intensity workloads, an individual's cognitive resources are often severely depleted. This not only leads to desensitization to auditory warning signals but also results in the complete ignoring of all input information, a significant slowing of decision-making processes, and an overall decline in alertness. Particularly in rehabilitation settings, this resource exhaustion manifests as ignoring task cues or difficulty in motor imagery. Research indicates that as task complexity increases, the alpha band power spectral density (PSD) in the parietal lobe decreases, while the theta band PSD in the frontal lobe increases accordingly.

[0066] As an example, the steps for calculating the psychological load index include:

[0067] 1) The data receiving and preprocessing module preprocesses the EEG signal, extracting non-overlapping EEG signal segments according to a preset time length (e.g., 2 seconds). Each EEG segment is then subjected to high-pass filtering (cutoff frequency 0.5Hz) and band-pass filtering (frequency range 2Hz to 50Hz) to effectively remove DC drift and enhance frequency signals closely related to cognitive activity. The signal is divided into four frequency bands: θ, α, β, and γ.

[0068] 2) PSD estimation: The index calculation module uses the Welch method to calculate the PSD of the filtered EEG signal. Then, the PSD value is integrated over four key frequency bands (θ: 4-8Hz, α: 8-12Hz, β: 12-30Hz, γ: 30-40Hz) to obtain a PSD feature matrix of shape (C,4), which lays a solid foundation for subsequent analysis. C is the number of EEG channels measured.

[0069] 3) The index calculation module inputs the PSD feature matrix as a feature vector into the trained regression model, uses the regression model to obtain the mapping value based on the input feature vector, and uses the mapping value as a psychological load index.

[0070] As an example, during formal rehabilitation training, an established regression model is used to map the PSD feature matrix of EEG signal segments to specific numerical values. Using the mean of the mapped values ​​of low-load and high-load signals during the calibration phase as a baseline, the index calculation module continuously monitors and evaluates the deviation of the EEG signal mapped values ​​from this baseline during rest and motor imagery tasks in formal training, obtaining the psychological load index deviation value. Whenever the psychological load index deviation value reaches ±0.2 standard deviations, a second state change information is generated. This second state change information is immediately transmitted to the training parameter adjustment module, serving as an important basis for dynamically adjusting training parameters, thereby achieving precise control and optimization of rehabilitation training. The psychological load index baseline is the mean of the mapped values ​​obtained using the regression model based on the PSD feature matrices of different categories of EEG signals.

[0071] The training parameter adjustment module can be configured to: retrieve a second adjustment strategy associated with the second state change information, and adjust the parameters of the second exercise rehabilitation system according to the second adjustment strategy; the second adjustment strategy includes: if the psychological load index is greater than the psychological load index baseline and exceeds the preset standard deviation according to the second state change information, then increase the setting ratio of the rehabilitation device's exercise distance; if the psychological load index is less than the psychological load index baseline and exceeds the preset standard deviation according to the second state change information, then decrease the setting ratio of the rehabilitation device's exercise distance.

[0072] In this embodiment, the training method for the regression model includes: acquiring historical EEG signals during the calibration phase; performing data preprocessing on the historical EEG signals, extracting non-overlapping EEG signal segments according to a preset time window, and sequentially performing high-pass filtering and band-pass filtering on each segment of the EEG signal; determining the power spectral density based on the filtered EEG signals, determining the total power in each of the four divided frequency bands (θ, α, β, γ) based on the power spectral density, and obtaining the historical PSD feature matrix.

[0073] The historical EEG signals were labeled with load categories based on the activity states corresponding to the EEG signals during the calibration phase. For example, signals during rest and preparation were labeled as low load, while signals during motor imagery tasks (such as left and right hand motor imagery) were considered high load. To ensure the validity and representativeness of the data, five left and right hand motor imagery tasks were collected during the calibration phase, each lasting four seconds, with each sample consisting of a two-second signal segment, resulting in a total of 20 high load samples. Simultaneously, the same number of samples were collected during the rest and preparation phases for constructing and training the regression model.

[0074] The historical PSD feature matrix and the corresponding load category are input into the regression model to train the regression model. The regression model includes an LDA module and a Softmax layer. The LDA module obtains the mapping value based on the input historical PSD feature matrix, and the Softmax layer determines the corresponding category based on the mapping value.

[0075] The regression model uses Linear Discriminant Analysis (LDA) to determine the optimal subspace projection, aiming to reduce the historical PSD feature matrix of high-dimensional EEG data to one dimension, i.e., the key dimension reflecting user workload. Furthermore, a Softmax layer is used to optimize this linear projection using logistic regression, assuming that the class-conditional probabilities of a given projection follow a logistic model:

[0076]

[0077] The weight vector w and bias b are adjusted by maximizing data likelihood to ensure a high degree of consistency between the data and the logistic model distribution of the labels. Leave-one-out cross-validation is used to comprehensively evaluate the performance of the regression model, ensuring its stability and reliability.

[0078] In other embodiments, cognitive state indicators also include mental effort (ME) indicators.

[0079] Mental effort is a key indicator for measuring the degree to which an individual allocates cognitive resources when performing a specific task, and it is closely linked to the level of task engagement and participation. From a cognitive science perspective, mental effort is not only closely related to the degree of attention focused on external stimuli, but also involves the in-depth processing of task-related information. This process is accompanied by enhanced activity in specific areas of the brain, thereby promoting improved neural efficiency.

[0080] As an example, determining the mental effort index includes: based on the frequency bands into which the EEG signal is divided by the data receiving and preprocessing module, the index calculation module determines the sum of power within the set frequency bands as the mental effort index.

[0081] Specifically, the process includes the following steps: Signal acquisition and frequency band segmentation. Electroencephalogram (EEG) signals from four key locations—Fz, Cz, Pz, and O1—are selected as the analysis targets. These signals are then subdivided into five specific frequency bands: the delta band (0-3Hz), theta band (4-7Hz), the alpha band (8-12Hz), the beta1 band (13-22Hz), and the beta2 band (23-30Hz). The sum of the powers within each band is calculated to comprehensively capture the characteristics of brain activity.

[0082] As an example, in order to establish an accurate reference baseline, the EEG signals during preparation were segmented into non-overlapping time windows of 2 seconds, and the power values ​​of the five frequency bands mentioned above were calculated in each window as baseline data for subsequent analysis.

[0083] Existing research indicates that increased mental effort is often accompanied by significant changes in the power of the theta and β2 bands in EEG. In this embodiment, the power dynamics of these two bands are continuously monitored at 2-second intervals. Once a difference of 0.2 standard deviations from the baseline in the mental effort index is detected, a third-state change information is generated and sent to the training parameter adjustment module. This mechanism provides a scientific basis for dynamically adjusting rehabilitation training parameters, aiming to achieve precise control and optimization of the rehabilitation training process, thereby maximizing training effectiveness.

[0084] In some embodiments, cognitive state indicators also include stress perception indicators.

[0085] Emotions play a crucial role in overall human performance, significantly influencing cognitive function, decision-making, and individual performance, with stress often stemming from emotional burden. In rehabilitation training settings, continuous monitoring of stress and timely adjustment of treatment parameters are of paramount importance for improving patients' rehabilitation outcomes.

[0086] EEG-based stress level monitoring techniques typically employ methods including analyzing the asymmetry of frontal lobe activity and the correlation between beta and delta waves. It is known that frontal alpha wave power is negatively correlated with task-related activity; therefore, observing the asymmetry in alpha wave power can effectively assess stress status.

[0087] As an example, the method for determining stress perception indicators includes the following steps:

[0088] 1) The data receiving and preprocessing module performs data preprocessing: with a time window of 5 seconds and an interval of 2 seconds, it extracts segments from the EEG signal for calculating the pressure index, and specifically selects channels F3 and F4 as signal sources.

[0089] 2) The index calculation module performs spectrum analysis: The Welch method of FFT is applied to extract the power changes of the α, β, δ and θ frequency bands from each 5-second segment, with the window length set to 1 second.

[0090] 3) The index calculation module performs correlation analysis: For the δ and β frequency bands, the relative power change is obtained by dividing their original power by the sum of the total spectral power. The Pearson correlation coefficient is then used to assess the correlation between the β wave and δ wave power. The calculation formula is as follows:

[0091]

[0092] Where x and y represent the relative power change curves of the β wave and δ wave in the 5-second signal segment, respectively.

[0093] 4) The index calculation module assesses alpha wave asymmetry: It measures the energy and complexity asymmetry of alpha waves by calculating the difference in average alpha wave power and fuzzy entropy (FuzzEn) between channels F3 and F4. Studies show that prefrontal asymmetry increases under stress conditions, as calculated below:

[0094] AlphaAsym(t)=P F4 (t)-P F3 (t) (4)

[0095] FuzzEnAsym(t)=FuzzEn F4 (t)-FuzzEn F3 (t) (5) Among them, the fuzzy entropy of information sequence is an effective tool for measuring the complexity of time series.

[0096] 5) After the above processing, each 5s EEG signal segment is mapped into a 3D feature vector, mapping the signal during the calibration period into a feature sequence, as follows:

[0097]

[0098] Where Stress(t) fit Let r(t) be the feature sequence at time t, r(t) be the correlation between the power of β wave and δ wave at time t, AlphaAsym(t) be the difference in average power of α wave on two pre-selected EEG channels (F3 channel and F4 channel), and FuzzEnAsym(t) be the difference in fuzzy entropy of energy on two pre-selected EEG channels (F3 channel and F4 channel).

[0099] As an example, PCA is used to analyze the above time-t characteristic sequence Stress(t). fit The stress perception index is mapped to a single-dimensional feature space as a baseline, and the same mapping matrix is ​​used to map the signal in the formal training to this feature space to obtain the stress perception index. The degree of deviation of the stress perception index from the baseline is monitored to obtain the stress perception index deviation value. When the stress perception index deviation value reaches ±0.2 standard deviations, the fourth state change information is generated and transmitted to the training parameter adjustment module in real time as the key basis for dynamically adjusting the training parameters, thereby realizing the precise control and optimization of the rehabilitation training process.

[0100] The training parameter adjustment module can be configured to retrieve a associated third adjustment strategy based on third and fourth state change information, and adjust the parameters of the third motor rehabilitation system according to the third adjustment strategy. The third adjustment strategy includes: if the mental effort index is determined to be greater than the baseline of the psychological load index by more than a preset standard deviation based on the third state change information, or if the stress perception index is determined to be greater than the baseline of the stress perception index by more than a preset standard deviation based on the fourth state change information, then the setting ratio of the motor intention recognition threshold is decreased; if the mental effort index is determined to be less than the baseline of the psychological load index by more than a preset standard deviation based on the third state change information, and the stress perception index is determined to be less than the baseline of the stress perception index by more than a preset standard deviation based on the fourth state change information, then the setting ratio of the motor intention recognition threshold is increased.

[0101] As an example, the training process in the formal training phase is as follows: Figure 3 As shown, the index calculation module calculates the baselines of the aforementioned four cognitive indices, including the mean and standard deviation, during the preparation and model calibration phases. During the formal training phase, it continuously sends monitoring results (i.e., state change information) to the training parameter adjustment module. Changes in state change information are recorded in units of 0.2 standard deviations; that is, if the change is less than ±0.2 standard deviations, the value from the previous time step is maintained. After completing the model calibration phase, the training parameter adjustment module continuously receives cognitive monitoring information x(t) = MF(t), MW(t), ME(t), Stress(t)} transmitted by the index calculation module during the formal training phase and performs dynamic parameter adjustment accordingly.

[0102] Some implementation examples propose to adjust the following training parameters: 1) the rest time t between tasks. rest 2) Movement intention recognition threshold T; 3) Movement distance S of upper limb rehabilitation equipment. By adjusting these parameters, training intensity and difficulty can be effectively controlled, thereby improving the patient's training experience. Specific adjustment strategies are as follows:

[0103] (1) Rest time t between tasks rest Regulation strategy

[0104] Adjustment Basis: This strategy primarily relies on changes in the mental fatigue (MF) index to dynamically adjust rest time t. baseline The MF index is an important indicator for assessing a patient's fatigue status. An increase in its level usually means that the patient is beginning to feel fatigued. Therefore, it is necessary to increase rest time to effectively alleviate fatigue and ensure training effectiveness and patient safety.

[0105] Adjustment Strategies: 1) Initial Setup: Before training begins, professionals manually set a standard rest time and a threshold for MF (Fatigue Factor) elevation based on the characteristics of the training task and the patient's physical condition. This threshold serves as a benchmark for determining whether the patient is beginning to feel fatigued. 2) Fatigue Monitoring and Increased Rest Time: During training, MF changes are continuously monitored. When the MF index rises and exceeds the preset threshold, it indicates that the patient has begun to show signs of fatigue. At this point, the rest time should be gradually increased, ideally by 5% to 10% after each trial, until the MF index shows a downward trend. This step aims to reduce the patient's fatigue by increasing rest, ensuring that training can continue. 3) Fatigue Relief and Reduced Rest Time: As the MF index begins to decline from a high level, it indicates that the patient's fatigue has been relieved to some extent. At this point, the rest time should be gradually reduced, ideally by 5% to 10% after each trial, to promote improved training efficiency. 4) Recovery Phase and Rapid Reduction of Rest Time: When the MF index returns to baseline or lower, it indicates that the patient's fatigue has been significantly relieved. At this stage, rest time should be reduced more rapidly, ideally by 10% to 20% after each trial, until the initial standard rest time is restored. This step aims to restore a normal training rhythm as quickly as possible. 5) Optimization at baseline: When the MF index is at or below baseline, and the mental workload (MW) index is also below baseline, it indicates that the patient is currently in good condition and has the potential to withstand higher training intensity. At this time, rest time should be gradually reduced, ideally by 5% to 10% after each trial, until the rest time is reduced to half of the standard rest time or the MF index returns to baseline. This step aims to further optimize training effects and improve the patient's training efficiency.

[0106] (2) Motion Intent Recognition Threshold (T) Adjustment Strategy

[0107] Adjustment rationale: The adjustment of the motor intention recognition threshold (T) is primarily based on changes in the patient's mental effort (ME) and stress perception indices. When the ME or stress indices are elevated, it may indicate that the patient is experiencing high cognitive stress. In this case, lowering the threshold can make the system more sensitive to capturing the patient's motor intentions, thereby reducing the patient's burden and improving the training experience.

[0108] Adjustment methods: 1) Initial calibration: Before training begins, based on the patient's motor ability and training goals, a professional sets an initial motor intention recognition threshold T. baseline2) High Load Response: During training, if ME or Stress continuously rises and exceeds the preset high threshold, it indicates that the patient is under high load. At this time, the motor intention recognition threshold should be gradually lowered, with each adjustment reducing it by 1% to 5%, until the ME and Stress indices return to levels below the threshold. This step aims to lower the recognition threshold, allowing the patient to complete the movement more easily and reducing physical and mental stress. 3) Load Stabilization and Threshold Fine-tuning: When the patient is judged to have correctly imagined the movement k times consecutively, and the training process is smooth, the current threshold should be maintained without further decrease. 4) Recovery Phase and Threshold Rebound: As training progresses, if the ME and Stress indices return to or fall below baseline levels, it indicates that the patient has adapted to the current training intensity. At this time, the motor intention recognition threshold can be gradually raised, with each adjustment increasing it by 5% to 10%, until the initially set T is reached. baseline This step aims to gradually restore normal recognition sensitivity as the patient's abilities improve, avoiding a decline in training effectiveness due to over-adaptation.

[0109] (3) Strategies for adjusting the range of motion (S) of upper limb rehabilitation equipment

[0110] Adjustment Basis: This strategy primarily uses the patient's psychological workload (MW) index to dynamically adjust the range of motion (S) of the upper limb rehabilitation equipment. Unlike traditional methods that mainly consider physical load, this strategy pays special attention to the patient's psychological burden, aiming to effectively alleviate the patient's psychological stress by adjusting the passive range of motion, while simultaneously promoting the recovery of upper limb function.

[0111] Adjustment strategies: 1) Baseline setting and monitoring: Before training begins, a basic movement distance S is set based on the patient's rehabilitation stage, upper limb function, and psychological tolerance. baseline And psychological load threshold. During training, the patient's psychological load (MW) index is continuously monitored to capture changes in their psychological state in a timely manner. 2) When the MW index rises and exceeds the preset threshold (set according to the baseline, exceeding the baseline set standard deviation), it indicates that the patient's psychological load is high and they may feel anxious, tense, or tired. At this time, contrary to the traditional view, this strategy suggests appropriately increasing S, by prolonging the patient's passive movement time to provide a "relaxed" rehabilitation experience, thereby indirectly relieving the patient's psychological stress. Specifically, when the MW index exceeds the threshold, the movement distance can be gradually increased, and it is recommended to increase it by 5% to 10% per trial until the MW index no longer rises or reaches S. baseline 200%. 3) When the MW index falls to or below the baseline level, gradually reduce S until the MW index begins to exceed the threshold again or S drops to S. baseline Increase the intensity of active exercise by 50% to promote more comprehensive rehabilitation.

[0112] Similar to the inventive concept of the brain-computer interface motor rehabilitation system parameter adaptive adjustment system provided in the above embodiments, this application also provides a brain-computer interface motor rehabilitation system parameter adaptive adjustment method, such as... Figure 4 As shown, it includes:

[0113] Step S1: Receive and preprocess the patient's electroencephalogram (EEG) signals;

[0114] Step S2: Determine the patient's cognitive state indicators based on the preprocessed EEG signals, determine the deviation value of the cognitive state from the corresponding baseline, and if the deviation value exceeds the preset standard deviation, generate state change information and send the state change information to the training parameter adjustment module.

[0115] Step S3: Retrieve the pre-associated adjustment strategy based on the state change information, and adaptively adjust the parameters of the sports rehabilitation system according to the adjustment strategy.

[0116] It is understood that the implementation methods of each step in this embodiment can refer to the implementation methods of each module in the above embodiments, and will not be repeated here.

[0117] The above provides a detailed description of the adaptive adjustment system and method for parameters of the brain-computer interface motor rehabilitation system provided in this application. Specific examples have been used in this article to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the concept of this application and should not be construed as a limitation on the scope of protection of this application.

Claims

1. A parameter adaptive adjustment system for a brain-computer interface motor rehabilitation system, characterized in that, The system includes: The data receiving and preprocessing module is configured to receive and preprocess the patient's electroencephalogram (EEG) signals. The index calculation module is configured to determine the patient's cognitive state index based on the preprocessed EEG signal, determine the deviation value of the cognitive state index from the corresponding baseline, and if the deviation value exceeds the preset standard deviation, generate state change information and send the state change information to the training parameter adjustment module. The training parameter adjustment module is configured to retrieve a pre-associated adjustment strategy based on the state change information, and to adaptively adjust the parameters of the sports rehabilitation system according to the adjustment strategy. The cognitive state indicators include psychological load indicators; The data receiving and preprocessing module is configured to: perform data preprocessing on the electroencephalogram (EEG) signal, extract non-overlapping EEG signal segments according to a preset time length, perform high-pass filtering and band-pass filtering on each segment of the EEG signal in sequence, and divide it into four frequency bands: θ, α, β, and γ. The indicator calculation module is configured to determine the psychological load indicator, including: The power spectral density of each frequency band is determined based on the preprocessed EEG signal, and the total power in each frequency band is determined according to the power spectral density to obtain the PSD feature matrix. The PSD feature matrix is ​​input into the trained regression model, and the mapping value is obtained using the regression model. The mapping value is then used as the psychological load index.

2. The adaptive parameter adjustment system for the brain-computer interface motor rehabilitation system according to claim 1, characterized in that, The cognitive state indicators include a mental fatigue indicator, and the indicator calculation module is configured to determine the expression for the mental fatigue indicator as follows: Among them, EI is an indicator of mental fatigue, E β E represents the proportion of beta wave power in the total power of the preselected frequency band in the EEG signal. α E represents the proportion of alpha wave power in the total power of the preselected frequency band in the electroencephalogram (EEG) signal. θ The theta wave power in the EEG signal accounts for the total power of the preselected frequency band, which includes the beta wave band, alpha wave band, and theta wave band.

3. The adaptive parameter adjustment system for the brain-computer interface motor rehabilitation system according to claim 2, characterized in that, The index calculation module is specifically configured to: compare the mental fatigue index with the mental fatigue index baseline to obtain the mental fatigue index deviation value; if the mental fatigue index deviation value exceeds the preset standard deviation, generate first state change information and send the first state change information to the training parameter adjustment module. The training parameter adjustment module is specifically configured to: retrieve a first adjustment strategy associated with the first state change information, and adjust the parameters of the first exercise rehabilitation system according to the first adjustment strategy.

4. The adaptive parameter adjustment system for the brain-computer interface motor rehabilitation system according to claim 1, characterized in that, The indicator calculation module is also configured to: The psychological load index is compared with the psychological load index baseline to obtain the psychological load index deviation value. If the psychological load index deviation value exceeds the preset standard deviation, a second state change information is generated and the second state change information is sent to the training parameter adjustment module. The training parameter adjustment module is specifically configured to: retrieve a second adjustment strategy associated with the second state change information, and adjust the parameters of the second exercise rehabilitation system according to the second adjustment strategy.

5. The adaptive parameter adjustment system for the brain-computer interface motor rehabilitation system according to claim 1, characterized in that, The training methods for the regression model include: Historical EEG signals were acquired during the calibration phase; The historical EEG signals are preprocessed, and non-overlapping EEG signal segments are extracted according to a preset time window. High-pass filtering and band-pass filtering are then performed on each segment of the EEG signal in sequence. The power spectral density is determined based on the filtered EEG signal, and the total power in each of the four divided frequency bands (θ, α, β, γ) is determined based on the power spectral density to obtain the historical PSD feature matrix. Label the load category of the historical EEG signals described in each segment; The historical PSD feature matrix and the corresponding load category are input into the regression model to train the regression model. The regression model includes an LDA module and a Softmax layer. The LDA module obtains the mapping value based on the input historical PSD feature matrix, and the Softmax layer determines the corresponding category based on the mapping value.

6. The adaptive parameter adjustment system for the brain-computer interface motor rehabilitation system according to claim 1, characterized in that, The cognitive state indicators include mental effort indicators. The indicator calculation module is configured to determine the mental effort indicators, including: dividing the electroencephalogram (EEG) signal into delta, theta, alpha, β1, and β2 bands, and determining the sum of power within the set bands as the mental effort indicators, wherein the delta band is 0-3Hz, the theta band is 4-7Hz, the alpha band is 8-12Hz, the β1 band is 13-22Hz, and the β2 band is 23-30Hz.

7. The adaptive parameter adjustment system for the brain-computer interface motor rehabilitation system according to claim 6, characterized in that, The cognitive state indicators also include stress perception indicators, and the indicator calculation module is configured to determine the stress perception indicators, including: For the EEG signals within each time window, the correlation between the power of beta waves and delta waves was assessed using the Pearson correlation coefficient. Determine the difference in average alpha wave power between two pre-selected EEG channels; Determine the fuzzy entropy difference in energy on the two pre-selected EEG channels; Principal component analysis is used to map the correlation, alpha wave average power difference, and fuzzy entropy difference to a single-dimensional feature space to obtain mapping features, which are then used as the pressure perception index.

8. The adaptive parameter adjustment system for the brain-computer interface motor rehabilitation system according to claim 7, characterized in that, The indicator calculation module is specifically configured to: compare the mental effort indicator with the mental effort indicator baseline to obtain a mental effort indicator deviation value, or compare the stress perception indicator with the stress perception indicator baseline to obtain a stress perception indicator deviation value; if the mental effort indicator deviation value exceeds a preset standard deviation, generate third state change information; if the stress perception indicator deviation value is less than a preset standard deviation, generate fourth state change information; and send the third state change information and the fourth state change information to the training parameter adjustment module. The training parameter adjustment module is specifically configured to: retrieve the associated third adjustment strategy based on the state change information and the fourth state change information, and adjust the parameters of the third exercise rehabilitation system according to the third adjustment strategy.

Images