Combined tbs stroke motor aphasia abnormality assessment system

Abstract

This invention relates to the field of healthcare technology, specifically to a stroke motor aphasia abnormality assessment system combined with TBS (Traumatic Brain Spectroscopy). The system includes a vascular abnormality analysis module, a point set acquisition module, a mismatch analysis module, and an operational adjustment module. The system generates microvascular abnormality indicators by detecting the concentrations of vascular endothelial growth factor (VEGF) and hypoxia-inducible factor (HIF); simultaneously extracts EEG frequency band energy and near-infrared oxyhemoglobin concentration, and obtains neural excitation point sets and blood flow response point sets through maximum search; based on the microvascular abnormality indicators, it amplifies the blood flow response amplitude and performs spatiotemporal weighted matching on the two types of point sets to obtain the degree of mismatch reflecting the true supply and demand relationship; furthermore, it calculates the cumulative abnormality coefficient by combining the multi-cycle mismatch degree to achieve TBS operational adjustment. This scheme overcomes the shortcomings of traditional methods that rely on strict signal alignment, effectively distinguishes between physiological time drift and decoupling, and improves the robustness of TBS operation.

Description

Technical Field

[0001] This invention relates to the field of healthcare technology, and more specifically to a stroke motor aphasia assessment system combined with TBS. Background Technology

[0002] Transcranial magnetic stimulation (TBS) can improve post-stroke motor aphasia; however, in practical abnormality assessment, the following technical challenges exist: the neurovascular coupling (NVC) response exhibits a non-rigid time lag. Microvascular abnormalities cause not only a delay in hemodynamic response relative to neuroelectrical activity, but this delay is also often variable (non-rigid), fluctuating within a few seconds. Existing techniques based on linear correlation analysis or synchronicity monitoring require strict signal alignment on the time axis, which is difficult to adapt to such fluctuations. This can easily lead to misinterpreting physiological time drift as functional decoupling, affecting TBS operational analysis and resulting in insufficient robustness of TBS operation. Summary of the Invention

[0003] To address the technical problem in related technologies that easily misinterpret physiological time drift as functional decoupling, affecting TBS operation analysis and resulting in insufficient robustness of TBS operation, this invention provides a stroke motor aphasia abnormality assessment system combined with TBS. The specific technical solution adopted is as follows: This invention proposes a stroke motor aphasia assessment system combined with TBS, the system comprising: The vascular abnormality analysis module is used to obtain the concentration values ​​of vascular endothelial growth factor and hypoxia-inducible factor, respectively; and to perform nonlinear weighted analysis by combining the two concentration values ​​to obtain microvascular abnormality indicators. The point set acquisition module is used to simultaneously acquire the frequency band energy of EEG signals and the oxygenated hemoglobin concentration of near-infrared brain functional imaging; to perform a maximum search on the frequency band energy to determine the neural excitation point set; and to perform a maximum search on the oxygenated hemoglobin concentration to determine the blood flow response point set. The mismatch analysis module is used to perform weighted gain on the blood flow response point set based on microvascular abnormality indicators, and to perform spatiotemporal weighted matching on the neural excitation point set and the gained blood flow response point set to obtain the degree of supply-demand mismatch. The operation adjustment module is used to determine the cumulative anomaly coefficient of global variables based on the degree of supply and demand mismatch under different evaluation periods; and to generate recommended parameters for the runtime of TBS equipment based on the cumulative anomaly coefficient.

[0004] Furthermore, the concentration values, the frequency band energy, and the oxygenated hemoglobin concentration are all standardized data.

[0005] Furthermore, the nonlinear weighted analysis combining the two concentration values ​​yields microvascular abnormality indicators, including: The concentration values ​​of vascular endothelial growth factor and hypoxia-inducible factor are respectively input into a pre-trained nonlinear mapping model; the output is a microvascular abnormality index characterizing the degree of microcirculation perfusion restriction.

[0006] Furthermore, the step of performing a maximum search on the frequency band energy to determine the set of neural excitation points includes: Local peak points of frequency band energy time series are detected within a preset time window, and local peak points exceeding a preset energy threshold are taken as neural excitation points; neural excitation points form a neural excitation point set.

[0007] Furthermore, the step of performing a maximum search on the oxygenated hemoglobin concentration to determine the blood flow response point set includes: Within a preset time window corresponding to the nerve excitation point, the local maximum value of oxygenated hemoglobin concentration is extracted as the blood flow response point; the blood flow response points form a blood flow response point set.

[0008] Furthermore, the weighting and gaining of the blood flow response point set based on microvascular abnormality indicators includes: The oxygenated hemoglobin concentration amplitude at each blood flow response point was compensated by using microvascular abnormality indicators to obtain the amplified oxygenated hemoglobin concentration.

[0009] Furthermore, spatiotemporal weighted matching is achieved based on the KM algorithm.

[0010] Furthermore, the spatiotemporal weighted matching of the neural excitation point set and the amplified blood flow response point set to obtain the degree of supply-demand mismatch includes: Construct a bipartite graph matching model, with the neural excitation points in the neural excitation point set as nodes on one side and the blood flow response points in the amplified blood flow response point set as nodes on the other side. For any neural excitation point and a blood flow response point, the weight of the matching edge between them is determined based on the time interval between them and the relative difference in their amplitudes after gain. The KM algorithm is used to find the minimum sum of weights of all paired edges, which is used as the degree of supply-demand mismatch.

[0011] Furthermore, determining the weight of the matching edge between the two based on the time interval between them and the relative difference in their post-gain amplitudes includes: If the time interval between the nerve excitation point and the blood flow response point is less than the preset physiological response interval, then the first interval weight is assigned; otherwise, the second interval weight is assigned. The first interval weight is less than the second interval weight. When the blood flow response point lags behind the nerve excitation point, a first lag weight is assigned; otherwise, a second lag weight is assigned. The first lag weight is less than the second lag weight. Calculate the absolute value of the difference between the amplitudes after gaining the nerve excitation point and the blood flow response point, and normalize the absolute value of the difference to obtain the amplitude difference cost. By combining interval weight, lag weight, and amplitude difference cost, the edge weights between the nerve excitation point and the blood flow response point are determined.

[0012] Furthermore, determining the cumulative anomaly coefficient of the global variable based on the degree of supply-demand mismatch under different evaluation periods includes: Calculate the sum of the supply-demand mismatch levels across all assessment periods as the cumulative anomaly coefficient.

[0013] The present invention has the following beneficial effects: This invention obtains microvascular abnormality indicators by acquiring the concentrations of vascular endothelial growth factor and hypoxia-inducible factor and performing nonlinear weighted analysis; simultaneously extracts EEG frequency band energy and near-infrared oxyhemoglobin concentration, and determines the neural excitation point set and blood flow response point set through maximum search; weights the blood flow response point set based on the microvascular abnormality indicators, and perform spatiotemporal weighted matching between the amplified blood flow response point set and the neural excitation point set to obtain the degree of supply-demand mismatch; and determines the cumulative abnormality coefficient based on the degree of supply-demand mismatch under different evaluation periods and adjusts the TBS operation accordingly. This invention abandons traditional linear correlation or synchronicity analysis methods that rely on strict signal time alignment, and instead adopts a spatiotemporal weighted matching mechanism based on discrete event point sets. This mechanism can effectively tolerate the inherent, non-rigid physiological time drift in the neurovascular response of stroke patients, avoiding the misjudgment of reasonable lag as functional decoupling. At the same time, the weighted gain of microvascular abnormality indicators eliminates the interference of individual vascular state differences on blood flow response, so that the degree of supply-demand mismatch truly reflects the efficacy of neuromodulation, thereby achieving accurate abnormality assessment and significantly improving the reliability and robustness of TBS operation. Attached Figure Description

[0014] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a structural diagram of a stroke motor aphasia abnormality assessment system combined with TBS provided in one embodiment of the present invention. Detailed Implementation

[0016] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a stroke motor aphasia assessment system combined with TBS proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0018] The following description, in conjunction with the accompanying drawings, details the specific scheme of the stroke motor aphasia abnormality assessment system provided by the present invention.

[0019] Please see Figure 1 The diagram illustrates a structural diagram of a stroke motor aphasia assessment system combined with TBS, according to an embodiment of the present invention. The system includes: The vascular abnormality analysis module 101 is used to obtain the concentration values ​​of vascular endothelial growth factor and hypoxia-inducible factor respectively; and to perform nonlinear weighted analysis by combining the two concentration values ​​to obtain microvascular abnormality indicators.

[0020] Because the microvascular status varies significantly among different participants, some participants, due to arteriosclerosis or endothelial dysfunction, experience limited blood flow response amplitude even with strong neural activity; while participants with good vascular function can generate sufficient perfusion. Directly using raw blood flow signals for neurovascular coupling assessment could misinterpret this low response caused by vascular structural abnormalities as functional decoupling, thus interfering with abnormal assessment and affecting the robustness of TBS operation.

[0021] Therefore, in this embodiment of the invention, a priori index reflecting the basic state of an individual's microcirculation is constructed: by obtaining vascular endothelial growth factor (VEGF-A) and hypoxia-inducible factor (HIF-A) from the subject respectively. The concentration values ​​of these two indicators reflect the angiogenesis and repair potential and the tissue's compensatory regulatory capacity in response to hypoxic environments, respectively. A nonlinear weighted analysis was then performed on these two indicators to generate microvascular abnormality indicators.

[0022] It should be noted that the concentration values ​​of vascular endothelial growth factor and hypoxia-inducible factor obtained in the embodiments of the present invention have all undergone data standardization processing and have been mapped to a unified [database / structure]. The interval is used to eliminate the differences in physical dimensions and numerical magnitudes of different biochemical indicators.

[0023] For example, a pre-set general medical reference dataset is called, which contains the reference minimum and maximum values ​​of the above biochemical indicators for people of the same age group. Then, the raw data is standardized using the restricted Min-Max normalization algorithm (Clipped Min-Max Normalization). Then, a boundary truncation operation (Clip) is performed to mark data greater than the reference maximum value as 1 and data less than the reference minimum value as 0. This effectively prevents calculation overflow or negative value errors caused by abnormal indicators of individual test subjects (such as exceeding the upper limit of the reference range), and ensures the stability of subsequent calculation inputs.

[0024] After data standardization, microvascular abnormality indicators are calculated, which can comprehensively reflect the assessment parameters of the degree of restriction of microvascular response to blood flow. Further, in some embodiments of the present invention, a nonlinear weighted analysis is performed combining two concentration values ​​to obtain microvascular abnormality indicators, including: inputting the concentration values ​​of vascular endothelial growth factor and hypoxia-inducible factor into a pre-trained nonlinear mapping model; and outputting microvascular abnormality indicators characterizing the degree of microcirculation perfusion restriction.

[0025] Specifically, the pre-trained nonlinear weighted model can be, for example: In the formula, Indicates an abnormal microvascular index. This represents the standardized concentration value of vascular endothelial growth factor. This represents the standardized concentration value of hypoxia-inducible factor. The weighting coefficient for the effect of hypoxia is used to adjust the proportion of the influence of hypoxia-inducible factors on the overall condition. In this embodiment, it is set to an empirical value of 0.3 (this value can be calibrated by fitting historical clinical data). To prevent tiny positive numbers with a denominator of zero (e.g., taking values...) ). The cutoff value is set to prevent numerical explosion caused by extreme anomalies in biochemical indicators. The specific value can be, for example, 3.0.

[0026] The lower the concentration of vascular endothelial growth factor (approaching 0), the smaller the denominator, and the higher the calculated value. The higher the value, the worse the physiological capacity for vascular repair, and the higher the restriction on blood flow (i.e., the state factor). The lower the concentration of hypoxia-inducible factor (approaching 0), the better. The larger the value, the larger the multiplier term inside the brackets, which in turn increases the multiplier term. Increased size. This corresponds to a weaker compensatory capacity and a poorer overall vascular condition. The final output... The higher the value, the less ideal the patient's microvascular condition, and the greater the reduction in blood flow response will be applied in subsequent algorithms during evaluation.

[0027] The point set acquisition module 102 is used to simultaneously acquire the frequency band energy of the electroencephalogram (EEG) signal and the oxygenated hemoglobin concentration of near-infrared brain functional imaging; perform a maximum search on the frequency band energy to determine the neural excitation point set; and perform a maximum search on the oxygenated hemoglobin concentration to determine the blood flow response point set.

[0028] In this embodiment of the invention, continuous signals are converted into a set of discrete event points with clear physiological significance. Specific frequency band energy of EEG signals and oxygenated hemoglobin concentration from near-infrared imaging are acquired simultaneously, and maximum values ​​are searched for both. Significant energy peak points are defined as a set of neural excitation points, and significant blood flow concentration peak points are defined as a set of blood flow response points.

[0029] The pulse trigger signal of the TBS device is used as the time synchronization reference. Whenever the end of a stimulus pulse sequence is detected, the system does not immediately truncate the data, but instead initiates a preset time window for hysteresis analysis. Let the current pulse be the trigger signal. The end time of the pulse in each stimulation cycle is The time period captured by the system is defined as .in, The pulse start time; A preset physiological hysteresis margin (e.g., set to 8 seconds) is used. This margin is designed to cover the longest possible response delay that may occur in damaged blood vessels, ensuring that the complete blood flow response waveform is included within the preset time window. Within this window, the system first extracts the Beta band energy sequence of the EEG signal. and fNIRS oxygenated hemoglobin concentration sequence And then standardize the process.

[0030] It should be noted that the frequency band energy of the EEG signal and the oxygenated hemoglobin concentration of the near-infrared brain functional imaging obtained in the embodiments of the present invention have been standardized, and the standardization can be specifically, for example, Z-Score standardization.

[0031] To eliminate dimensional differences and highlight the degree of signal variation relative to the baseline, the system pre-collects background noise parameters (average frequency band energy of the EEG signal). Standard deviation Mean blood oxygen Standard deviation Z-Score normalization was performed on the frequency band energy of the above EEG signals and the oxygenated hemoglobin concentration of near-infrared brain functional imaging.

[0032] For EEG signals, the standardized sequence The calculation is as follows: For near-infrared spectral signals in near-infrared brain functional imaging, the standardized sequence The calculation is as follows: After this processing step, all signals are converted into dimensionless signal-to-noise ratio (SNR) values ​​(Standard Score). If This indicates that the signal strength at the current moment is higher than the resting baseline. This indicates that the current signal strength is 3 times the standard deviation of the background noise, which is statistically significant. To prevent tiny positive numbers with a denominator of zero (e.g., taking values...) ), whose dimensions are in the same position as the denominator.

[0033] This processing enables electrical and optical signals, which originally had different units, to achieve numerical alignment in terms of the statistical dimension of the degree of deviation from the baseline, providing a basis for comparability in subsequent weighted calculations.

[0034] In this embodiment of the invention, subsequent neural-blood flow analysis is performed using frequency band energy and oxygenated hemoglobin concentration. When evaluating the efficacy of combined TBS for stroke-induced motor aphasia, the raw electroencephalogram (EEG) signals and functional near-infrared spectroscopy (fNIRS) data are both continuous, high-dimensional time series. Direct coupling analysis is easily affected by background noise, rhythm fluctuations, and non-task-related activities, making it difficult to accurately locate effective neural-blood flow events related to language tasks or stimulus responses.

[0035] To transform continuous waveforms into discrete events that can be used for subsequent bipartite graph matching, the system needs to identify key physiological event points in the standardized data. This embodiment of the invention specifically introduces positive filtering logic to avoid interference from invalid or inhibitory signals.

[0036] Furthermore, in some embodiments of the present invention, performing a maximum search on the frequency band energy to determine a set of neural excitation points includes: detecting local peak points of the time series of frequency band energy within a preset time window, and taking local peak points exceeding a preset energy threshold as neural excitation points; the neural excitation points constitute a set of neural excitation points.

[0037] The maximum value search of oxygenated hemoglobin concentration is performed to determine the blood flow response point set, including: extracting the local maximum value of oxygenated hemoglobin concentration as the blood flow response point within a preset time window corresponding to the nerve excitation point; the blood flow response points form the blood flow response point set.

[0038] The system performs a local maximum search algorithm on the standardized values ​​of the frequency band energy. To ensure that the extracted points represent real neural excitation rather than inhibition or noise, the system sets dual screening conditions: the peak amplitude must be greater than a preset neural saliency threshold (e.g., a value of 2); if the peak amplitude is greater than 0, fluctuations below the baseline are excluded. Points that meet the above conditions are taken as neural excitation points, forming a set of neural excitation points.

[0039] Similarly, the system for Perform a local maximum search. The amplitude must be greater than a preset blood flow significance threshold (e.g., 1.5); the amplitude must be greater than 0 (i.e.,...). Points that meet the conditions are designated as blood flow response points, forming a blood flow response point set.

[0040] In this embodiment of the invention, the continuous physiological process is reduced to two discrete sets of points. This approach not only significantly reduces the amount of data required for subsequent calculations but also simplifies the complex waveform morphology into two core features: occurrence time and response intensity. This allows subsequent algorithms to focus on evaluating the degree of matching between these two features on the supply and demand sides. The explicit positive filtering logic effectively avoids logical errors in subsequent algebraic operations caused by negative Z-Score values ​​(representing a suppression state below the baseline).

[0041] The mismatch analysis module 103 is used to perform weighted gain on the blood flow response point set based on microvascular abnormality indicators, and to perform spatiotemporal weighted matching on the neural excitation point set and the gained blood flow response point set to obtain the degree of supply-demand mismatch.

[0042] The degree of substantial supply-demand mismatch that still exists after tolerating physiological time lags and individual differences in vascular status is calculated using a globally optimal matching algorithm. To quantify the reasonable cost of pairing any neural excitation event with any blood flow response event, a weighted evaluation system can be constructed.

[0043] If the original blood flow response amplitude is directly matched with the intensity of nerve excitation, systematic bias will occur due to individual differences in microvascular function (such as arteriosclerosis and endothelial damage). Under the same neural activity, the blood flow response of patients with vascular abnormalities will naturally be lower. To address this issue, this step focuses on eliminating the interference of vascular structural limitations on coupling discrimination: First, the amplitude of the blood flow response point set is individually amplified based on microvascular abnormality indicators so that the amplified blood flow value reflects the true effective perfusion capacity; then, the nerve excitation point set and the amplified blood flow response point set are spatiotemporally weighted and matched, taking into account time lag, causal order and amplitude consistency, to calculate the degree of mismatch that represents the true supply and demand relationship.

[0044] Furthermore, in some embodiments of the present invention, weighting the blood flow response point set based on microvascular abnormality indicators includes: using microvascular abnormality indicators to compensate for the amplitude of oxygenated hemoglobin concentration at each blood flow response point, thereby obtaining the amplified oxygenated hemoglobin concentration.

[0045] Specifically, the weighted gain can be used as an example, where microvascular abnormality indicators are used as gain indicators, and then the product of the gain indicator and the amplitude of oxyhemoglobin concentration is calculated. This product represents the parameter data obtained after overcoming vascular limitations. The worse the vascular condition (the larger the microvascular abnormality indicators), the greater the weighted gain on the original measurement should be, in order to estimate the response level that should be present when the blood vessels are normal.

[0046] In this embodiment of the invention, the KM algorithm is used to achieve spatiotemporal weighted matching. Further, in some embodiments of the invention, spatiotemporal weighted matching is performed on the set of neural excitation points and the set of blood flow response points after amplification to obtain the degree of supply-demand mismatch, including: constructing a bipartite graph matching model, with neural excitation points in the set of neural excitation points as nodes on one side and blood flow response points in the set of blood flow response points as nodes on the other side; for any neural excitation point and a blood flow response point, the weight of the matching edge between them is determined based on the time interval between them and the relative difference in their amplified amplitudes; the KM algorithm is used to solve for the minimum sum of the weights of all paired edges, which is taken as the degree of supply-demand mismatch.

[0047] In this embodiment of the invention, the previously extracted set of neural excitation points and the set of blood flow response points after amplification by microvascular abnormality indicators are regarded as two independent event sets, and a bipartite graph matching model is constructed, with nodes on one side representing neural excitation points and nodes on the other side representing blood flow response points. For any pair of nodes (i.e., a neural event and a blood flow event), the weight of its matching edge is determined by both factors. Finally, the KM algorithm (Kuhn-Munkres algorithm) is used to solve for the minimum weight perfect matching of the bipartite graph, and its total weight serves as the quantitative output of the degree of supply-demand mismatch.

[0048] In this embodiment of the invention, minimum cost matching is achieved by calculating edge weights. The calculation method for edge weights is as follows: Furthermore, if the time interval between the nerve excitation point and the blood flow response point is less than the preset physiological response interval, a first interval weight is assigned; otherwise, a second interval weight is assigned. The first interval weight is less than the second interval weight. If the blood flow response point lags behind the nerve excitation point, a first lag weight is assigned; otherwise, a second lag weight is assigned. The first lag weight is less than the second lag weight. The absolute value of the difference in amplitude between the nerve excitation point and the blood flow response point after gain is calculated, and the absolute value of the difference is normalized to obtain the amplitude difference cost. Combining the interval weight, lag weight, and amplitude difference cost, the edge weights between the nerve excitation point and the blood flow response point are determined.

[0049] In this embodiment of the invention, due to damage to the neurovascular unit, the lag time of blood flow response varies significantly among individuals (typically fluctuating within 1–6 seconds). Fixed-delay models are prone to misclassifying reasonable lags as desynchronization. Furthermore, existing methods do not explicitly model the fundamental physiological causal law that "neural activity must precede blood flow response," leading to inverse pairing interference. Additionally, amplitude matching is often simplified, failing to reflect the effective supply-demand balance after microvascular amplification. To address these issues, this step proposes a multi-dimensional, hierarchically weighted edge weight construction mechanism. By explicitly encoding temporal rationality, causal directionality, and amplitude consistency, it achieves a refined assessment of the quality of neuro-blood flow event pairs.

[0050] The interval weight is used to determine whether the time delay is within a physiologically reasonable range. If the time interval between two points is less than the preset physiological response upper limit (e.g., 5 seconds), a lower first interval weight (e.g., 0.2) is assigned; otherwise, it is considered an out-of-window anomaly and a higher second interval weight (e.g., 1.0) is assigned. The lag weight is used to enforce the causal order of nerve-blood flow. If the blood flow response point timestamp is strictly greater than the nerve excitation point (i.e., blood flow lags behind nerve), a low first lag weight (e.g., 0.1) is assigned; if blood flow is earlier than or equal to nerve (violating physiology), a high second lag weight (e.g., 2.0) is assigned. The amplitude difference cost is used to measure the degree of matching between supply and demand intensity after gain. The absolute difference between the gain amplitudes of two events is calculated and linearly normalized to the [0,1] interval to obtain the amplitude difference cost. The higher the amplitude difference cost, the more uncoordinated the neural drive and blood flow response are, and the worse the consistency.

[0051] Ultimately, the edge weights are determined by a weighted fusion of the three factors (e.g., simple averaging). After obtaining the edge weights, the KM algorithm is used to identify the globally optimal event correspondence from a large number of potential matches, so as to objectively reflect the true state of neurovascular coupling.

[0052] The set of neural excitation points and the amplified blood flow response points are each considered as a subset of two vertices in a bipartite graph. Since the number of events in the two classes is usually unequal (e.g., a language task might induce multiple gamma wave bursts but only one slow blood flow rise), the system transforms the non-matrix matching problem into a matrix problem by introducing dummy nodes: if the number of neural excitation points is m and the number of blood flow response points is n, and m > n, then mn dummy blood flow points are added to the blood flow side, and vice versa. The edge weight between each dummy node and a real node is set to a high penalty value (e.g., 5.0), much higher than the reasonable weight of any real event pair, ensuring that the algorithm only matches to dummy nodes when there is no better choice, thus including unpaired real events in the total mismatch cost.

[0053] Subsequently, the KM algorithm operates on this expanded matrix, using steps such as alternating labeling and augmented path search to find the perfect matching scheme that minimizes the sum of the weights of all matching edges. This minimum weight sum is defined as the degree of supply-demand mismatch; the smaller the value, the more coordinated the neural activity and blood flow response are in the time, causal, and amplitude dimensions, and the tighter the coupling. The larger the value, the more likely there are numerous delayed windows, causal inversions, amplitude imbalances, or completely isolated events, reflecting significant neurovascular decoupling.

[0054] The operation adjustment module 104 is used to determine the cumulative anomaly coefficient of global variables based on the degree of supply and demand mismatch under different evaluation periods; and to generate recommended parameters for the runtime of TBS equipment based on the cumulative anomaly coefficient.

[0055] Since microcirculatory failure is usually the result of continuous accumulation of metabolic stress, a risk score at a single moment is insufficient as a basis for circuit breaker. The system employs a monotonically increasing time integral strategy to track the total risk throughout the entire process. That is, data acquisition at different times is taken as an assessment cycle. Then, the degree of supply-demand mismatch in each assessment cycle is calculated, and the degree of supply-demand mismatch in all assessments is cumulatively analyzed to obtain the cumulative anomaly coefficient.

[0056] Furthermore, the sum of the supply-demand mismatches across all assessment periods is calculated as the cumulative anomaly coefficient. The cumulative anomaly coefficient is a monotonically decreasing value that records in real time the total amount of uncompensated metabolic debt borne by the brain's microcirculation from the start to the end of the test. The larger the value, the closer the subject is to the critical point of microcirculatory dysfunction (such as latent hypoxia or vasospasm).

[0057] Therefore, in this embodiment of the invention, the recommended runtime parameters for the TBS device are generated based on the cumulative anomaly coefficient. That is, the larger the value of the cumulative anomaly coefficient, the higher the degree of anomaly, and the larger the corresponding recommended runtime parameter value.

[0058] In this embodiment of the invention, a dual control threshold can be implemented, that is, a preset maximum risk index for the test subject can be determined. This indicator is the maximum value of the cumulative abnormal coefficient of an individual subject obtained based on prior experience. It is a statistical upper limit determined based on the average cumulative abnormal coefficient of patients with microcirculatory dysfunction in historical clinical data, such as 20. Then, two control thresholds are determined; the threshold is optimized. Circuit breaker threshold: It also implements hierarchical control based on optimized thresholds and circuit breaker thresholds.

[0059] If the cumulative outlier coefficient The system determines that it is within a safe range and uses the recommended runtime parameter as the standard recommended runtime (e.g., 8-second pulse interval). The system determined that the cumulative risk was high and sent a parameter optimization command, generating a recommended runtime parameter of 12 seconds. This extended interval allows for longer recovery time for microvessels, thus mitigating the risk. The growth rate; if The system determines that the risk has overflowed, immediately sends a safety circuit breaker command, forcibly interrupts the TBS pulse output, and sends a microcirculation metabolism abnormality alarm to the operation terminal.

[0060] This invention obtains microvascular abnormality indicators by acquiring the concentrations of vascular endothelial growth factor and hypoxia-inducible factor and performing nonlinear weighted analysis; simultaneously extracts EEG frequency band energy and near-infrared oxyhemoglobin concentration, and determines the neural excitation point set and blood flow response point set through maximum search; weights the blood flow response point set based on the microvascular abnormality indicators, and perform spatiotemporal weighted matching between the amplified blood flow response point set and the neural excitation point set to obtain the degree of supply-demand mismatch; and determines the cumulative abnormality coefficient based on the degree of supply-demand mismatch under different evaluation periods and adjusts the TBS operation accordingly. This invention abandons traditional linear correlation or synchronicity analysis methods that rely on strict signal time alignment, and instead adopts a spatiotemporal weighted matching mechanism based on discrete event point sets. This mechanism can effectively tolerate the inherent, non-rigid physiological time drift in the neurovascular response of stroke patients, avoiding the misjudgment of reasonable lag as functional decoupling. At the same time, the weighted gain of microvascular abnormality indicators eliminates the interference of individual vascular state differences on blood flow response, so that the degree of supply-demand mismatch truly reflects the efficacy of neuromodulation, thereby achieving accurate abnormality assessment and significantly improving the reliability and robustness of TBS operation.

[0061] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0062] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A stroke motor aphasia assessment system combined with TBS, characterized in that, The system includes: The vascular abnormality analysis module is used to obtain the concentration values ​​of vascular endothelial growth factor and hypoxia-inducible factor, respectively; and to perform nonlinear weighted analysis by combining the two concentration values ​​to obtain microvascular abnormality indicators. The point set acquisition module is used to simultaneously acquire the frequency band energy of EEG signals and the oxygenated hemoglobin concentration of near-infrared brain functional imaging; to perform a maximum search on the frequency band energy to determine the neural excitation point set; and to perform a maximum search on the oxygenated hemoglobin concentration to determine the blood flow response point set. The mismatch analysis module is used to perform weighted gain on the blood flow response point set based on microvascular abnormality indicators, and to perform spatiotemporal weighted matching on the neural excitation point set and the gained blood flow response point set to obtain the degree of supply-demand mismatch. The operation adjustment module is used to determine the cumulative anomaly coefficient of global variables based on the degree of supply and demand mismatch under different evaluation periods; and to generate recommended parameters for the runtime of TBS equipment based on the cumulative anomaly coefficient.

2. The stroke motor aphasia assessment system combined with TBS as described in claim 1, characterized in that, The concentration values, the frequency band energy, and the oxygenated hemoglobin concentration are all standardized data.

3. The stroke motor aphasia assessment system combined with TBS as described in claim 1, characterized in that, The nonlinear weighted analysis combining the two concentration values ​​yields microvascular abnormality indicators, including: The concentration values ​​of vascular endothelial growth factor and hypoxia-inducible factor are respectively input into a pre-trained nonlinear mapping model; the output is a microvascular abnormality index characterizing the degree of microcirculation perfusion restriction.

4. The stroke motor aphasia assessment system combined with TBS as described in claim 1, characterized in that, The process of performing a maximum search on the frequency band energy to determine the set of neural excitation points includes: Local peak points of frequency band energy time series are detected within a preset time window, and local peak points exceeding a preset energy threshold are taken as neural excitation points; neural excitation points form a neural excitation point set.

5. The stroke motor aphasia assessment system combined with TBS as described in claim 4, characterized in that, The process of maximizing the concentration of oxyhemoglobin to determine the blood flow response point set includes: Within a preset time window corresponding to the nerve excitation point, the local maximum value of oxygenated hemoglobin concentration is extracted as the blood flow response point; the blood flow response points form a blood flow response point set.

6. The stroke motor aphasia assessment system combined with TBS as described in claim 1, characterized in that, The weighting and gaining of the blood flow response point set based on microvascular abnormality indicators includes: The oxygenated hemoglobin concentration amplitude at each blood flow response point was compensated by using microvascular abnormality indicators to obtain the amplified oxygenated hemoglobin concentration.

7. The stroke motor aphasia assessment system combined with TBS as described in claim 1, characterized in that, Spatiotemporal weighted matching based on the KM algorithm.

8. The stroke motor aphasia assessment system combined with TBS as described in claim 7, characterized in that, The spatiotemporal weighted matching of the neural excitation point set and the amplified blood flow response point set to obtain the degree of supply-demand mismatch includes: Construct a bipartite graph matching model, with the neural excitation points in the neural excitation point set as nodes on one side and the blood flow response points in the amplified blood flow response point set as nodes on the other side. For any neural excitation point and a blood flow response point, the weight of the matching edge between them is determined based on the time interval between them and the relative difference in their amplitudes after gain. The KM algorithm is used to find the minimum sum of weights of all paired edges, which is used as the degree of supply-demand mismatch.

9. The stroke motor aphasia assessment system combined with TBS as described in claim 8, characterized in that, The determination of the weight of the matching edge between the two components based on the time interval between them and the relative difference in their post-gain amplitudes includes: If the time interval between the nerve excitation point and the blood flow response point is less than the preset physiological response interval, then the first interval weight is assigned; otherwise, the second interval weight is assigned. The first interval weight is less than the second interval weight. When the blood flow response point lags behind the nerve excitation point, a first lag weight is assigned; otherwise, a second lag weight is assigned. The first lag weight is less than the second lag weight. Calculate the absolute value of the difference between the amplitudes after gaining the nerve excitation point and the blood flow response point, and normalize the absolute value of the difference to obtain the amplitude difference cost. By combining interval weight, lag weight, and amplitude difference cost, the edge weights between the nerve excitation point and the blood flow response point are determined.

10. The stroke motor aphasia assessment system combined with TBS as described in claim 1, characterized in that, The determination of the cumulative anomaly coefficient of the global variable based on the degree of supply-demand mismatch under different assessment periods includes: Calculate the sum of the supply-demand mismatch levels across all assessment periods as the cumulative anomaly coefficient.

Images