Attention-enhanced cognitive closed-loop rehabilitation intervention method based on brain-computer interface

Abstract

The application discloses a brain-computer interface-based attention-enhanced cognitive closed-loop rehabilitation intervention method, which comprises the following steps: firstly, collecting electroencephalogram signals and related behavior information synchronously and completing time alignment processing; secondly, pre-processing and extracting features of the collected signals; on the basis, constructing a multi-dimensional brain state fusion model to realize comprehensive evaluation of attention and cognitive participation state; further generating control instructions through dynamic threshold determination and a gated decision mechanism, and adaptively adjusting cognitive task parameters and intervention intensity according to the control instructions; and updating system parameters through a feedback mechanism, so as to form a closed-loop intervention process. The application solves the problems of single attention representation, unstable state determination and lack of adaptive closed-loop regulation in the existing brain-computer interface cognitive intervention method, and realizes an attention-enhanced cognitive closed-loop intervention mechanism based on multi-dimensional brain state fusion.

Description

Technical Field

[0001] This invention relates to the field of information processing technology, and specifically to a brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method. Background Technology

[0002] Brain-computer interface (BCI) is a technology that enables information exchange between the brain and external systems by collecting and analyzing neural signals such as electroencephalograms (EEGs). With the development of non-invasive EEG acquisition technology, signal processing methods, and human-computer interaction systems, BCI has gradually moved from experimental research to practical applications, demonstrating significant value in fields such as neurorehabilitation, cognitive training, and assistive interaction. In the field of cognitive rehabilitation, BCI identifies brain states such as attention and cognitive load, transforming brain activity into control signals or feedback information, thereby providing a new technological path for attention regulation and cognitive ability reconstruction.

[0003] In existing technologies, brain-computer interface-based cognitive training systems typically include an EEG signal acquisition module, a signal preprocessing module, a feature extraction and state recognition module, and a feedback or control output module. EEG signals are generally acquired via scalp electrodes and, after filtering, artifact removal, and normalization, power features of specific frequency bands are extracted. Research shows that the alpha and beta bands are closely related to an individual's attention level and cognitive state; therefore, some systems construct attention indices by calculating the power relationships between different frequency bands to characterize the user's level of focus. Based on this, the system maps attention-related indicators to feedback information or control parameters, enabling cognitive training and intervention based on brain state.

[0004] In terms of training modes, existing brain-computer interface cognitive intervention methods mainly include single-task cognitive training and composite training methods combining multimodal stimulation. Single-task modes typically focus on attention maintenance or attention allocation abilities, guiding users to maintain specific brain states through continuous feedback. To enhance training effectiveness, some studies have introduced multimodal interactive environments such as virtual reality, embedding cognitive tasks into immersive scenarios. These methods enhance user engagement through visual, operational, and task-driven approaches, making the attention regulation process more closely resemble real-world situations. Related research indicates that cognitive training in multi-task or complex environments helps improve an individual's attention allocation ability and information processing efficiency.

[0005] While the aforementioned technologies have made some progress in cognitive intervention, their overall technical system still has significant shortcomings. In terms of brain state modeling, existing methods mostly rely on signals from single brain regions or single frequency bands to characterize attention levels, lacking a comprehensive portrayal of the complexity of cognitive processes. Attention, as a multi-dimensional cognitive function, involves not only sustained attention but also selective attention and executive control; relying solely on power relationships in a single frequency band is insufficient to accurately reflect its dynamic changes. Furthermore, different cognitive processes may exhibit overlapping characteristics in EEG signals, making the system susceptible to interference when recognizing specific cognitive states, thus affecting recognition accuracy.

[0006] Regarding state determination and control mechanisms, existing technologies generally employ a fixed threshold-based determination strategy, triggering feedback or control output when the attention index reaches a set level. This approach fails to adequately consider the temporal fluctuations of brain signals and does not constrain the persistence and stability of states, making the system susceptible to transient fluctuations and resulting in unstable responses. Furthermore, control logic driven by a single index lacks the ability to comprehensively judge multidimensional cognitive states, making it difficult to adapt to dynamic changes in complex training environments.

[0007] Regarding cognitive intervention mechanisms, existing systems mostly employ preset or manually adjusted task parameters, lacking the ability to adaptively adjust based on the user's real-time cognitive state. During training, task difficulty, stimulus intensity, and feedback rhythm are typically kept fixed or only subject to limited adjustments, failing to dynamically optimize intervention strategies based on changes in individual attention levels and cognitive load. This static or semi-static intervention approach easily leads to a mismatch between training load and individual capabilities, thus affecting training efficiency and effectiveness.

[0008] Current brain-computer interface systems remain imperfect in terms of closed-loop control. Most systems only achieve a unidirectional regulatory process driven by brain state feedback output, lacking a mechanism to incorporate behavioral performance, task completion, and system response results into the feedback loop. A large amount of behavioral data generated during cognitive training is not fully utilized, making it difficult for the system to optimize subsequent intervention strategies based on training effects, thus limiting the effectiveness of closed-loop regulation.

[0009] In terms of user engagement and system effectiveness assessment, existing technologies lack real-time monitoring mechanisms for brain signal quality, cognitive engagement, and brain-behavioral consistency. The system struggles to determine whether a user is in an effective training state and cannot promptly identify and address abnormal states, thus affecting the reliability and stability of the intervention process.

[0010] In summary, existing brain-computer interface (BCI)-based cognitive intervention technologies still have shortcomings in attention representation, state determination, adaptive adjustment, and closed-loop control. The core problem lies in the lack of multi-dimensional cognitive state fusion modeling and dynamic regulation mechanisms, making it difficult to achieve stable, precise, and individualized attention-enhancing cognitive closed-loop interventions. Therefore, it is necessary to propose a new technical solution to improve the application effect of BCIs in the field of cognitive rehabilitation. Summary of the Invention

[0011] The main objective of this invention is to provide an attention-enhancing cognitive closed-loop rehabilitation intervention method based on brain-computer interface, in order to address the shortcomings of related technologies.

[0012] To achieve the above objectives, according to a first aspect of the present invention, a brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method is provided, comprising: collecting electroencephalogram (EEG) signals and behavioral information during the execution of cognitive tasks; assigning timestamps to the EEG signals, behavioral information, and device operating status data, and segmenting the continuous data using a fixed-length sliding time window; and performing frequency domain analysis on the EEG signals within each sliding time window to calculate the α-band power of the EEG signals in the prefrontal cortex. and β band power And based on the two, determine the basic attention features. Within each sliding time window, determine the quality assessment indicators of the EEG signal; and assess the basic characteristics of attention. Normalization is performed to obtain the normalized attention intensity. And based on the normalized attention intensity Determine the average attention intensity over M consecutive time windows preceding the current moment. And based on the mean attention intensity Determine the attention stability index Feature extraction is performed on the collected behavioral information to obtain specified feature information; cognitive engagement index information is determined based on the specified feature information. The intensity of attention Attention stability index EEG signal quality evaluation indicators and cognitive engagement indicators Weighted fusion was performed to determine comprehensive brain state indicators. Based on comprehensive brain state indicators A gating mechanism with dynamic thresholds and multi-time window constraints is used to determine the state; control commands are generated based on the state determination results, and the operating parameters of external intervention equipment are adjusted based on the control commands.

[0013] According to a second aspect of the present invention, a brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention system is provided, comprising a preprocessing unit for collecting electroencephalogram (EEG) signals and behavioral information during the execution of cognitive tasks; assigning timestamps to the EEG signals, behavioral information, and device operating status data, and segmenting continuous data using a fixed-length sliding time window; and a processing unit for performing frequency domain analysis on the EEG signals within each sliding time window to calculate the α-band power of the EEG signals in the frontal region. and β band power And based on the two, determine the basic attention features. Within each sliding time window, determine the quality assessment indicators of the EEG signal; and assess the basic characteristics of attention. Normalization is performed to obtain the normalized attention intensity. And based on the normalized attention intensity Determine the average attention intensity over M consecutive time windows preceding the current moment. And based on the mean attention intensity Determine the attention stability index Feature extraction is performed on the collected behavioral information to obtain specified feature information; cognitive engagement index information is determined based on the specified feature information. The intensity of attention Attention stability index EEG signal quality evaluation indicators and cognitive engagement indicators Weighted fusion was performed to determine comprehensive brain state indicators. Intervention unit, used for comprehensive brain state indicators A gating mechanism with dynamic thresholds and multi-time window constraints is used to determine the state; control commands are generated based on the state determination results, and the operating parameters of external intervention equipment are adjusted based on the control commands.

[0014] This embodiment of the attention-enhancing cognitive closed-loop rehabilitation intervention method based on brain-computer interface includes collecting electroencephalogram (EEG) signals and behavioral information during the execution of cognitive tasks; assigning timestamps to the EEG signals, behavioral information, and device operating status data; and segmenting continuous data using a fixed-length sliding time window; within each sliding time window, performing frequency domain analysis on the EEG signals to calculate the α-band power of the EEG signals in the prefrontal cortex. and β band power And based on the two, determine the basic attention features. Within each sliding time window, determine the quality assessment indicators of the EEG signal; and assess the basic characteristics of attention. Normalization is performed to obtain the normalized attention intensity. And based on the normalized attention intensity Determine the average attention intensity over M consecutive time windows preceding the current moment. And based on the mean attention intensity Determine the attention stability index Feature extraction is performed on the collected behavioral information to obtain specified feature information; cognitive engagement index information is determined based on the specified feature information. The intensity of attention Attention stability index EEG signal quality evaluation indicators and cognitive engagement indicators Weighted fusion was performed to determine comprehensive brain state indicators. Based on comprehensive brain state indicators This study employs a gating mechanism with dynamic thresholds and multi-time windows for state determination. Control commands are generated based on the state determination results, and the operating parameters of external intervention devices are adjusted accordingly. First, EEG signals and related behavioral information are simultaneously acquired and time-aligned. Second, the acquired signals are preprocessed and feature extracted. Based on this, a multi-dimensional brain state fusion model is constructed to comprehensively assess attention and cognitive engagement. Furthermore, control commands are generated through dynamic threshold determination and gating decision-making mechanisms, and cognitive task parameters and intervention intensity are adaptively adjusted accordingly. System parameters are updated through a feedback mechanism, thus forming a closed-loop intervention process. This approach addresses the problems of single attention representation, unstable state determination, and lack of adaptive closed-loop regulation in existing brain-computer interface cognitive intervention methods, realizing an attention-enhanced cognitive closed-loop intervention mechanism based on multi-dimensional brain state fusion. Attached Figure Description

[0015] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0016] Figure 1 This is a flowchart of the attention-enhancing cognitive closed-loop rehabilitation intervention method based on brain-computer interface according to an embodiment of the present invention; Figures 2-4 This is a schematic diagram illustrating the application of the attention-enhanced cognitive closed-loop rehabilitation intervention method based on brain-computer interface according to an embodiment of the present invention; Figure 5 This is a schematic diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0017] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0018] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of the invention described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0019] It should be noted that, where there is no conflict, the various embodiments and features of this invention can be combined with each other. The invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] According to embodiments of the present invention, a brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method is provided, such as... Figure 1 As shown, steps 101 to 103 are included below: Step 101: Collect EEG signals and behavioral information during the execution of cognitive tasks; assign timestamps to the EEG signals, behavioral information and device operating status data, and use a fixed-length sliding time window to segment the continuous data.

[0021] In this step, the subject's electroencephalogram (EEG) signals are acquired using an EEG acquisition device. These signals include at least prefrontal cortex signals, reflecting cognitive activities related to attention. Simultaneously, EEG signals from the central or sensorimotor cortex are acquired to characterize neural processes related to task performance. The sampling frequency of the EEG signals is set to no less than [missing information]. This is to ensure effective differentiation of the characteristics of each frequency band.

[0022] While acquiring EEG signals, behavioral information generated during the cognitive task execution is recorded simultaneously. This behavioral information includes task response results, reaction time, operational process data, and error rate, used to characterize the subject's actual participation in the task. Furthermore, the operational status of external intervention devices is recorded, including device output parameters and their changes, for correlation analysis during subsequent closed-loop control.

[0023] To achieve time alignment between different types of signals, a unified timestamp is assigned to all types of acquired data, and continuous data is segmented based on a sliding time window of fixed length. Let the time window length be... Then at any time The corresponding data range is represented as ,in, Indicates the current time The time window is the endpoint, and the window contains a length of Continuous data segments. By continuously sliding along the time axis, the original continuous signal can be divided into multiple overlapping time windows, thus providing a unified data structure for subsequent feature extraction and state analysis.

[0024] As an optional implementation of this embodiment, frequency domain analysis is performed on the EEG signal within each sliding time window to calculate the α-band power of the EEG signal in the prefrontal region. and β band power And based on the two, determine the basic attention features. Within each sliding time window, determine the quality assessment indicators of the EEG signal; and assess the basic characteristics of attention. Normalization is performed to obtain the normalized attention intensity. And based on the normalized attention intensity Determine the average attention intensity over M consecutive time windows preceding the current moment. And based on the mean attention intensity Determine the attention stability index Feature extraction is performed on the collected behavioral information to obtain specified feature information; cognitive engagement index information is determined based on the specified feature information. The intensity of attention Attention stability index EEG signal quality evaluation indicators and cognitive engagement indicators Weighted fusion was performed to determine comprehensive brain state indicators. .

[0025] In this optional implementation, frequency domain analysis is performed on the EEG signal within each time window, and the energy characteristics of different frequency bands are calculated. Let the power of the prefrontal EEG signal in the α and β frequency bands be respectively... and Then, the fundamental feature quantity used to characterize the attention level can be defined as follows: ,in, Indicates within the time window The frequency band energy ratio obtained from internal calculation; The power in the alpha band reflects a relatively relaxed or low-attention state; The power in the β band reflects a relatively active or high-attention state. To prevent extremely small positive numbers with a denominator of zero, this ratio is used to initially characterize the subject's attention intensity level within the corresponding time window. It should be noted that the above feature is only one of the basic inputs for subsequent multidimensional brain state fusion modeling and is not directly used for control decisions. Instead, it participates in the comprehensive evaluation along with behavioral information and other features in subsequent processing. This achieves the synchronous expression of EEG signals, behavioral information, and device status data within a unified time frame, providing a data foundation for subsequent processing.

[0026] After completing multi-source signal acquisition and time alignment, preprocessing and feature extraction were performed on the EEG signals and behavioral data to improve signal quality and construct stable feature representations suitable for subsequent analysis. First, bandpass filtering was applied to the EEG signals to keep them within a preset frequency range, thereby removing low-frequency drift and high-frequency noise interference. Simultaneously, artifact removal methods were used to suppress abnormal fluctuations caused by blinking, electromyographic activity, or environmental interference, enhancing signal effectiveness. Based on this, the processed EEG signals were normalized to ensure comparability of features across different time windows and individuals.

[0027] To further evaluate the reliability of EEG signals within the current time window, in each time window We will construct a built-in signal quality evaluation index. Let the filtered EEG signal be... Its mean is The standard deviation is Then the normalized fluctuation level can be defined as ,in, Indicates the first time within the time window The signal value at each sampling point This represents the number of sampling points within the window. To prevent extremely small positive numbers with a denominator of zero. This index reflects the degree of fluctuation of the signal relative to its statistical characteristics. An excessively large value indicates potential abnormal interference within the current window. Based on this metric, a signal quality factor can be further defined. This is used to measure the reliability of EEG signals within that time window.

[0028] Furthermore, in obtaining basic attentional features Based on this, to eliminate the differences between individuals and the influence of signal amplitude variations, normalization is performed. Let the minimum and maximum values ​​of the attention features obtained in the initial stage be respectively... and Then the normalized attention intensity can be expressed as ,in, This represents the normalized attention intensity, with its value ranged from 0 to 1, used to unify the attention representation across different individuals and time periods.

[0029] Furthermore, the attention stability index is defined as... ,in, It is used to reflect the degree of fluctuation in attention over a period of time. When its value is high, it indicates that the attention state is relatively stable; when its value is low, it indicates that there is a large fluctuation in attention.

[0030] Simultaneously, feature extraction was performed on the behavioral data collected during the cognitive task execution, including indicators such as task completion status, error rate, and reaction time. These features were then normalized to ensure they were within the same metric as the EEG characteristics. These behavioral features are used to characterize the subject's participation and performance in the task, providing supplementary information for subsequent comprehensive cognitive state assessment.

[0031] After obtaining attention intensity, stability characteristics, signal quality factors, and behavioral features, the above multidimensional information is fused and modeled to achieve a comprehensive assessment of the subject's cognitive state. In this process, features from different sources are uniformly mapped to the same evaluation scale, and a comprehensive brain state index is constructed through weighted fusion.

[0032] Specifically, the normalized attention intensity Attention stability Signal quality factor And cognitive engagement metrics extracted from behavioral data Joint modeling was performed. This included cognitive engagement. This is used to characterize a subject's actual engagement in a task, and can be constructed based on behavioral characteristics such as task completion rate, error rate, and reaction time. For example, the task completion rate can be denoted as... Error rate is denoted as Normalized reaction time is denoted as Cognitive engagement can then be expressed as ,in, , , Let be the weighting coefficient, satisfying This is used to adjust the contribution ratio of each behavioral characteristic in the participation assessment.

[0033] Based on this, the aforementioned multidimensional features are weighted and fused to construct a comprehensive brain state index. ,in, Indicates within the time window Comprehensive brain state score within the body; , , , The corresponding weight coefficients are such that the sum of all weights is 1, which is used to reflect the importance of different features in the overall state assessment.

[0034] This comprehensive indicator This comprehensive approach, considering attention level, attention duration, task engagement, and signal reliability, provides a more complete picture of the subject's true cognitive state at any given moment compared to a single EEG feature. When attention intensity is high but stability is poor or signal quality is low, the overall index decreases accordingly, preventing misjudgments due to transient fluctuations or abnormal signals. Conversely, when behavioral performance aligns with brain state, the overall index increases, enhancing the response to effective cognitive engagement. Furthermore, dynamic thresholds and gating mechanisms can be introduced to hierarchically assess the overall brain state, achieving stable and reliable closed-loop control.

[0035] Step 103: Based on comprehensive brain state indicators A gating mechanism with dynamic thresholds and multi-time window constraints is used to determine the state; control commands are generated based on the state determination results, and the operating parameters of external intervention equipment are adjusted based on the control commands.

[0036] In obtaining comprehensive brain state indicators Subsequently, dynamic judgment and control decisions are made to achieve stable and reliable intervention outputs. To overcome the sensitivity of fixed threshold methods to instantaneous fluctuations, dynamic thresholds and multi-time window constraints are introduced in this process to classify brain states hierarchically.

[0037] refer to Figure 2 A schematic diagram illustrating the closed-loop cognitive regulation mechanism of a brain-computer interface, and references. Figure 3 This diagram illustrates the closed-loop interactive process of cognitive training. Based on data from the subjects' baseline state and the initial training phase, the distribution characteristics of comprehensive brain state indicators are statistically analyzed to obtain the initial threshold. In subsequent training, the threshold is dynamically adjusted based on fluctuations in the overall brain state within the recent time window. Let's assume that the continuous [state of mind] before the current moment... The standard deviation of the comprehensive brain state index within each time window is: At the same time, training load or fatigue level indicators are introduced. Then the dynamic threshold can be expressed as ,in, This is the dynamic threshold for the current moment; and The threshold is set as an adjustment factor, used to control the effects of volatility and fatigue level on the threshold, respectively. When the overall brain state fluctuates significantly, the judgment threshold is increased to enhance the robustness of the system; when the subject experiences fatigue or a decline in state, the threshold is appropriately decreased to maintain the continuity of training.

[0038] After the threshold is determined, gating is applied to the comprehensive brain state index. To avoid false triggering due to transient anomalies, a continuous window constraint is introduced in the time dimension, with the minimum number of continuously satisfied windows set as follows: Then the gating decision function can be expressed as in, This represents the gated output at time t. If the overall brain state meets the threshold condition for L consecutive time windows, it is considered a valid state, and the output is 1; otherwise, the output is 0. This mechanism effectively filters short-term fluctuations and improves the stability of control decisions.

[0039] After completing the gating decision, a control command is generated based on the decision result, and the comprehensive brain state index is mapped to the intervention intensity or control parameter. Let the control output be... Then it can be represented as ,in, , Here, b is the proportionality constant, and b is the bias term. This indicates the difficulty level of the current cognitive task. This expression shows that the system only outputs a valid control signal when the gating conditions are met; furthermore, the control output is not only related to the current brain state but also coupled with the task difficulty, thus achieving more reasonable intervention and regulation.

[0040] Obtaining control output Subsequently, based on the current comprehensive brain state and gating judgment results, the cognitive task parameters and intervention intensity are adaptively adjusted to achieve an individualized cognitive intervention process. In this process, the brain state assessment results are directly applied to the task structure and feedback mechanism, so that the training process is dynamically adjusted according to the subject's real-time state to maintain an appropriate level of cognitive load.

[0041] As an optional implementation method in this embodiment, based on comprehensive brain state indicators Compared with the preset reference level The difference between them updates the difficulty of the cognitive task, where the current task difficulty is set to 1. Then the difficulty of the task in the next moment is expressed as , and These are adjustment coefficients, used to control the effects of brain state bias and error rate on changes in task difficulty, respectively. The task error rate within the current time window; adjustments to specified parameters in the cognitive task based on task difficulty; and control output according to control instructions. Adjust the operating parameters of external intervention equipment.

[0042] In this optional implementation, based on comprehensive brain state indicators Compared with the preset reference level The difference between them updates the difficulty of the cognitive task. Let the current task difficulty be... Then the task difficulty in the next moment can be expressed as ,in, and These are adjustment coefficients, used to control the effects of brain state bias and error rate on changes in task difficulty, respectively. This represents the task error rate within the current time window. This update method ensures that the task difficulty gradually increases when the subject's attention level is high and performance is good, and decreases accordingly when the error rate rises or attention declines, thus achieving a dynamic balance.

[0043] Once the task difficulty is determined, specific parameters in the cognitive task are adjusted, including the task presentation pace, number of targets, intensity of interference, and feedback methods, to match the current difficulty level. Simultaneously, based on the control output... Adjust the operating parameters of external intervention devices, such as the intensity of stimulation, the speed of execution, or the degree of assistance, to make them consistent with the cognitive state of the subject.

[0044] As an optional implementation in this embodiment, during the adjustment of operating parameters, the range of changes in task difficulty and intervention intensity is limited based on a change constraint mechanism. The intervention intensity is determined by a comprehensive brain state index. The mapping is obtained; let the maximum allowable range of change be... Then the task difficulty update satisfies .

[0045] In this optional implementation, to ensure the continuity and stability of the adjustment process, a change constraint mechanism is introduced during parameter updates to limit the range of changes in task difficulty and intervention intensity, preventing drastic parameter changes due to state fluctuations. Let the maximum allowable range of change be... Then the task difficulty update satisfies This constraint makes the task adjustment process smoother, improves user experience, and enhances system stability. After completing the adaptive adjustment of cognitive tasks and intervention parameters, a closed-loop feedback and parameter update are performed throughout the intervention process to achieve continuous optimization and individual adaptation of the system during multiple training sessions. In this process, not only are responses made to the state within a single time window, but also the trends of state changes across time scales are analyzed, thereby improving the long-term stability and intervention effect of the system.

[0046] As an optional implementation of this embodiment, after completing an adjustment cycle, the comprehensive brain state index C(t), normalized attention intensity A(t), attention stability index S(t), cognitive engagement E(t), and EEG signal quality evaluation index information of each time window within that cycle are statistically analyzed. Based on the statistical results, the normalized parameters of the basic attention features are corrected using a weighted update method, and the weight coefficients and initial thresholds of the dynamic thresholds corresponding to the comprehensive brain state index C(t), normalized attention intensity A(t), attention stability index S(t), cognitive engagement E(t), and EEG signal quality evaluation index information are updated. The updated parameters are then applied to the next cycle.

[0047] In this optional implementation, after each training cycle, the comprehensive brain state index for each time window within that cycle is calculated. Attention intensity Stability indicators Cognitive participation and signal quality factor Perform statistical analysis to calculate its mean, variance, and trend. Assume there are a total of [number] values ​​within one training period. If there are multiple time windows, then the average level of overall brain state can be expressed as: ,in, It reflects the overall cognitive state level within that period. This is achieved by analyzing the levels of cognitive function across different training periods. By comparing the changes, we can obtain the trend of changes in the subject's cognitive ability during the training process.

[0048] Based on this, the initial normalization parameters are updated. Let the minimum and maximum values ​​of the attention features observed in the current period be respectively... and Then, a weighted update method can be used to correct the original parameters. , ,in, This is a smoothing coefficient used to control the weighting distribution between historical parameters and current observations. This allows the normalization interval to gradually adapt to changes in the individual during training.

[0049] Simultaneously, the fusion weight parameters are adjusted based on the performance of each feature in the current cycle. Let the comprehensive brain state index be... The correlations between the features of each component are as follows: Then, the weight coefficients can be normalized and updated according to the correlation magnitude, thereby enhancing the responsiveness to key features.

[0050] In addition, for dynamic threshold parameters Update it to adapt to changes in individual states. This can be done based on the distribution of overall brain states within the current cycle. The threshold is adjusted to a certain quantile value of the distribution to ensure that the threshold is reasonable at different stages.

[0051] After updating the parameters as described above, the updated parameters are applied to the next training cycle, enabling the system to more accurately reflect the individual's state and respond accordingly during subsequent interventions. By continuously iterating this process, the system gradually converges from the initial settings to the individualized parameters.

[0052] As an optional implementation method in this embodiment, comprehensive brain state indicators are continuously monitored during real-time intervention. EEG signal quality evaluation indicators and cognitive engagement indicators When the EEG signal quality evaluation index is lower than the preset threshold, or the rate of change of the comprehensive brain state index exceeds the preset threshold, or the cognitive participation index E(t) is significantly lower than the expected level, it is judged as an abnormal state, and the control output is suspended or the intervention intensity is reduced until the abnormality is resolved.

[0053] In this optional implementation, based on the aforementioned closed-loop feedback and parameter updates, an abnormal state detection and protection mechanism is introduced into the intervention process to further improve the stability and security of the system in practical applications. When abnormal fluctuations occur in the comprehensive brain state indicators, signal quality, or behavioral performance, the system can promptly identify and limit or adjust the control output to avoid misjudgment or over-intervention.

[0054] For example, during real-time operation, the comprehensive brain state index is analyzed. Signal quality factor Continuous monitoring of behavioral characteristics is performed. An abnormal state is determined when any indicator exceeds a preset range. Let the signal quality threshold be... Then when If significant interference is detected in the current EEG signal, the control output is paused or the intervention intensity is reduced. Simultaneously, the rate of change of the comprehensive brain state indicators is constrained, assuming its rate of change is... ,when Exceeding the preset threshold If the state changes too drastically, the system will smooth out the control output or temporarily suspend updates to prevent control instability caused by instantaneous fluctuations.

[0055] In addition, consistency testing was performed on behavioral performance. (When cognitive engagement metrics...) When the level is significantly lower than expected, it indicates that the subject did not effectively participate in the task; even if the EEG characteristics meet certain conditions, they do not directly trigger effective control output. This mechanism achieves joint constraints between brain signals and behavioral performance, thereby improving the system's ability to determine the true cognitive state.

[0056] After the abnormal state is handled, the system re-enters the normal closed-loop control process and continues to perform brain state assessment, control decision-making, and intervention adjustment. By introducing the above-mentioned abnormality detection and protection mechanisms, the entire cognitive closed-loop intervention system can maintain stable operation in complex environments, reduce the risk of false triggering, and improve the safety and reliability of the intervention process.

[0057] This embodiment's method reference Figure 4 The cognitive closed-loop intervention technology roadmap shown constructs a multi-dimensional brain state fusion and closed-loop control system for cognitive intervention. In terms of brain state modeling, it breaks through the single EEG feature representation method by fusing multi-dimensional information such as attention intensity, attention stability, cognitive engagement, and signal quality to achieve a unified description of cognitive states, improving recognition accuracy and stability. Regarding control strategies, a hierarchical gating mechanism based on dynamic thresholds and time window constraints is introduced to perform multi-level judgments of brain states, effectively reducing the impact of instantaneous fluctuations on system control. In terms of intervention regulation, the difficulty of cognitive tasks and intervention parameters are dynamically adjusted according to real-time brain states, ensuring that the training process remains within an appropriate cognitive load range. Regarding the closed-loop structure, brain signal features and behavioral performance results are jointly incorporated into the feedback loop to achieve continuous optimization and adaptive adjustment of the intervention process.

[0058] This embodiment provides a brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method. By collaboratively collecting and processing EEG signals and task behavior information, a multidimensional brain state fusion model is constructed to comprehensively evaluate an individual's attention level, cognitive engagement, and signal effectiveness. Based on this evaluation result, cognitive tasks and external intervention parameters are dynamically adjusted, thus forming a continuous and adaptive cognitive closed-loop intervention process. Its core technology lies in using multidimensional brain state fusion as a foundation, combined with dynamic threshold determination and hierarchical gating control strategies, to achieve closed-loop coupling control between brain state recognition, state evaluation, and intervention adjustment, thereby improving the stability, accuracy, and individual adaptability of the cognitive intervention process. Compared to existing cognitive training methods based on single EEG features or fixed threshold control, this invention, by introducing multidimensional brain state fusion and an adaptive closed-loop regulation mechanism, achieves comprehensive modeling and dynamic adjustment of cognitive states. The technical approach includes: first, synchronously collecting EEG signals and related behavioral information and completing time alignment processing; second, preprocessing and extracting features from the collected signals; third, constructing a multidimensional brain state fusion model to achieve a comprehensive assessment of attention and cognitive participation; fourth, generating control commands through dynamic threshold determination and gating decision-making mechanisms, and adaptively adjusting cognitive task parameters and intervention intensity accordingly; and finally, updating system parameters through a feedback mechanism to form a closed-loop intervention process.

[0059] The key innovation lies in constructing a multidimensional brain state fusion and closed-loop control system for cognitive intervention. In terms of brain state modeling, it breaks through the limitations of single EEG feature representation methods. By integrating multidimensional information such as attention intensity, attention stability, cognitive engagement, and signal quality, it achieves a unified description of cognitive states, improving recognition accuracy and stability. Regarding control strategies, a hierarchical gating mechanism based on dynamic thresholds and time window constraints is introduced to make multi-level judgments on brain states, effectively reducing the impact of instantaneous fluctuations on system control. In terms of intervention regulation, the difficulty of cognitive tasks and intervention parameters are dynamically adjusted according to real-time brain states, ensuring that the training process remains within an appropriate cognitive load range. Regarding the closed-loop structure, brain signal features and behavioral performance results are jointly incorporated into the feedback loop, achieving continuous optimization and adaptive adjustment of the intervention process. In summary, this forms a cognitive closed-loop intervention method with state perception, decision-making regulation, and adaptive optimization capabilities.

[0060] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0061] According to an embodiment of the present invention, an attention-enhancing cognitive closed-loop rehabilitation intervention system based on a brain-computer interface is also provided, comprising a preprocessing unit for collecting electroencephalogram (EEG) signals and behavioral information during the execution of cognitive tasks; assigning timestamps to the EEG signals, behavioral information, and device operating status data, and segmenting continuous data using a fixed-length sliding time window; and a processing unit for performing frequency domain analysis on the EEG signals within each sliding time window to calculate the α-band power of the EEG signals in the frontal region. and β band power And based on the two, determine the basic attention features. Within each sliding time window, determine the quality assessment indicators of the EEG signal; and assess the basic characteristics of attention. Normalization is performed to obtain the normalized attention intensity. And based on the normalized attention intensity Determine the average attention intensity over M consecutive time windows preceding the current moment. And based on the mean attention intensity Determine the attention stability index Feature extraction is performed on the collected behavioral information to obtain specified feature information; cognitive engagement index information is determined based on the specified feature information. The intensity of attention Attention stability index EEG signal quality evaluation indicators and cognitive engagement indicators Weighted fusion was performed to determine comprehensive brain state indicators. Intervention unit, used for comprehensive brain state indicators A gating mechanism with dynamic thresholds and multi-time window constraints is used to determine the state; control commands are generated based on the state determination results, and the operating parameters of external intervention equipment are adjusted based on the control commands.

[0062] According to embodiments of the present invention, the present invention also provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to implement the method described in any of the above embodiments.

[0063] According to embodiments of the present invention, the present invention also provides a readable storage medium storing computer instructions that enable a computer to perform the methods described in any of the above embodiments when executed.

[0064] According to embodiments of the present invention, the present invention also provides a computer program product that, when executed by a processor, can implement the methods described in any of the above embodiments.

[0065] Figure 5 A schematic block diagram of an example electronic device 300 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices.

[0066] like Figure 5 As shown, the electronic device 300 includes a computing unit 301, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 302 or a computer program loaded from a storage unit 308 into a random access memory (RAM) 303. The RAM 303 may also store various programs and data required for the operation of the electronic device 300. The computing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0067] Multiple components in electronic device 300 are connected to I / O interface 305, including: input unit 306, such as keyboard, mouse, etc.; output unit 307, such as various types of displays, speakers, etc.; storage unit 308, such as disk, optical disk, etc.; and communication unit 309, such as network card, modem, wireless transceiver, etc. Communication unit 309 allows electronic device 300 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0068] The computing unit 301 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 301 performs the various methods and processes described above, such as the object matching method. For example, in some embodiments, the object matching method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 308. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 300 via ROM 302 and / or communication unit 309. When the computer program is loaded into RAM 303 and executed by the computing unit 301, one or more steps of the methods described above may be performed.

[0069] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0070] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0071] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

Claims

1. A brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method, characterized in that, include: Collecting EEG signals and behavioral information during cognitive task execution; The EEG signals, behavioral information, and device operating status data are assigned timestamps, and continuous data are segmented using a fixed-length sliding time window. Within each sliding time window, frequency domain analysis was performed on the EEG signals to calculate the alpha band power of the EEG signals in the prefrontal cortex. and β band power And based on the two, determine the basic attention features. Within each sliding time window, determine the quality assessment indicators of the EEG signal; and assess the basic characteristics of attention. Normalization is performed to obtain the normalized attention intensity. And based on the normalized attention intensity Determine the average attention intensity over M consecutive time windows preceding the current moment. And based on the mean attention intensity Determine the attention stability index ; Extract features from the collected behavioral information to obtain the specified feature information; Cognitive engagement index information is determined based on the specified feature information. The intensity of attention Attention stability index EEG signal quality evaluation indicators and cognitive engagement indicators Weighted fusion was performed to determine comprehensive brain state indicators. ; Based on comprehensive brain state indicators A gating mechanism with dynamic thresholds and multi-time window constraints is used to determine the state; control commands are generated based on the state determination results, and the operating parameters of external intervention equipment are adjusted based on the control commands.

2. The brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method according to claim 1, characterized in that, Before adjusting the operating parameters of external intervention equipment based on control commands, the methods include: Based on comprehensive brain state indicators Compared with the preset reference level The difference between them updates the difficulty of the cognitive task, where the current task difficulty is set to 1. Then the difficulty of the task in the next moment is expressed as , and These are adjustment coefficients, used to control the effects of brain state bias and error rate on changes in task difficulty, respectively. This represents the task error rate within the current time window. Adjust specified parameters in the cognitive task based on the task difficulty; and simultaneously adjust the control output according to the control instructions. Adjust the operating parameters of external intervention equipment.

3. The brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method according to claim 2, characterized in that, During the adjustment of operational parameters, a change constraint mechanism was used to limit the range of changes in task difficulty and intervention intensity. The intervention intensity was determined by a comprehensive brain state index. The mapping is obtained; let the maximum allowable range of change be... Then the task difficulty update satisfies .

4. The brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method according to claim 3, characterized in that, After completing a conditioning cycle, the comprehensive brain state index C(t), normalized attention intensity A(t), attention stability index S(t), cognitive engagement E(t), and EEG signal quality evaluation index information of each time window within the cycle are statistically analyzed. Based on the statistical results, a weighted update method is used to correct the normalization parameters of the basic attention features. and Update the weighting coefficients corresponding to the comprehensive brain state index C(t), normalized attention intensity A(t), attention stability index S(t), cognitive engagement E(t), and EEG signal quality evaluation index. 、 、 、 and the initial threshold of the dynamic threshold The updated parameters will then be applied to the next cycle.

5. The brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method according to claim 4, characterized in that, During real-time intervention, comprehensive brain state indicators are continuously monitored. EEG signal quality evaluation indicators and cognitive engagement indicators ; When the EEG signal quality evaluation index is lower than the preset threshold, or the rate of change of the comprehensive brain state index exceeds the preset threshold, or the cognitive participation index E(t) is significantly lower than the expected level, it is judged as an abnormal state, and the control output is suspended or the intervention intensity is reduced until the abnormality is resolved.

6. The brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method according to claim 1, characterized in that, Based on comprehensive brain state indicators The state determination mechanism employs dynamic thresholds and multi-time-window constraints, including: Set initial threshold Combined with the standard deviation of the comprehensive brain state index over the M consecutive time windows preceding the current moment and fatigue level indicators Dynamic adjustment Introducing the minimum number of consecutively satisfied windows L, a gating decision function is constructed. Based on the gating decision function, the comprehensive brain state index is judged. The validity of the status is determined; Generating control commands based on state determination results includes using a gating determination function. Generate control output .

7. The brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method according to claim 4, characterized in that, The normalization parameters of the attention base features are corrected using a weighted update method. and include: Let the minimum and maximum values ​​of the attention features observed in the current period be respectively... and Then, a weighted update method is used to correct the original parameters: in, This is the smoothing coefficient.

8. The brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method according to claim 1, characterized in that, Within each sliding time window, the information for determining the quality assessment indicators of EEG signals includes: Determining the normalized fluctuation level based on EEG signals within a sliding time window ; Based on the above Determine the signal quality factor , wherein Used to measure the reliability of EEG signals within a time window.

9. The brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention method according to claim 1, characterized in that, Cognitive engagement index information is determined based on the specified feature information. This includes: recording the task completion rate as... Error rate is denoted as Normalized reaction time is denoted as Cognitive engagement is then expressed as ,in, , , Let be the weighting coefficient, satisfying This is used to adjust the contribution ratio of each behavioral characteristic in the participation assessment.

10. A brain-computer interface-based attention-enhancing cognitive closed-loop rehabilitation intervention system, characterized in that, include: The preprocessing unit is used to collect EEG signals and behavioral information during the execution of cognitive tasks; The EEG signals, behavioral information, and device operating status data are assigned timestamps, and continuous data are segmented using a fixed-length sliding time window. The processing unit is used to perform frequency domain analysis on the EEG signal within each sliding time window to calculate the alpha band power of the EEG signal in the frontal region. and β band power And based on the two, determine the basic attention features. Within each sliding time window, determine the quality assessment indicators of the EEG signal; and assess the basic characteristics of attention. Normalization is performed to obtain the normalized attention intensity. And based on the normalized attention intensity Determine the average attention intensity over M consecutive time windows preceding the current moment. And based on the mean attention intensity Determine the attention stability index ; Extract features from the collected behavioral information to obtain the specified feature information; Cognitive engagement index information is determined based on the specified feature information. The intensity of attention Attention stability index EEG signal quality evaluation indicators and cognitive engagement indicators Weighted fusion was performed to determine comprehensive brain state indicators. ; Intervention unit, used for comprehensive brain state indicators A gating mechanism with dynamic thresholds and multi-time window constraints is used to determine the state; control commands are generated based on the state determination results, and the operating parameters of external intervention equipment are adjusted based on the control commands.

Images