A cognitive impairment screening method and system based on resting-state electroencephalogram and multi-task paradigm

Abstract

The application discloses a cognitive disorder screening method and system based on resting state electroencephalogram and multi-task paradigm, wherein the method comprises the following steps: collecting baseline closed-eye resting state electroencephalogram signals of a subject under a task-free stimulation condition to establish an individualized neural steady state reference; presenting a multi-dimensional cognitive task covering multiple cognitive function networks; collecting closed-eye resting state electroencephalogram signals in a preset time threshold after the task ends to represent the steady state regression process after neural disturbance. Based on the baseline and recovery period electroencephalogram signals, the neural recovery amplitude feature, the neural recovery rate feature and the brain network reconstruction feature are calculated, the neural recovery feature set is mapped to the healthy neural recovery norm space, the deviation degree of the neural steady state regulation ability is evaluated, and the cognitive disorder risk assessment result is output. The application overcomes the problem of insufficient sensitivity of the traditional resting state or task state single analysis mode to early cognitive disorders, and is suitable for early screening and clinical auxiliary evaluation of cognitive disorders.

Description

Technical Field

[0001] This invention relates to the fields of biomedical signal processing and cognitive neuroscience, and in particular to a method and system for screening cognitive impairment based on resting-state electroencephalography and a multi-task paradigm. Background Technology

[0002] Alzheimer's disease (AD) and its early mild cognitive impairment (MCI) have become a major global public health challenge. Achieving early and accurate screening for cognitive impairment is of vital clinical significance for slowing disease progression and developing intervention strategies.

[0003] Currently, clinical screening for cognitive impairment mainly relies on neuropsychological scales, such as the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA). However, these methods are primarily question-and-answer based, which can lead to subjectivity and time consumption. More importantly, scale scores are easily influenced by the subject's education level, cultural background, and language ability, and have low sensitivity to early, subtle changes, making it difficult to detect latent cognitive impairments during the compensatory phase.

[0004] With the development of neuroscience, electroencephalography (EEG) signals have been increasingly applied to cognitive function assessment due to their high temporal resolution and low cost. Traditional EEG examination modalities are mainly divided into two categories: resting-state analysis, which focuses on static indicators such as increased slow waves and decreased alpha wave frequency; and task-based analysis, which mainly observes event-related potentials (ERPs) under specific stimuli. However, existing single-mode detection has significant limitations: simple resting-state analysis lacks a dynamic assessment of the brain's information processing capacity; while simple task-based analysis often only focuses on the degree of activation at the moment of task execution, ignoring the brain's recovery process after coping with cognitive load.

[0005] In fact, early cognitive impairment often manifests as a weakening of the nervous system's self-regulation ability. That is, subjects may manage to complete cognitive tasks by mobilizing more neural resources (resulting in normal or only slightly abnormal behavioral data), but after the task, their neural activity struggles to quickly return to baseline homeostasis, exhibiting a phenomenon of "recovery lag" or "network reconstruction failure." Existing screening technologies struggle to capture this dynamic process, leading to a large number of patients in the early stages of MCI being missed.

[0006] Therefore, there is an urgent need for a cognitive impairment screening method that can break through the limitations of traditional static or single-task analysis and quantify the dynamic changes in the nervous system's ability to adapt throughout the "steady-state-perturbation-recovery" process. Summary of the Invention

[0007] This invention provides a cognitive impairment screening method and system based on resting-state EEG and a multi-task paradigm. It uses the brain's ability to recover from neural homeostasis as the core assessment indicator and constructs a strong time-constrained detection process of "baseline homeostasis - neural disturbance - recovery homeostasis" to achieve highly sensitive screening for mild cognitive impairment.

[0008] This invention provides a cognitive impairment screening method based on resting-state EEG and a multi-task paradigm, comprising:

[0009] S1. Acquire the subject's eye-closed resting-state EEG signal for a first preset duration under conditions without cognitive task stimulation, and process the eye-closed resting-state EEG signal to construct a baseline EEG feature set to characterize the subject's initial neural homeostasis.

[0010] S2. Present a multi-dimensional cognitive task paradigm to the subject in a preset order. During the subject's execution of the multi-dimensional cognitive task paradigm, record behavioral data simultaneously to confirm the task completion rate. The multi-dimensional cognitive task paradigm is used to apply neural load perturbation to different cognitive networks in the brain.

[0011] S3. Within a preset time threshold after the end of the multidimensional cognitive task paradigm, collect the subject's eye-closed resting-state EEG signal for a second preset duration during the recovery period, and process the eye-closed resting-state EEG signal during the recovery period to construct a recovery period EEG feature set for characterizing the steady-state recovery process after neural disturbance.

[0012] S4. Based on the baseline EEG feature set and the recovery period EEG feature set, a comparative analysis is performed to calculate the neural recovery feature set; wherein, the neural recovery feature set includes neural recovery amplitude features, neural recovery rate features, and brain network remodeling features;

[0013] S5. Map the neural recovery feature set to a pre-constructed healthy neural recovery norm space, calculate the degree of deviation of the neural recovery feature set from the healthy neural recovery norm, and output the screening assessment results used to characterize the subject's cognitive impairment risk based on the degree of deviation.

[0014] Furthermore, S1 specifically includes:

[0015] S101. Using a portable EEG acquisition device, collect the EEG signal of the subject in a resting state with eyes closed and without engaging in specific thinking activities for the first preset duration.

[0016] S102. Perform artifact removal processing on the closed-eye resting-state EEG signal, and decompose the artifact-removed signal into Delta band, Theta band, Alpha band and Beta band.

[0017] S103. Based on the decomposed frequency band signals, calculate the baseline relative power spectral density, the baseline alpha wave steady-state reference value, and the baseline brain functional connectivity network model benchmark value to construct the baseline EEG feature set.

[0018] Further, in S103, the formula for calculating the baseline relative power spectral density is:

[0019] ;

[0020] in, This represents the baseline average relative power spectral density of channel c within frequency band b; The power spectrum of channel c; and These are the lower and upper frequency limits for frequency band b, respectively. and These are the lower and upper frequency limits across the entire frequency band used in calculating relative power, respectively.

[0021] The formula for calculating the steady-state reference value of the baseline Alpha wave is as follows:

[0022] ;

[0023] in, This represents the steady-state reference value for the Alpha band during the baseline resting period. Indicates the duration of the baseline resting period; This represents the global average power in the Alpha band at time t during the baseline resting period;

[0024] The calculation formula for the baseline brain functional connectivity network model is as follows:

[0025] ;

[0026] in, This represents the baseline average brain functional connectivity matrix; Indicates the number of sliding windows during the baseline period; This represents the brain functional connectivity matrix within the k-th sliding time window during the baseline period.

[0027] Furthermore, S2 specifically includes:

[0028] S201. After the baseline EEG acquisition is completed, the control interaction device is switched to task mode, and the multidimensional cognitive task paradigm is presented to the subject in a preset order to apply neural load to the brain's memory network, attention network and spatial processing network; wherein, the multidimensional cognitive task paradigm includes episodic memory task, working memory task and visuospatial memory task.

[0029] S202. During the subject's execution of the multidimensional cognitive task paradigm, the task-state EEG signals are continuously recorded, and the subject's behavioral data are recorded simultaneously.

[0030] Furthermore, S3 specifically includes:

[0031] S301. Immediately terminate the cognitive task stimulation when the multidimensional cognitive task paradigm ends, and switch to the EEG resting acquisition mode within a preset time threshold.

[0032] S302. After the subject enters a closed-eye resting state, collect the closed-eye resting state EEG signal during the recovery period of the second preset duration.

[0033] S303. Extract features from the resting-state EEG signal with eyes closed during the recovery period to construct a recovery period EEG feature set including the recovery period Alpha band power time series, the recovery period average brain functional connectivity network matrix, and the recovery period average relative power spectral density; wherein, the recovery period average relative power spectral density is the average relative power spectral density of each channel in the preset frequency band during the entire recovery period, which is used to perform differential calculation with the baseline relative power spectral density in the baseline EEG feature set.

[0034] Furthermore, in S303, the calculation formula for the Alpha band power time series during the recovery period is as follows:

[0035] ;

[0036] in, The global alpha band average power at time t during the recovery period; C represents the selected set of EEG channels related to the central-parietal lobe or default network. This represents the number of channels in set C; This represents the relative power spectral density of channel c in the Alpha band at time t; t is the time variable during the recovery period, and ;

[0037] The formula for calculating the average brain functional connectivity matrix during the recovery period is as follows:

[0038] ;

[0039] in, This represents the average brain functional connectivity matrix during the recovery period; This indicates the total number of sliding windows during the recovery period; This represents the brain functional connectivity matrix within the k-th sliding time window of the recovery period. This matrix is ​​calculated from the correlation between signals from different EEG channels within the k-th sliding time window.

[0040] Furthermore, S4 specifically includes:

[0041] S401. Based on the relative power spectral density difference between the baseline EEG feature set and the recovery period EEG feature set, calculate the neural recovery amplitude characteristics, and calculate the deviation of the neural recovery amplitude characteristics from the healthy norm. The calculation formula is as follows:

[0042] ;

[0043] ;

[0044] in, This indicates the characteristics of the neural recovery amplitude in frequency band b; This represents the average relative power spectral density within frequency band b during the recovery period; This represents the average relative power spectral density within frequency band b during the baseline period; This indicates the degree of neural recovery deviation in frequency band b; This represents the mean recovery index of the healthy subject norm in frequency band b; This represents the standard deviation of the recovery index of healthy subjects in frequency band b;

[0045] S402. Based on the Alpha band power time series in the recovery period EEG feature set and the baseline Alpha steady-state reference value in the baseline EEG feature set, calculate the Alpha rebound slope as a characteristic of neural recovery rate. The calculation formula is as follows:

[0046] ;

[0047] Where ARS represents the Alpha bounce slope; This represents the i-th sampling time point within the recovery period; This represents the mean value at the sampling time points; Indicates a point in time The corresponding Alpha power value; This represents the steady-state reference value for the Alpha band during the baseline period;

[0048] S403. Based on the correlation between the brain functional connectivity network matrix of the baseline EEG feature set and the recovery period EEG feature set, calculate the brain network reconstruction features using the following formula:

[0049] ;

[0050] in, Indicates brain network remodeling characteristics; This represents the baseline average brain functional connectivity matrix; This represents the average brain functional connectivity matrix during the recovery period; This represents the operation of expanding a matrix into a vector in a preset order; This represents the Spearman rank correlation coefficient calculation function;

[0051] S404. Based on the deviation of the neural recovery amplitude characteristics, the neural recovery rate characteristics, and the brain network remodeling characteristics, a comprehensive neural recovery score is calculated using preset weights. The calculation formula is as follows:

[0052] ;

[0053] NRS stands for Comprehensive Neurological Recovery Score; Indicates the degree of deviation in the extent of nerve recovery; This represents the normalized Alpha bounce slope; Indicates brain network remodeling characteristics; These are preset weighting coefficients.

[0054] Furthermore, S5 specifically includes:

[0055] S501. Map the neural recovery feature set and comprehensive neural recovery score to a pre-constructed healthy neural recovery norm space, and calculate its deviation from the healthy norm; wherein, the healthy neural recovery norm space is a multi-dimensional feature distribution interval constructed based on large-scale cognitively normal population data, including confidence intervals for neural recovery amplitude, neural recovery rate and brain network reconstruction features;

[0056] S502. Based on the deviation state, the subject's cognitive impairment risk level is determined using a preset grading judgment logic;

[0057] S503. Generate and output a screening assessment report containing visualized quantitative indicators and risk assessment conclusions based on the cognitive impairment risk level.

[0058] This invention also provides a cognitive impairment screening system based on resting-state EEG and a multi-task paradigm. Based on the cognitive impairment screening method based on resting-state EEG and a multi-task paradigm described above, the system includes:

[0059] The first acquisition module is used to acquire the subject's eye-closed resting-state EEG signal for a first preset duration under conditions without cognitive task stimulation, and to process the eye-closed resting-state EEG signal to construct a baseline EEG feature set for characterizing the subject's initial neural homeostasis.

[0060] The task execution module is used to present multidimensional cognitive task paradigms to subjects in a preset order. During the subjects' execution of the multidimensional cognitive task paradigms, behavioral data is recorded simultaneously to confirm the task completion rate. The multidimensional cognitive task paradigms are used to apply neural load perturbations to different cognitive networks in the brain.

[0061] The second acquisition module is used to acquire the subject's eye-closed resting-state EEG signal for a second preset duration within a preset time threshold after the end of the multidimensional cognitive task paradigm, and to process the eye-closed resting-state EEG signal to construct a recovery-state EEG feature set for characterizing the steady-state recovery process after neural disturbance.

[0062] The calculation module is used to perform comparative analysis based on the baseline EEG feature set and the recovery period EEG feature set to calculate the neural recovery feature set; wherein, the neural recovery feature set includes neural recovery amplitude features, neural recovery rate features, and brain network remodeling features;

[0063] The output module is used to map the neural recovery feature set to a pre-constructed healthy neural recovery norm space, calculate the degree of deviation of the neural recovery feature set from the healthy neural recovery norm, and output the screening assessment results used to characterize the subject's cognitive impairment risk based on the degree of deviation.

[0064] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0065] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described method.

[0066] The beneficial effects of this invention are as follows:

[0067] This invention, by introducing neural recovery ability as a core screening indicator, overcomes the limitations of traditional single resting-state or task-based EEG analysis, which only reflects the brain's instantaneous state. It transforms multi-dimensional cognitive tasks into standardized neural perturbation sources and utilizes a strongly time-constrained process of "baseline steady state - neural perturbation - recovery steady state" to quantitatively assess the steady-state regulatory dynamics of the brain's nervous system after load. This not only significantly reduces the subjective interference of subjects' education level and cultural background on screening results but also keenly captures neural compensation impairment signals (such as alpha wave rebound lag or network reconstruction failure) that have already appeared in patients with early-stage latent cognitive impairment when their behavioral performance is still acceptable. This greatly improves the sensitivity and objectivity of screening for early and mild cognitive impairment in Alzheimer's disease. Attached Figure Description

[0068] Figure 1 This is a schematic diagram of a method flow according to an embodiment of the present invention.

[0069] Figure 2 This is a schematic diagram of the device structure according to an embodiment of the present invention.

[0070] Figure 3This is a schematic diagram of the internal structure of a computer device according to an embodiment of the present invention.

[0071] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0072] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0073] This invention utilizes electroencephalography (EEG) signals to capture changes in brain neural plasticity, thereby screening for early and mild cognitive impairment (MCI) in Alzheimer's disease (AD). It transforms cognitive tasks from direct assessment objects into sources of neural perturbation. By modeling the dynamics of resting-state EEG recovery, it overcomes the problem of insufficient sensitivity of traditional single analysis modes of resting-state or task-state to early cognitive impairment, reduces the impact of educational background and cultural differences on screening results, and is suitable for early screening and clinical auxiliary assessment of cognitive impairment.

[0074] To address the issue that existing cognitive impairment screening technologies focus only on static EEG states or single-task performance, making it difficult to reveal early impairment of neuromodulation, this study uses a multidimensional cognitive task as a standardized source of neural perturbation. By differentially modeling the resting-state EEG signals before and after the perturbation, it assesses the brain's ability to recover neural homeostasis, rather than simply analyzing the task execution process itself. For example... Figure 1 As shown, the present invention provides a cognitive impairment screening method based on resting-state EEG and a multi-task paradigm, comprising:

[0075] S1. Acquire the subject's eye-closed resting-state EEG signal for a first preset duration under conditions without cognitive task stimulation, and process the eye-closed resting-state EEG signal to construct a baseline EEG feature set to characterize the subject's initial neural homeostasis.

[0076] In one embodiment, step S1 includes the following sub-steps:

[0077] S101. The subject wears a portable EEG acquisition device (e.g., an EEG cap) to detect the impedance value of each electrode channel and ensure that the contact impedance of all channels is lower than a preset threshold in order to ensure signal acquisition quality.

[0078] After confirming that the device is properly fitted, the subject is prompted via voice or screen to remain with their eyes closed, body relaxed, and without engaging in any specific mental activity. Under these conditions, without any external cognitive task stimulation, data acquisition is initiated, recording the resting-state EEG signals with eyes closed for a first preset duration (e.g., 2-5 minutes). The signals collected during this phase reflect the subject's brain's intrinsic neural oscillation patterns and default mode network connectivity under no-load conditions.

[0079] S102. The collected resting-state EEG signals with eyes closed are processed to remove artifacts, including power frequency interference, electrooculography (EOG) artifacts and electromyography (EMG) artifacts, in order to obtain a clean EEG time series.

[0080] Subsequently, frequency band decomposition processing is performed. Through filtering or time-frequency analysis methods, the EEG signal is decomposed into multiple preset frequency bands with specific physiological significance. The preset frequency bands include at least: Delta band (usually 1-4Hz), Theta band (usually 4-8Hz), Alpha band (usually 8-13Hz), and Beta band (usually 13-30Hz).

[0081] S103. Based on the preprocessed frequency-segmented signals, calculate baseline characteristic indices for quantifying the initial neural homeostasis, and construct a baseline EEG feature set. The baseline EEG feature set shall contain at least the following baseline values:

[0082] ① Baseline relative power spectral density For each acquisition channel c and each frequency band b, the relative power spectral density during the baseline period is calculated using Fourier transform or power spectral estimation. The specific calculation formula is as follows:

[0083] ;

[0084] in, Here is the power spectrum of channel c. and These are the lower and upper frequency limits for frequency band b, respectively. and The lower and upper frequency limits for the entire frequency band used when calculating relative power; this index is used as the subtraction benchmark when subsequently calculating the neural recovery amplitude characteristic (RI).

[0085] ② Baseline Alpha wave steady-state reference value Extract the global average power time series of the Alpha band and calculate its value for a first preset duration. The mean value within the range is used as a benchmark for assessing the rate of neural recovery. ;

[0086] in, This is the time series of Alpha power during the baseline resting period.

[0087] ③ Baseline brain functional connectivity model A sliding time window method was used to construct a dynamic functional connectivity matrix for the baseline period. By calculating the correlations of signals from different channels or brain regions within the time window (such as Pearson correlations or Spearman correlations), a series of transient connectivity matrices were obtained, and their average values ​​were calculated as a reference for the baseline network topology.

[0088] ;

[0089] in, This represents the functional connectivity matrix within the k-th sliding time window of the baseline period. This represents the total number of windows during the baseline period; this model is used to subsequently assess the stability of brain network remodeling.

[0090] S2. Present a multidimensional cognitive task paradigm to the subject in a preset order. During the subject's execution of the multidimensional cognitive task paradigm, record behavioral data simultaneously to confirm the task completion rate. The multidimensional cognitive task paradigm is used to apply neural load perturbation to different cognitive networks in the brain.

[0091] In one embodiment, step S2 includes the following sub-steps:

[0092] S201. After baseline EEG acquisition, the control and interaction device (such as a display screen, speaker, and input terminal) immediately switches to task mode and presents task stimuli covering different cognitive dimensions to the subject in a preset order. The multi-dimensional cognitive task paradigm includes the following three core task types, which apply perturbations to different brain networks respectively:

[0093] ① Episodic Memory Task: This task applies neural load to the brain's memory network. A story or list of phrases with a narrative is presented to the subject via audio or image, and the subject is then asked to repeat it. During this process, a speech-semantic comparison unit is invoked to automatically collect the subject's repetition, convert it into text using speech recognition technology, and perform a semantic comparison with keywords in a standard content library to calculate the accuracy of recall.

[0094] ② Working memory task: This task applies neural load to the brain's attentional network; it presents a series of information (e.g., number sequences, alphabetical order, or N-back paradigms) that the subject needs to retain and process within a short period of time. Subjects input answers via buttons or touchscreen, and the system identifies their input in real time, compares it with a standard answer database, and records their reaction time and accuracy.

[0095] ③ Visuospatial Memory Task: This task applies neural load to the brain's spatial processing network; it requires subjects to perform operations such as drawing shapes, building blocks, or judging spatial rotation. For drawing tasks, the outlines and stroke order of the shapes drawn by the subjects are captured using touch screen or electronic pen trajectory recognition technology, and then compared with a standard image library for geometric similarity, providing an accuracy score.

[0096] S202. Throughout the period during which the subject performs the above cognitive task, the following data collection operations are performed in parallel:

[0097] Task-based EEG acquisition: Continuously record the subject's EEG signals as a marker of the "perturbation period" on the timeline to define the time boundary between the baseline period and the recovery period.

[0098] Behavioral data recording: Real-time recording of subjects' reaction times and accuracy rates in various tasks. Behavioral data is not used as the sole or primary basis for diagnosing cognitive impairment, but rather as auxiliary parameters to verify the effectiveness of neural perturbations. By determining whether the behavioral data reaches a minimum preset threshold (e.g., accuracy rate > 60%), it is confirmed whether the subject truly and effectively participated in the cognitive task, thereby ensuring that the subsequent recovery period EEG recordings are based on the regression process after a state of high neural load, rather than an invalid test.

[0099] S3. Within a preset time threshold after the end of the multidimensional cognitive task paradigm, collect the subject's eye-closed resting-state EEG signal for a second preset duration during the recovery period, and process the eye-closed resting-state EEG signal during the recovery period to construct a recovery period EEG feature set for characterizing the steady-state recovery process after neural disturbance.

[0100] In one embodiment, step S3 includes the following sub-steps:

[0101] S301. At the moment the multidimensional cognitive task paradigm ends, the subject is prompted via voice with "Task ended, please close your eyes and relax immediately." To capture the fleeting neural recovery effect (especially the alpha wave rebound phenomenon), a strict preset time threshold is set (e.g., within 0-5 seconds after the task ends). The task end signal is monitored to ensure a seamless switch to the EEG resting-state acquisition mode within the preset time threshold. If the switching time exceeds the threshold, the data is marked as invalid to avoid losing the information required for screening due to the complete recovery of neural activity caused by the time delay.

[0102] S302. After the subject re-enters the closed-eye resting state, collect the closed-eye resting-state EEG signal during the recovery period for a second preset duration (e.g., 2-5 minutes, usually consistent with or equivalent to the first preset duration in step S1). The EEG signal collected at this time contains rich information about the transition of the nervous system from a high-load state to a low-energy steady state.

[0103] S303. The acquired recovery period signals are processed for artifact removal and frequency band decomposition, and recovery period feature indicators are extracted for subsequent differential analysis to construct a recovery period EEG feature set. This feature set includes at least:

[0104] ① Average relative power spectral density during the recovery period Calculate the average relative power spectral density over the entire recovery period. For any frequency band b, the eigenvalues ​​of the recovery period are... Defined as the average relative power spectral density of this frequency band during the recovery period, used to compare with baseline period eigenvalues. Perform a difference comparison.

[0105] ② Alpha power time series during recovery period The dynamic sequence of alpha band power change over time is extracted from the recovery signal. A preset set of EEG channels related to the central-parietal lobe or default network is selected, and the average power of these channels in the alpha band changes with time t is calculated. The calculation formula is as follows:

[0106] ;

[0107] in, This represents the global Alpha band average power at time t during the recovery period; Indicates the number of channels in the selected channel set; This represents the relative power spectral density of channel c in the Alpha band at time point t; t is a time point within the recovery period, satisfying... This time series It is directly used in step S4 for regression analysis with the baseline steady-state reference value to calculate the Alpha wave bounce slope.

[0108] ③ Average functional connectivity matrix of the brain during the recovery period Using the same sliding window parameters as the baseline period (window length w and step size), dynamic functional connectivity calculations are performed on the recovery period signal. First, the correlation between channel i and channel j within the k-th sliding time window is calculated. ,in, represents the functional connectivity strength between brain region i and brain region j within the k-th sliding time window; i,j are EEG channels or brain region indices. This refers to the signal of the i-th channel within the k-th sliding window; This refers to the signal of the j-th channel within the k-th sliding window; The correlation calculation function can be either Pearson or Spearman correlation; then, the average value of the connectivity matrix generated by all sliding windows during the recovery period is calculated, as shown in the following formula: ;

[0109] in, This is the functional connectivity matrix within the k-th sliding time window of the recovery period. The total number of windows during the recovery period; using the same sliding window parameters as S1, a sequence of dynamic functional connectivity matrices for the recovery period is constructed, and their average matrix is ​​calculated. This model characterizes the reorganization of the functional network topology after the brain experiences a cognitive task shock, and is used to compare with a baseline network model to calculate brain network remodeling features.

[0110] S4. Based on the baseline EEG feature set and the recovery period EEG feature set, a comparative analysis is performed to calculate the neural recovery feature set; wherein, the neural recovery feature set includes neural recovery amplitude features, neural recovery rate features, and brain network remodeling features.

[0111] In one embodiment, step S4 includes the following sub-steps:

[0112] S401. Calculate the characteristics of neural recovery amplitude, compare the power spectral density between the baseline period and the recovery period, and calculate the energy difference in a specific frequency band.

[0113] ① Band recovery index calculation: For each preset frequency band b (such as Delta, Theta, Alpha or Beta), the baseline relative power spectral density calculated in S1 is used. The relative power spectral density during the recovery period calculated in S3 The recovery index is calculated using the following formula. :

[0114] ;

[0115] in, The neural recovery index represents frequency band b. If... , indicates full recovery; if This indicates that there is residual enhancement in this frequency band after the task disturbance (e.g., an increase in slow waves); if This indicates excessive nerve inhibition or recovery.

[0116] ② Neurological recovery deviation assessment: The calculated deviation will be used to assess the deviation. Mapped to a pre-constructed healthy neural recovery norm space Calculate the deviation :

[0117] ;

[0118] in, and These represent the mean and standard deviation of the recovery index of healthy subjects in frequency band b, respectively.

[0119] S402. Calculate the neural recovery rate characteristics, and call the Alpha band power time series extracted in step S3. and the baseline Alpha steady-state reference value calculated in step S1. During the recovery period, a linear regression was performed on the difference between the power series and the baseline, and the regression slope ARS was calculated.

[0120] ;

[0121] Where ARS represents the Alpha bounce slope; This represents the i-th sampling time point within the recovery period; This represents the mean value at the sampling time points; Indicates a point in time The corresponding Alpha power value; The ARS value represents the steady-state reference value of the Alpha band during the baseline period; the larger the ARS value, the faster the brain returns to steady state; if the ARS is close to 0 or negative, it indicates slow neural recovery or failure of the regulatory mechanism.

[0122] S403. Calculate brain network reconstruction features, and call the baseline average brain functional connectivity matrix generated in step S1. The average brain functional connectivity matrix generated during the recovery period in step S3 First, we utilize vectorization operations. Expand the two-dimensional matrix into a one-dimensional vector, and then calculate the Spearman rank correlation coefficient between the two, as shown in the following formula:

[0123] ;

[0124] in, This is the Spearman rank correlation coefficient calculation function. A value close to 1 indicates that the network topology has successfully reverted to the baseline state during the recovery period; a significant decrease in this value indicates impaired reconstruction of neural synergistic structures.

[0125] S404. Calculate the comprehensive neural recovery score. Based on the above three feature dimensions and combined with preset weights, calculate the comprehensive neural recovery score:

[0126] ;

[0127] in, Deviation in the magnitude of nerve recovery; The normalized Alpha rebound slope; These are features of brain network reconstruction; These are pre-defined weighting coefficients based on the relevance of clinical data. A higher overall neurological recovery score indicates poorer neurological homeostasis recovery and a higher risk of cognitive impairment.

[0128] S5. Map the neural recovery feature set to a pre-constructed healthy neural recovery norm space, calculate the degree of deviation of the neural recovery feature set from the healthy neural recovery norm, and output the screening assessment results used to characterize the subject's cognitive impairment risk based on the degree of deviation.

[0129] In one embodiment, step S5 includes the following sub-steps:

[0130] S501. Pre-store healthy neural recovery norms constructed based on large-scale cognitively normal population data. These norms are not single numerical values, but a multi-dimensional feature space encompassing neural recovery amplitude (RI), recovery rate (ARS), and network reconstruction stability under different age and gender groups. The distribution range of ) (usually defined as (Or 95% confidence interval). The neural recovery feature set calculated by the subject in step S4 is mapped to this multidimensional norm space. For a single feature dimension, it is determined whether the subject's feature value falls within the normal range of the norm. If the feature value exceeds the normal range (e.g., the alpha rebound slope is significantly lower than the norm mean), the specific degree of deviation is calculated. For the comprehensive score dimension, the subject's comprehensive neural recovery score (NRS) is compared with the NRS threshold of the norm population.

[0131] S502. Based on the calculation results of the above-mentioned degree of deviation, a preset grading judgment logic is used to determine the subject's cognitive impairment risk level. A specific example of the judgment logic is as follows:

[0132] Normal: All core characteristics (RI, ARS, dFNC) fall within the healthy norm range, or the overall NRS score is below the preset risk threshold. In this case, the subject's neurological homeostasis regulation ability is considered intact.

[0133] Mild risk: A slight deviation in a single trait (such as only the alpha rebound rate), or the NRS being in the borderline risk range. This suggests that although the subject can complete the cognitive task, there are early signs of impairment in the intrinsic recovery dynamics of their nervous system.

[0134] High risk: Multiple characteristics significantly deviate from the norm (e.g., high amplitude residuals and network reconstruction failure), or NRS significantly exceeds the risk threshold. This suggests that the subject's neural homeostasis regulation mechanism is severely impaired, and there is a high possibility of cognitive impairment.

[0135] S503. Generate a screening assessment report based on the judgment results. The report should include at least visual quantitative indicators, risk assessment conclusions, and neurodynamic explanations.

[0136] like Figure 2 As shown, the present invention also provides a cognitive impairment screening system based on resting-state EEG and a multi-task paradigm. Based on the cognitive impairment screening method based on resting-state EEG and a multi-task paradigm described above, the system includes:

[0137] The first acquisition module 1 is used to acquire the subject's eye-closed resting-state EEG signal for a first preset duration under conditions without cognitive task stimulation, and to process the eye-closed resting-state EEG signal to construct a baseline EEG feature set for characterizing the subject's initial neural homeostasis.

[0138] Task execution module 2 is used to present a multi-dimensional cognitive task paradigm to the subject in a preset order. During the subject's execution of the multi-dimensional cognitive task paradigm, behavioral data is recorded simultaneously to confirm the task completion rate. The multi-dimensional cognitive task paradigm is used to apply neural load perturbation to different cognitive networks in the brain.

[0139] The second acquisition module 3 is used to acquire the subject's eye-closed resting-state EEG signal for a second preset duration within a preset time threshold after the end of the multidimensional cognitive task paradigm, and to process the eye-closed resting-state EEG signal to construct a recovery-state EEG feature set for characterizing the steady-state recovery process after neural disturbance.

[0140] The calculation module 4 is used to perform comparative analysis based on the baseline EEG feature set and the recovery period EEG feature set to calculate the neural recovery feature set; wherein, the neural recovery feature set includes neural recovery amplitude features, neural recovery rate features, and brain network reconstruction features;

[0141] Output module 5 is used to map the neural recovery feature set to a pre-constructed healthy neural recovery norm space, calculate the degree of deviation of the neural recovery feature set from the healthy neural recovery norm, and output the screening assessment results for characterizing the subject's cognitive impairment risk based on the degree of deviation.

[0142] Each of the above modules is used to perform the respective steps in the cognitive impairment screening method based on resting-state EEG and multi-task paradigm. The specific implementation method is as described in the above method embodiments, and will not be repeated here.

[0143] like Figure 3 As shown, the present invention also provides a computer device, which may be a server, and its internal structure may be as follows: Figure 3 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores all data required for the process of a cognitive impairment screening method based on resting-state EEG and a multi-task paradigm. The network interface allows communication with external terminals via a network connection. The computer program is executed by the processor to implement the cognitive impairment screening method based on resting-state EEG and a multi-task paradigm.

[0144] Those skilled in the art will understand that Figure 3The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer equipment on which the present application is applied.

[0145] An embodiment of this application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements any of the above-described cognitive impairment screening methods based on resting-state EEG and a multi-task paradigm.

[0146] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by hardware related to computer program instructions. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the above method embodiments. Any references to memory, storage, databases, or other media provided in this application and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM), such as dynamic RAM (used as main storage) or static RAM (commonly used as cache memory). By way of illustration and not limitation, RAM has various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), and Rambus DRAM (RDRAM).

[0147] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0148] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A cognitive impairment screening method based on resting-state EEG and a multi-task paradigm, characterized in that, include: S1. Acquire the subject's eye-closed resting-state EEG signal for a first preset duration under conditions without cognitive task stimulation, and process the eye-closed resting-state EEG signal to construct a baseline EEG feature set to characterize the subject's initial neural homeostasis. S2. Present a multi-dimensional cognitive task paradigm to the subject in a preset order. During the subject's execution of the multi-dimensional cognitive task paradigm, record behavioral data simultaneously to confirm the task completion rate. The multi-dimensional cognitive task paradigm is used to apply neural load perturbation to different cognitive networks in the brain. S3. Within a preset time threshold after the end of the multidimensional cognitive task paradigm, collect the subject's eye-closed resting-state EEG signal for a second preset duration during the recovery period, and process the eye-closed resting-state EEG signal during the recovery period to construct a recovery period EEG feature set for characterizing the steady-state recovery process after neural disturbance. S4. Based on the baseline EEG feature set and the recovery period EEG feature set, a comparative analysis is performed to calculate the neural recovery feature set; wherein, the neural recovery feature set includes neural recovery amplitude features, neural recovery rate features, and brain network remodeling features; S5. Map the neural recovery feature set to a pre-constructed healthy neural recovery norm space, calculate the degree of deviation of the neural recovery feature set from the healthy neural recovery norm, and output the screening assessment results used to characterize the subject's cognitive impairment risk based on the degree of deviation.

2. The cognitive impairment screening method based on resting-state EEG and multi-task paradigm according to claim 1, characterized in that, S1 specifically includes: S101. Using a portable EEG acquisition device, collect the EEG signal of the subject in a resting state with eyes closed and without engaging in specific thinking activities for the first preset duration. S102. Perform artifact removal processing on the closed-eye resting-state EEG signal, and decompose the artifact-removed signal into Delta band, Theta band, Alpha band and Beta band. S103. Based on the decomposed frequency band signals, calculate the baseline relative power spectral density, the baseline alpha wave steady-state reference value, and the baseline brain functional connectivity network model benchmark value to construct the baseline EEG feature set.

3. The cognitive impairment screening method based on resting-state EEG and multi-task paradigm according to claim 2, characterized in that, In step S103, the formula for calculating the baseline relative power spectral density is: ； in, This represents the baseline average relative power spectral density of channel c within frequency band b; The power spectrum of channel c; and These are the lower and upper frequency limits for frequency band b, respectively. and These are the lower and upper frequency limits across the entire frequency band used in calculating relative power, respectively. The formula for calculating the baseline Alpha wave steady-state reference value is as follows: ； in, This represents the steady-state reference value for the Alpha band during the baseline resting period; Indicates the duration of the baseline resting period; This represents the global average power in the Alpha band at time t during the baseline resting period; The calculation formula for the baseline brain functional connectivity network model is as follows: ； in, This represents the baseline average brain functional connectivity matrix; Indicates the number of sliding windows during the baseline period; This represents the brain functional connectivity matrix within the k-th sliding time window during the baseline period.

4. The cognitive impairment screening method based on resting-state EEG and multi-task paradigm according to claim 1, characterized in that, S2 specifically includes: S201. After the baseline EEG acquisition is completed, the control interaction device is switched to task mode, and the multidimensional cognitive task paradigm is presented to the subject in a preset order to apply neural load to the brain's memory network, attention network and spatial processing network; wherein, the multidimensional cognitive task paradigm includes episodic memory task, working memory task and visuospatial memory task. S202. During the subject's execution of the multidimensional cognitive task paradigm, the task-state EEG signals are continuously recorded, and the subject's behavioral data are recorded simultaneously.

5. The cognitive impairment screening method based on resting-state EEG and multi-task paradigm according to claim 1, characterized in that, S3 specifically includes: S301. Immediately terminate the cognitive task stimulation when the multidimensional cognitive task paradigm ends, and switch to the EEG resting acquisition mode within a preset time threshold. S302. After the subject enters a closed-eye resting state, collect the closed-eye resting state EEG signal during the recovery period of the second preset duration. S303. Extract features from the resting-state EEG signal with eyes closed during the recovery period to construct a recovery period EEG feature set including the recovery period Alpha band power time series, the recovery period average brain functional connectivity network matrix, and the recovery period average relative power spectral density; wherein, the recovery period average relative power spectral density is the average relative power spectral density of each channel in the preset frequency band during the entire recovery period, which is used to perform differential calculation with the baseline relative power spectral density in the baseline EEG feature set.

6. The cognitive impairment screening method based on resting-state EEG and multi-task paradigm according to claim 5, characterized in that, In step S303, the calculation formula for the Alpha band power time series during the recovery period is as follows: ； in, The global alpha band average power at time t during the recovery period; C represents the selected set of EEG channels related to the central-parietal lobe or default network. This represents the number of channels in set C; This represents the relative power spectral density of channel c in the Alpha band at time t; t is the time variable during the recovery period, and ; The formula for calculating the average brain functional connectivity matrix during the recovery period is as follows: ； in, This represents the average brain functional connectivity matrix during the recovery period; This indicates the total number of sliding windows during the recovery period; This represents the brain functional connectivity matrix within the k-th sliding time window of the recovery period. This matrix is ​​calculated from the correlation between signals from different EEG channels within the k-th sliding time window.

7. The cognitive impairment screening method based on resting-state EEG and multi-task paradigm according to claim 1, characterized in that, S4 specifically includes: S401. Based on the relative power spectral density difference between the baseline EEG feature set and the recovery period EEG feature set, calculate the neural recovery amplitude characteristics, and calculate the deviation of the neural recovery amplitude characteristics from the healthy norm. The calculation formula is as follows: ； ； in, This indicates the characteristics of the neural recovery amplitude in frequency band b; This represents the average relative power spectral density within frequency band b during the recovery period; This represents the average relative power spectral density within frequency band b during the baseline period; This indicates the degree of neural recovery deviation in frequency band b; This represents the mean recovery index of the healthy subject norm in frequency band b; This represents the standard deviation of the recovery index of healthy subjects in frequency band b; S402. Based on the Alpha band power time series in the recovery period EEG feature set and the baseline Alpha steady-state reference value in the baseline EEG feature set, calculate the Alpha rebound slope as a characteristic of neural recovery rate. The calculation formula is as follows: ； Where ARS represents the Alpha bounce slope; This represents the i-th sampling time point within the recovery period; This represents the mean value at the sampling time points; Indicates a point in time The corresponding Alpha power value; This represents the steady-state reference value for the Alpha band during the baseline period; S403. Based on the correlation between the brain functional connectivity network matrix of the baseline EEG feature set and the recovery period EEG feature set, calculate the brain network reconstruction features using the following formula: ； in, Indicates brain network remodeling characteristics; This represents the baseline average brain functional connectivity matrix; This represents the average brain functional connectivity matrix during the recovery period; This represents the operation of expanding a matrix into a vector in a preset order; This represents the Spearman rank correlation coefficient calculation function; S404. Based on the deviation of the neural recovery amplitude characteristics, the neural recovery rate characteristics, and the brain network remodeling characteristics, a comprehensive neural recovery score is calculated using preset weights. The calculation formula is as follows: ； NRS stands for Comprehensive Neurological Recovery Score; Indicates the degree of deviation in the extent of nerve recovery; This represents the normalized Alpha bounce slope; Indicates brain network remodeling characteristics; These are preset weighting coefficients.

8. The cognitive impairment screening method based on resting-state EEG and multi-task paradigm according to claim 1, characterized in that, S5 specifically includes: S501. Map the neural recovery feature set and comprehensive neural recovery score to a pre-constructed healthy neural recovery norm space, and calculate its deviation from the healthy norm; wherein, the healthy neural recovery norm space is a multi-dimensional feature distribution interval constructed based on large-scale cognitively normal population data, including confidence intervals for neural recovery amplitude, neural recovery rate and brain network reconstruction features; S502. Based on the deviation state, the subject's cognitive impairment risk level is determined using a preset grading judgment logic; S503. Generate and output a screening assessment report containing visualized quantitative indicators and risk assessment conclusions based on the cognitive impairment risk level.

9. A cognitive impairment screening system based on resting-state EEG and a multi-task paradigm, based on the cognitive impairment screening method based on resting-state EEG and a multi-task paradigm as described in any one of claims 1 to 8, characterized in that, The system includes: The first acquisition module is used to acquire the subject's eye-closed resting-state EEG signal for a first preset duration under conditions without cognitive task stimulation, and to process the eye-closed resting-state EEG signal to construct a baseline EEG feature set for characterizing the subject's initial neural homeostasis. The task execution module is used to present multidimensional cognitive task paradigms to subjects in a preset order. During the subjects' execution of the multidimensional cognitive task paradigms, behavioral data is recorded simultaneously to confirm the task completion rate. The multidimensional cognitive task paradigms are used to apply neural load perturbations to different cognitive networks in the brain. The second acquisition module is used to acquire the subject's eye-closed resting-state EEG signal for a second preset duration within a preset time threshold after the end of the multidimensional cognitive task paradigm, and to process the eye-closed resting-state EEG signal to construct a recovery-state EEG feature set for characterizing the steady-state recovery process after neural disturbance. The calculation module is used to perform comparative analysis based on the baseline EEG feature set and the recovery period EEG feature set to calculate the neural recovery feature set; wherein, the neural recovery feature set includes neural recovery amplitude features, neural recovery rate features, and brain network remodeling features; The output module is used to map the neural recovery feature set to a pre-constructed healthy neural recovery norm space, calculate the degree of deviation of the neural recovery feature set from the healthy neural recovery norm, and output the screening assessment results used to characterize the subject's cognitive impairment risk based on the degree of deviation.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

Images