Hypertensive brain injury early monitoring and evaluation system and method based on multi-mode MRI data

Abstract

The invention relates to the technical field of intelligent monitoring medical treatment of senile chronic diseases, and discloses a hypertensive brain injury early-stage monitoring and evaluation system and method based on multi-mode MRI data, and the system comprises a data collection module, a data preprocessing module, a virtual elastography analysis module, a microcirculation analysis module, a feature extraction and prediction model module, and a data visualization and report module. The method further comprises the steps of multi-modal MRI data collection, data processing, virtual elastic tensor calculation, lymphatic-like system function analysis, image feature extraction, prediction model construction, quantitative report generation and personalized suggestion providing. According to the method, virtual elastography analysis and microcirculation analysis technologies are combined, multi-dimensional data such as cerebrovascular elasticity modulus, blood flow velocity and lymphatic-like system function indexes are integrated, more comprehensive hypertension cerebrovascular injury assessment is provided, and in addition, the accuracy of hypertension cerebrovascular injury assessment is improved by combining a machine learning algorithm and radiomics characteristics. And a high-precision cerebrovascular injury prediction model is constructed.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medical monitoring technology for chronic diseases in the elderly, and specifically to an early monitoring and evaluation system and method for hypertensive brain damage using multimodal MRI data. Background Art

[0002] Hypertension, a major chronic disease worldwide, has become a major cause of premature death and decreased quality of life. According to the World Health Organization, cardiovascular and cerebrovascular diseases caused by hypertension are the leading cause of death worldwide, with the toll particularly pronounced among the elderly. Hypertension not only leads to diseases such as arteriosclerosis and cerebral atherosclerosis, but also accelerates the aging and functional degeneration of cerebral blood vessels, causing long-term damage to the structure and function of brain tissue.

[0003] Hypertension increases the risk of stroke, cognitive impairment, and other conditions by directly altering cerebral vascular structure and function. Specifically, hypertension thickens and increases the fragility of cerebral vascular walls, altering cerebral hemodynamics and making blood vessels more susceptible to rupture or blockage, ultimately leading to stroke and brain dysfunction. Furthermore, hypertension damages the microcirculatory system, weakening microvascular function in the brain and causing localized oxygen and nutrient depletion in the brain, which in turn triggers neuronal damage and brain cell death.

[0004] Stroke, one of the most serious consequences of hypertension, has become a leading cause of death and disability worldwide. According to the "China Stroke Prevention and Control Report 2021," stroke is the leading cause of death and disability in both urban and rural areas of my country, with cerebral infarction accounting for over 80% of these cases. Stroke often leads to the partial death of brain cells, which in turn causes loss of nervous system function and, in severe cases, can be life-threatening.

[0005] Cerebrovascular damage caused by hypertension is often a long-term, cumulative process, with its pathological changes often being relatively subtle in the early stages. Traditional imaging techniques struggle to detect subtle changes in cerebral vessels at this early stage. While existing imaging techniques, such as conventional MRI and CT, can effectively demonstrate morphological changes in cerebral vessels, they are significantly limited in assessing functional indicators such as cerebral vascular elasticity and hemodynamic changes. This makes it often difficult to diagnose cerebral vascular damage in hypertensive patients early on, missing the optimal opportunity for treatment. Summary of the Invention

[0006] In response to the deficiencies in the prior art, the present invention provides a system and method for early monitoring and evaluation of hypertensive brain damage using multimodal MRI data, which solves the problem of insufficient early diagnosis of hypertensive brain damage in the prior art.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: an early monitoring and evaluation system for hypertensive brain damage based on multimodal MRI data, comprising: A data acquisition module, which is used to acquire multi-modal image data of hypertensive patients through an MRI device, and the data includes DWI-VEMRI sequences and DTI-ALPS sequences; A data preprocessing module, which is used to perform denoising, registration, and normalization processing on the acquired DWI-VEMRI and DTI-ALPS image data to generate standardized high-quality image data; A virtual elastography analysis module, which is used to calculate the virtual elastic tensor of cerebral blood vessels based on DWI-VEMRI data, and obtain the elastic modulus of cerebral blood vessels through this tensor, and the elastic modulus is used to evaluate the hardness or elastic properties of cerebral blood vessels; A microcirculation analysis module, which is used to perform fiber bundle tracking analysis through DTI-ALPS technology, calculate the glymphatic system function index, and evaluate the changes in cerebral microcirculation caused by hypertension; A feature extraction and prediction model module, which is used to extract radiomics features from the DWI-VEMRI and DTI-ALPS image data, and combine machine learning algorithms to construct a prediction model for cerebral blood vessel injury related to hypertension; A data visualization and reporting module, which is used to generate a quantitative report of cerebral blood vessel injury and display the analysis results of the prediction model in the form of charts.

[0008] Preferably, the virtual elastography analysis module includes: A DWI data acquisition unit, which is used to acquire high-resolution DWI-VEMRI image data and collect diffusion-weighted image data in different directions; An elastic tensor calculation unit, which is based on the acquired DWI data and uses a diffusion tensor imaging model to calculate the virtual elastic tensor of the vascular region; An elastic modulus calculation unit, which calculates the elastic modulus of cerebral blood vessels according to the obtained virtual elastic tensor, so as to evaluate the hardness or elastic properties of cerebral blood vessels; A blood flow velocity calculation unit, which calculates the cerebral blood flow velocity through a virtual elastic tensor model and estimates the flow velocity of blood in cerebral blood vessels; A cerebral blood vessel perfusion analysis unit, which calculates the perfusion index of cerebral blood vessels and provides a comprehensive evaluation of cerebral blood vessel function by comparing and analyzing the differences between the healthy group and hypertensive patients.

[0009] Preferably, the formula for calculating the virtual elastic tensor of the vascular region in the elastic modulus calculation unit is as follows: Where E represents the elastic modulus of cerebral blood vessels, D xx 、D yy 、D zz are the three main components of the diffusion tensor respectively.

[0010] Preferably, the microcirculation analysis module includes: A water molecule diffusion tracking unit, which tracks the diffusion direction of water molecules in the brain through DTI-ALPS technology, obtains the diffusion trajectory of water molecules along the fiber bundle, and analyzes the structural connectivity of white matter fibers in the brain; A glymphatic system function analysis unit, which calculates the function index of the glymphatic system based on the water molecule diffusion tracking data; A microcirculation evaluation unit, which evaluates the cerebrovascular microcirculation function of hypertensive patients according to the obtained glymphatic system function index and compares it with that of the healthy control group.

[0011] Preferably, the formula for calculating the function index of the glymphatic system in the glymphatic system function analysis unit is as follows: where C ij is the connectivity between the i-th fiber bundle and the j-th fiber bundle, N is the total number of fiber bundles, and L represents the function index of the glymphatic system.

[0012] Preferably, the feature extraction and prediction model module includes: An image feature extraction unit, which extracts relevant radiomics features from DWI-VEMRI and DTI-ALPS image data, such as texture features, morphological features, vascular elasticity, blood flow velocity, and glymphatic system function index; A machine learning model construction unit, which uses support vector machine, random forest or deep learning algorithm to construct a prediction model based on radiomics features and predicts the risk level of cerebrovascular injury; A personalized treatment recommendation generation unit, which provides personalized treatment recommendations for patients according to the prediction results combined with the clinical data of the patients.

[0013] Preferably, the data visualization and report module includes: A quantitative analysis result display unit, which integrates the quantitative results of the virtual elastography analysis module and the microcirculation analysis module to generate quantitative data on cerebrovascular injury; A prediction result visualization unit, which displays the analysis results of the prediction model in the form of charts to help doctors intuitively understand the health status of patients; A customized report generation unit, which generates a personalized report for patients, including specific indicators of cerebrovascular injury, prediction results, recommended treatment plans, and follow-up monitoring plans.

[0014] The present invention also provides an early monitoring and evaluation method for hypertensive brain injury of multimodal MRI data, which is applied to the above-mentioned early monitoring and evaluation system for hypertensive brain injury of multimodal MRI data, and includes the following steps: S1. Collect multi-modal MRI data of hypertensive patients, including DWI-VEMRI sequences and DTI-ALPS sequences; S2. Denoise, register, and normalize the collected image data to generate high-quality normalized image data; S3. Calculate the virtual elastic tensor based on DWI-VEMRI data, and then calculate the elastic modulus, blood flow velocity, perfusion index of cerebral blood vessels, and conduct quantitative analysis of cerebral vascular injury; S4. Use DTI-ALPS data for fiber tractography, analyze the diffusion direction of water molecules in the brain, and calculate the glymphatic system function index; S5. Extract radiomics features from DWI-VEMRI and DTI-ALPS data, and construct a prediction model for hypertensive-related cerebral vascular injury through machine learning algorithms; S6. Generate a quantitative report, display the analysis results of cerebral vascular injury, and provide personalized suggestions based on the prediction model.

[0015] The present invention provides an early monitoring and evaluation system and method for hypertensive brain injury using multi-modal MRI data. It has the following Beneficial effects: 1. The present invention adopts a technical solution that combines virtual elastography analysis and microcirculation analysis, achieving precise evaluation of hypertensive cerebral vascular injury. Compared with the existing diagnostic methods that solely rely on imaging indicators, the present invention combines virtual elastography analysis and microcirculation analysis, integrating multi-dimensional data such as the elastic modulus of cerebral blood vessels, blood flow velocity, and glymphatic system function index, providing a more comprehensive assessment of cerebral vascular health. Specifically, virtual elastography analysis can quantify the elastic properties of blood vessels, reflecting whether cerebral blood vessels have sclerosis or loss of elasticity, while microcirculation analysis focuses on the health of cerebral microvessels and can detect early vascular function damage. By combining these two, the system can early detect vascular structural and functional abnormalities caused by hypertension, especially microcirculation problems that are difficult to capture by traditional imaging techniques. In addition, by combining these quantitative data with the prediction model, the present invention can provide accurate disease warnings and treatment plans for clinicians. Therefore, compared with the existing method of only relying on a single imaging sequence for evaluation, the present invention not only solves the deficiency of being unable to comprehensively evaluate cerebral vascular injury but also can provide more accurate and early health interventions.

[0016] 2. The present invention combines machine learning algorithms with radiomics features to construct a high-precision prediction model for cerebrovascular injury, which can accurately predict the risk level of patients. The present invention extracts radiomics features from DWI-VEMRI and DTI-ALPS imaging data, and combines machine learning algorithms to construct a high-precision prediction model. These radiomics features include not only the geometric structure features of blood vessels, but also multi-dimensional information such as blood vessel elasticity, blood flow velocity, and microcirculation index. These features can comprehensively reflect the functional and structural state of cerebrovascular vessels. Through the training of the machine learning model, the present invention can accurately evaluate the risk of cerebrovascular injury of patients based on these features. Compared with the traditional method that relies on the subjective judgment of doctors in the prior art, the present invention can provide a more objective and accurate risk prediction, reducing the errors and biases of human judgment. In addition, the machine learning algorithm can mine potential laws from a large amount of data, making the prediction model have strong generalization ability and can adapt to the personalized data of different patients, thereby further improving the accuracy and reliability of the prediction. This intelligent prediction model has important value in the early warning of cerebrovascular injury, solves the deficiency that multi-dimensional feature integration is often ignored in traditional evaluation methods, and improves the efficiency and accuracy of cerebrovascular health management. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is the system framework diagram of the present invention; Figure 2 is the framework diagram of the virtual elastography analysis module of the present invention; Figure 3 is the framework diagram of the microcirculation analysis module of the present invention; Figure 4 is the framework diagram of the feature extraction and prediction model module of the present invention; Figure 5 is the framework diagram of the data visualization and reporting module of the present invention; Figure 6 is the flow chart of the method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0019] Please refer to the attached Figure 1 , the embodiments of the present invention provide an early monitoring and evaluation system for hypertensive brain injury with multi-modal MRI data, including: Data acquisition module, used to acquire multi-modal imaging data of hypertensive patients through an MRI device, where the data includes DWI-VEMRI sequences and DTI-ALPS sequences; The data acquisition module aims to provide high-quality and accurate cerebrovascular imaging data for the subsequent analysis module. This module acquires multi-modal imaging data through a high-resolution MRI device, especially for quantitative monitoring of cerebrovascular damage in hypertensive patients. This module is mainly responsible for obtaining DWI-VEMRI (Diffusion-Weighted Imaging-Virtual Elastography) sequence and DTI-ALPS (Diffusion Tensor Imaging-Lymphatic System Functional Imaging) sequence data. These data provide the necessary inputs for the subsequent virtual elastography analysis and microcirculation analysis modules, ensuring the accuracy and reliability of the assessment of hypertensive cerebrovascular damage. The implementation of the data acquisition module ensures the quality of the imaging data so that subsequent steps can effectively perform virtual elastic tensor calculation, lymphatic system function index analysis, etc.

[0020] In this embodiment, the workflow of the data acquisition module includes the following main links to ensure high-quality MRI data for the subsequent analysis module.

[0021] Imaging data acquisition: The data acquisition module acquires brain imaging data of hypertensive patients through an MRI device, using DWI-VEMRI (Diffusion-Weighted Imaging-Virtual Elastography) sequence and DTI-ALPS (Diffusion Tensor Imaging-Lymphatic System Functional Imaging) sequence. These two sequences provide different imaging information to ensure a comprehensive assessment of cerebrovascular damage and microcirculation function.

[0022] DWI-VEMRI sequence: Generally, the DWI-VEMRI sequence is used to obtain elastic modulus data of cerebrovascular through sensitive imaging of water molecule diffusion. This sequence scans the elastic characteristics of intracranial blood vessels in detail, and through multi-directional diffusion-weighted imaging, obtains the diffusion characteristics of the cerebrovascular region in each direction. The data is finely recorded through magnetic resonance imaging technology, thus providing the required diffusion coefficient data for the subsequent calculation of virtual elastic tensor.

[0023] Diffusion tensor calculation: During the data acquisition process, the diffusion tensor \(\mathbf{D}\) generated by the DWI-VEMRI sequence is the basis for subsequent elastic modulus calculation. The calculation of this tensor depends on diffusion-weighted imaging data, which contains diffusion coefficients in three-dimensional directions, as follows: where \(D\) xx , \(D\) yy , \(D\) zz respectively represent the diffusion coefficients of cerebrovascular in the X, Y, and Z axis directions, \(D\) xy , \(D\) xz , \(D\)yz It is the off-diagonal term that describes the diffusion interaction of blood vessels in different directions.

[0024] Elastic modulus calculation: After the diffusion tensor data is collected, the calculation of the virtual elastic tensor is used to evaluate the elasticity or stiffness of the cerebral blood vessels. The elastic modulus is calculated by the following formula: where E represents the elastic modulus of the cerebral blood vessels, which is used to quantify the elastic properties of the cerebral blood vessels and reflects the response ability of the cerebral blood vessels to external forces, D xx 、D yy 、D zz are the three main components of the diffusion tensor respectively.

[0025] DTI-ALPS sequence: In one possible implementation, the DTI-ALPS sequence is mainly used to analyze the microcirculation in the brain and the connectivity of fiber bundles, providing a quantitative analysis of the glymphatic system function. The DTI-ALPS sequence evaluates the structural connectivity of white matter fibers in the brain through the directionality of water molecule diffusion, helping to analyze the microcirculation function of the cerebral vascular system.

[0026] Glymphatic system function index calculation: During the data acquisition process, the DTI-ALPS sequence is used to calculate the glymphatic system function index L, which represents the connectivity of the cerebral vascular network. The formula is: where C ij is the connectivity between the i-th fiber bundle and the j-th fiber bundle, N is the total number of fiber bundles, and L represents the glymphatic system function index. This index can quantitatively evaluate the microcirculation function of the cerebral blood vessels and provide necessary data support for subsequent analysis.

[0027] Data acquisition accuracy and quality control: To ensure the high quality and high accuracy of the image data, the data acquisition module has implemented various quality control measures during the implementation process. For example, for the influence of noise, denoising techniques are used to optimize the image quality, and at the same time, the imaging contrast and resolution are improved by adjusting the scanning parameters (such as the b value of the diffusion-weighted sequence, the scanning direction, etc.).

[0028] Specifically, the DWI-VEMRI sequence adjusts the diffusion sensitivity by setting different b values, thereby optimizing the imaging effect of the cerebral vascular region. Generally, the selection of the b value needs to be adjusted according to the experimental requirements and the performance of the equipment to ensure the best diffusion signal.

[0029] Optimizing Scanning Parameters: By adjusting the scanning parameters, the data acquisition module can adapt to the body shapes and conditions of different patients. For example, the scanning time, slice thickness, and signal reception settings may be optimized to balance image quality and scanning time. On this basis, the selection of the scanning direction and imaging sequence will ensure the capture of optimal information in the vascular region.

[0030] Data Processing and Storage: After data acquisition is completed, all the acquired imaging data will be preprocessed, including denoising, registration, and normalization. These processes ensure that data from different sources and time points can be compared with each other, avoiding the impacts caused by equipment differences or operation errors. The processed data will be stored in the database and provide available data for the virtual elastography analysis and microcirculation analysis modules.

[0031] Data Storage and Sharing: During the storage process, all data will be saved in a standardized medical imaging format (such as DICOM format), and patient privacy will be protected through encryption. The data storage system supports fast retrieval and sharing, facilitating the invocation by subsequent analysis modules.

[0032] Through the data acquisition module in this embodiment, high-quality multi-modal MRI data acquisition can be achieved, providing accurate data support for subsequent virtual elastography analysis, microcirculation analysis, and cerebrovascular injury assessment. This module can acquire comprehensive data including cerebrovascular elasticity, microcirculation, and fiber bundle structure, ensuring the high precision and high reliability of the entire system and providing a strong data foundation for the early diagnosis of hypertensive cerebrovascular injury.

[0033] A data preprocessing module, which is used to perform denoising, registration, and normalization on the acquired DWI-VEMRI and DTI-ALPS imaging data to generate standardized high-quality imaging data; The data preprocessing module is an important part of the present invention and is responsible for performing processing operations such as denoising, registration, and normalization on the imaging data obtained from the data acquisition module (including DWI-VEMRI and DTI-ALPS sequences). Since the imaging data may be affected by various factors (such as equipment noise, patient movement, scanning parameter differences, etc.), these processing steps can ensure that subsequent analysis modules (such as the virtual elastography analysis and microcirculation analysis modules) can process high-quality and standardized data. The accuracy of data preprocessing is crucial for ensuring the accuracy of the entire system. Through technical means such as denoising, registration, normalization, and reconstruction, this module can provide reliable data input for subsequent steps, ensuring the accuracy of virtual elastic modulus calculation and glymphatic system function index analysis.

[0034] In this embodiment, the workflow of the data preprocessing module includes the following main steps to ensure the quality and consistency of the imaging data and provide high-quality and standardized inputs for subsequent virtual elastography analysis and microcirculation analysis.

[0035] Denoising processing: Generally, DWI-VEMRI and DTI-ALPS imaging data are affected by noise to varying degrees, and noise may lead to subsequent analysis errors. Therefore, denoising processing is one of the key steps in the data preprocessing module.

[0036] Denoising method selection: In this embodiment, the denoising methods selected are the non-local means (NLM) denoising algorithm and the wavelet transform denoising algorithm. The non-local means method can better preserve the details of the image, especially without losing edge information while removing image noise. The wavelet transform can effectively remove the high-frequency noise in the image without affecting the low-frequency information of the image (such as the structural information of cerebral blood vessels).

[0037] Denoising technology implementation: In the specific denoising steps, for DWI-VEMRI data, non-local means denoising is adopted. First, local regions in the image are selected, and noise is removed by calculating the average value of similar regions. For DTI-ALPS data, wavelet transform denoising is used to perform multi-scale analysis on the original data, extract low-frequency information, and remove high-frequency noise.

[0038] Image registration: Since the imaging data may be from different time points or different scanning sequences, image registration is an important step to ensure the accurate alignment of different image data in the same spatial coordinate system. The purpose of registration is to correct the spatial deviation between different images to be consistent for subsequent analysis.

[0039] Registration method selection: Generally, algorithms based on rigid or non-rigid registration are used for image registration. In DWI-VEMRI and DTI-ALPS data, non-rigid registration methods can usually better handle the deformation of different tissue regions, especially when there are patient movements or anatomical structure differences. Specifically, this module uses a registration algorithm based on the maximization of mutual information (MI), which can effectively align different imaging sequences, especially in the case of different scanning parameters (such as b-values).

[0040] Registration steps: Initial registration: Perform initial rigid registration on different images to ensure the position correction of image rotation, translation, etc.

[0041] Non-rigid registration: Perform non-rigid registration on the images that have undergone initial registration to further eliminate the distortion caused by scanning angles and individual differences by minimizing mutual information or maximizing the correlation degree.

[0042] Normalization: Data normalization is a crucial step in ensuring the comparability of different image data in subsequent analyses. Due to differences among different patients, scanning devices, and sequence settings, the normalization of imaging data can eliminate the influence of these factors, thus providing data on a unified scale.

[0043] Selection of normalization methods: In this embodiment, two normalization methods are adopted: min-max normalization and z-score normalization.

[0044] Min-max normalization: This method is calculated using the following formula: where X is the gray value of the original image, X min and X max are the minimum and maximum gray values in the image data respectively, and X norm is the normalized gray value. This method compresses the gray values of the image data into the range of 0 to 1.

[0045] Z-score normalization: This method normalizes according to the mean and standard deviation of the image data, and the formula is: where X is the gray value of the original image, μ is the mean of the image, σ is the standard deviation of the image, and X norm is the value after normalization. Through this method, the mean of the image data is adjusted to 0 and the standard deviation is 1, ensuring that the data has a unified scale.

[0046] Reconstruction and interpolation: In some cases, due to factors such as scanning device or patient movement, the imaging data may be lost or distorted. To fill in the missing areas or repair the distortion, this module also includes reconstruction and interpolation steps.

[0047] Selection of interpolation methods: In this embodiment, bilinear interpolation and cubic interpolation are adopted for image reconstruction.

[0048] Bilinear interpolation: Pixel value estimation is performed on a two-dimensional plane, and the value of the unknown pixel is deduced through the weighted average of the surrounding four pixel points.

[0049] Cubic interpolation: This method estimates the missing pixels through a higher-order polynomial. Compared with bilinear interpolation, cubic interpolation can better preserve the details and edge information of the image.

[0050] Data quality assessment: To ensure the quality of the data, the preprocessing module also includes a data quality assessment step. By calculating the signal-to-noise ratio (SNR) and contrast sensitivity (CNR) of the image, the image quality is evaluated to ensure that the data can meet the analysis requirements.

[0051] SNR Calculation: Signal-to-noise ratio is an important indicator for evaluating image quality, defined as the ratio of signal intensity to noise intensity. A high SNR value indicates less noise, stronger signal, and higher data quality in the image.

[0052] CNR Calculation: Contrast sensitivity is used to evaluate the contrast between different regions in an image. In medical imaging, a high CNR value usually indicates a higher contrast between the lesion area and normal tissues in the image, enabling better highlighting of the lesion area.

[0053] The data preprocessing module ensures that the DWI-VEMRI and DTI-ALPS image data obtained from the data acquisition module have high quality, high consistency, and high reliability through processing such as denoising, registration, normalization, reconstruction, and data quality assessment. In this way, it provides reliable input data for the subsequent virtual elastography analysis and microcirculation analysis modules, ensuring the accuracy and stability of cerebrovascular injury assessment.

[0054] Please refer to the appendix Figure 2 , the virtual elastography analysis module, which is used to calculate the virtual elastic tensor of cerebrovascular vessels based on DWI-VEMRI data, and obtain the elastic modulus of cerebrovascular vessels through this tensor. The elastic modulus is used to evaluate the hardness or elastic properties of cerebrovascular vessels; The main task of the virtual elastography analysis module is to calculate the virtual elastic tensor of cerebrovascular vessels using DWI-VEMRI data and further derive the elastic modulus of cerebrovascular vessels based on this tensor. As an important biomechanical index, the elastic modulus is used to describe the hardness or elastic properties of cerebrovascular vessels, which is directly related to the health status and function of cerebrovascular vessels. The function of this module is not limited to the calculation of the elastic modulus, but also includes the evaluation of blood flow velocity and the analysis of cerebrovascular perfusion status. Through the extraction of these quantitative indicators, this module provides important technical support for the assessment of cerebrovascular injury, pathological analysis, and early intervention in hypertensive patients. The steps involved in the module, such as data acquisition, tensor calculation, and blood flow velocity estimation, ensure the accurate assessment of the health status of cerebrovascular vessels with the support of high-precision image data.

[0055] In this embodiment, the specific implementation of the virtual elastography analysis module is divided into the following sub-units.

[0056] Generally, DWI-VEMRI data is obtained through an MRI scanning device. Using the diffusion-weighted imaging (DWI) technique, diffusion-weighted images in different directions are acquired. This step is the basic data source for the entire virtual elastography analysis module, ensuring the quality and accuracy of the image data. Specifically, DWI-VEMRI data provides the diffusion coefficients of cerebrovascular vessels in different directions by collecting the diffusion information of the cerebrovascular region. These data are used for subsequent virtual elastic tensor calculation to ensure a comprehensive evaluation of the elastic properties of blood vessels in different directions.

[0057] In this step, the key of the data acquisition unit is the acquisition of high-resolution DWI-VEMRI images. Through multi-directional diffusion-weighted imaging, accurate diffusion signals are obtained, providing sufficient data support for subsequent virtual elastic tensor calculation.

[0058] After the imaging data acquisition is completed, the elastic tensor calculation unit calculates the virtual elastic tensor of the cerebral blood vessels based on the DWI-VEMRI data through the diffusion tensor imaging (DTI) model. The diffusion tensor is a mathematical model that describes the diffusion direction and rate of water molecules in the cerebral blood vessels and can reflect the elastic and stiffness characteristics of the cerebral blood vessels.

[0059] The calculation of the elastic tensor depends on the diffusion coefficients in the DWI-VEMRI data, and the formula is as follows: where D xx , D yy , D zz represent the diffusion coefficients of the cerebral blood vessels in the X, Y, and Z directions respectively, reflecting the diffusion ability of water molecules in these directions; D xy , D xz , D yz are the off-diagonal terms of the diffusion tensor, describing the interactive diffusion of water molecules between different directions. By modeling these diffusion coefficients, the virtual elastic tensor can be calculated, providing a basis for the calculation of the elastic modulus.

[0060] The elastic modulus calculation unit is the core of this module, mainly responsible for further obtaining the elastic modulus of the cerebral blood vessels based on the calculation results of the virtual elastic tensor. The elastic modulus, as an index reflecting the hardness of blood vessels, is usually used to evaluate the health status of blood vessels. Excessively high hardness is usually a sign of arteriosclerosis or hypertension. The calculation formula of the elastic modulus EEE is: where E represents the elastic modulus of the cerebral blood vessels, D xx , D yy , D zz are the three main components of the diffusion tensor, representing the diffusion coefficients of the blood vessels in the X, Y, and Z axis directions respectively. The square terms in the formula reflect the contributions of the diffusion coefficients in each direction to the elastic modulus. By weighting the contributions in these three directions, the comprehensive elastic modulus E is calculated to quantify the elastic characteristics of the cerebral blood vessels.

[0061] Specifically, the value of the elastic modulus can reflect the deformation ability of the cerebral blood vessels under external forces. A higher elastic modulus usually means that the blood vessels are relatively hard, which may lead to a decrease in blood vessel function and the occurrence of diseases.

[0062] Blood flow velocity is an important physiological parameter for evaluating cerebrovascular health. The blood flow velocity calculation unit estimates the blood flow velocity in the cerebrovascular by means of a virtual elastic tensor model. Specifically, the calculation formula for blood flow velocity v is as follows: where v is the blood flow velocity, Δx represents the displacement of blood within a unit of time, and Δt is the time interval. This calculation can provide quantitative data for blood flow velocity, thereby assisting in judging the functional status of the cerebrovascular.

[0063] In practical applications, the calculation of blood flow velocity not only depends on the elastic modulus of the cerebrovascular, but also takes into account the geometric shape of the blood vessels and the hemodynamic model. By estimating the blood flow velocity, the perfusion status and flow resistance of the blood vessels can be evaluated more comprehensively.

[0064] Cerebrovascular perfusion is one of the core indicators of cerebrovascular function, representing the blood flow through the cerebrovascular within a unit of time. The cerebrovascular perfusion analysis unit calculates the perfusion index of the cerebrovascular by the ratio of blood flow volume and blood vessel cross-sectional area.

[0065] The calculation formula for perfusion index C is: where Q is the blood flow volume through the blood vessel within a unit of time, and A is the cross-sectional area of the blood vessel. This index reflects the perfusion ability of the cerebrovascular. Through comparative analysis with the healthy group, the functional changes of the cerebrovascular in hypertensive patients can be evaluated.

[0066] During the analysis process, the cerebrovascular perfusion analysis unit also compares the perfusion data of hypertensive patients with the healthy control group. Through the differential perfusion index, the damage of blood vessel function and microcirculation disorder are further revealed.

[0067] The virtual elastography analysis module, through in-depth analysis of DWI-VEMRI data, can not only provide the elastic modulus of the cerebrovascular, but also evaluate physiological parameters such as cerebral blood flow velocity and perfusion status. Through the extraction of these quantitative indexes, this module can comprehensively reflect the cerebrovascular health status of hypertensive patients, providing a scientific basis for the early diagnosis, risk assessment and personalized treatment of cerebrovascular injury. This module can accurately quantify the hardness, elasticity and microcirculation status of the cerebrovascular, ensuring a comprehensive assessment of hypertension-related cerebrovascular injury.

[0068] Please refer to Appendix Figure 3 , the microcirculation analysis module, which is used to perform fiber tract tracing analysis through DTI-ALPS technology, calculate the glymphatic system function index, and evaluate the changes in cerebral microcirculation caused by hypertension; The microcirculation analysis module is an important part of the present invention for evaluating the changes in cerebral microcirculation caused by hypertension. Through the DTI-ALPS technology, the module can track the diffusion direction of water molecules in the brain and obtain the diffusion trajectories of water molecules along the cerebral fiber tracts. This process provides data support for the subsequent calculation of the glymphatic system function index. The glymphatic system function index (L), as the core index for evaluating the health of cerebral vascular microcirculation, can quantify the health status of cerebral microcirculation. By calculating and comparing the glymphatic system function index, this module can effectively reveal the impact of hypertension on cerebral vascular microcirculation and provide a scientific basis for the early diagnosis and treatment of cerebral vascular injury.

[0069] The functions of this module not only include water molecule diffusion tracking and the calculation of the glymphatic system function index, but also the evaluation of microcirculation function. Through these analyses, the module can achieve a comprehensive health assessment of the cerebral blood vessels of hypertensive patients, especially the changes in the microcirculation state, so as to provide valuable reference for clinical practice.

[0070] In this embodiment, the specific implementation of the microcirculation analysis module includes the following key units.

[0071] Generally, the DTI-ALPS technology accurately tracks the diffusion direction of water molecules through diffusion tensor imaging (DTI), so as to obtain the diffusion trajectories of water molecules along the cerebral fiber tracts. By analyzing these trajectories, the structural information of white matter fibers in the brain can be obtained, and then the microcirculation state of cerebral blood vessels can be revealed.

[0072] Specifically, this unit calculates the diffusion paths of water molecules in the brain tissue through the DTI-ALPS technology. These diffusion paths provide detailed information about the microvascular structure in the brain. The directionality and diffusion coefficient of diffusion provide the structural information of the vascular network for subsequent analysis.

[0073] In a possible implementation, the DTI-ALPS technology determines the diffusion trajectories of water molecules in the cerebral fiber tracts by analyzing the diffusion signals in different directions. This trajectory analysis helps to determine the connectivity of white matter fibers in the brain and provides data support for the calculation of the glymphatic system function index.

[0074] The function of the glymphatic system function analysis unit is to calculate the function index of the glymphatic system based on the water molecule diffusion tracking data. The glymphatic system function index (L) is an important index reflecting the health status of cerebral vascular microcirculation.

[0075] The calculation formula of the glymphatic system function index (L) is as follows: Where C ij$C_{ij}$ is the connectivity between the $i$-th fiber bundle and the $j$-th fiber bundle, $N$ is the total number of fiber bundles, and $L$ represents the glymphatic system function index.

[0076] Connectivity $C$ ij : Connectivity $C$ ij represents the structural connectivity between the $i$-th fiber bundle and the $j$-th fiber bundle. Specifically, this connectivity is calculated based on the similarity of diffusion signals. If the diffusion directions of two fiber bundles are relatively consistent, their connectivity is higher.

[0077] Total number of fiber bundles $N$: The total number of fiber bundles $N$ represents the number of all fiber bundles considered in the microcirculation analysis. The total number of fiber bundles is determined according to the number of fiber bundles traced by the DTI-ALPS technique.

[0078] The glymphatic system function index $L$ calculated by the above formula provides a quantitative index for the microcirculation function of cerebral blood vessels. A higher glymphatic system function index usually indicates a better health status of the cerebral microcirculation, while a lower $L$ value may indicate a decline in the function of cerebral blood vessels, possibly indicating blood flow obstruction or microvascular dysfunction.

[0079] The role of the microcirculation assessment unit is to evaluate the microcirculation function of the cerebral blood vessels in hypertensive patients based on the calculated glymphatic system function index ($L$). This unit reveals the impact of hypertension on the cerebral microcirculation by comparing the glymphatic system function index of hypertensive patients with that of the healthy control group.

[0080] In the microcirculation assessment unit, the glymphatic system function index ($L$) will be used to evaluate the microcirculation function of the cerebral blood vessels. Specifically, the system compares the $L$ value of hypertensive patients with that of the healthy control group to quantify the impact of hypertension on the cerebral microcirculation. A lower $L$ value indicates impaired microcirculation function, which may be a manifestation of cerebrovascular lesions caused by hypertension. In this way, the microcirculation assessment unit can not only reveal the microcirculation health status of a single patient but also provide a basis for early warning of cerebrovascular damage by comparing with the healthy control group.

[0081] The microcirculation analysis module can accurately evaluate the changes in cerebral microcirculation caused by hypertension through water molecule diffusion tracking of the DTI-ALPS technique and the calculation of the glymphatic system function index. By calculating the glymphatic system function index and comparing it with the healthy control group, this module can reveal the functional health status of the microcirculation.

[0082] The advantage of this module is that it can provide high-precision and quantitative microcirculation assessment data, providing important support for the early detection of cerebrovascular damage and the formulation of treatment plans for hypertensive patients. Through the quantitative glymphatic system function index, doctors can better understand the impact of hypertension on cerebrovascular microcirculation and provide strong decision-making basis for clinical treatment.

[0083] Please refer to the appendix Figure 4 , a feature extraction and prediction model module, which is used to extract radiomics features from the DWI-VEMRI and DTI-ALPS image data, and combine machine learning algorithms to construct a prediction model for hypertension-related cerebrovascular damage; the feature extraction and prediction model module is the core technical module for the assessment of hypertension-related cerebrovascular damage in the present invention. It extracts radiomics features from DWI-VEMRI and DTI-ALPS image data, and combines machine learning algorithms to establish a prediction model to evaluate the risk level of cerebrovascular damage. This module can extract multi-dimensional information such as texture features, morphological features, vascular elasticity, blood flow velocity, and glymphatic system function index from the image data to comprehensively evaluate the health status of cerebral blood vessels.

[0084] Through machine learning algorithms (such as support vector machines, random forests, deep learning), the module can train an efficient prediction model, predict the risk of cerebrovascular damage based on radiomics features, and provide personalized treatment suggestions for clinicians. This module provides technical support for the early diagnosis of cerebrovascular damage and the formulation of personalized treatment plans by predicting and evaluating the cerebrovascular health of hypertensive patients.

[0085] In this embodiment, the implementation of the feature extraction and prediction model module is divided into the following sub-units.

[0086] The main task of the image feature extraction unit is to extract radiomics features from the DWI-VEMRI and DTI-ALPS image data. These features can reflect the structure, elasticity, hemodynamics and microcirculation health status of cerebral blood vessels.

[0087] Texture features: Texture features are usually used to describe the gray-level distribution pattern in a local area of an image. In the cerebrovascular region, common texture features include contrast, entropy, uniformity, correlation, etc.

[0088] Contrast: It represents the degree of change in gray level and can reflect the structural changes of the blood vessel wall. Areas with higher contrast may indicate abnormal thickening of the blood vessel wall or blood vessel dilation.

[0089] Entropy: It describes the complexity or irregularity of an image and reflects the complexity of the blood vessel region. A higher entropy value usually indicates a more complex morphological structure of the blood vessels.

[0090] Homogeneity: Describes the uniformity of the gray-level distribution of an image and is usually used to reflect the uniformity of blood vessel structures.

[0091] Morphological features: Morphological features include the geometric shape of blood vessels, such as the length, diameter, and branching angle of blood vessels. These features can reflect the health status of blood vessels, especially during the process of vascular remodeling caused by hypertension.

[0092] Vascular elasticity and blood flow velocity: Vascular elasticity is calculated by the virtual elasticity imaging analysis module, and the blood flow velocity is provided by the virtual elastic tensor model. These features provide important quantitative data for predicting the risk of cerebrovascular injury.

[0093] Lymphatic system function index: The lymphatic system function index (L) reflects the health status of the microcirculation in the brain. After obtaining fiber tractography data through DTI-ALPS technology, the lymphatic system function index is calculated. This index reflects the microcirculation ability of blood vessels, and the calculation formula is: where C ij is the connectivity between the i-th fiber bundle and the j-th fiber bundle, and N is the total number of fiber bundles. Through this index, the microcirculation function of cerebral blood vessels can be quantified, and a lower L value may indicate microcirculation dysfunction.

[0094] Once the radiomics features are extracted, the machine learning model construction unit begins to construct a prediction model based on these features, with the goal of evaluating the risk level of cerebrovascular injury in hypertensive patients. Algorithm selection: In this embodiment, the machine learning model construction unit uses several mainstream machine learning algorithms, including: Support Vector Machine (SVM): SVM is an algorithm commonly used for binary classification that can find the optimal hyperplane by maximizing the classification margin. In the prediction of cerebrovascular injury, SVM can classify different risk levels based on image features.

[0095] Random Forest (RF): Random Forest is an ensemble learning method that makes predictions by constructing multiple decision trees and combining their outputs. RF performs well in dealing with high-dimensional features and noisy data and has wide applications in the classification tasks of medical image data.

[0096] Deep Learning (Convolutional Neural Network CNN): Deep learning methods, especially convolutional neural networks (CNN), automatically extract features from image data through multi-level learning networks. CNN has extremely strong performance in medical image analysis, can automatically learn the deep features of images, and perform efficient classification.

[0097] Training and Validation: During the training process of a machine learning model, cross-validation is used to evaluate the performance of the model and ensure that the model has good generalization ability. The cross-validation method divides the data into multiple subsets and conducts multiple trainings and evaluations through different combinations of training sets and validation sets.

[0098] Grid Search: Grid search is an optimization method used to adjust the hyperparameters of a model. By traversing different combinations of hyperparameters, grid search can find the model configuration that best suits the current dataset.

[0099] Predicted Output: The trained machine learning model can predict the risk level of cerebrovascular injury based on the input image features. The risk level can be divided into multiple categories (such as low risk, medium risk, high risk), providing a scientific basis for subsequent clinical interventions.

[0100] Based on the risk level generated by the prediction model, the personalized treatment recommendation generation unit combines the patient's clinical data to generate a personalized treatment plan. Data Integration and Analysis: The personalized treatment recommendation generation unit combines the patient's clinical data (such as age, gender, medical history, blood pressure, lifestyle habits, etc.) with the prediction results and provides customized treatment recommendations by analyzing the risk level of hypertensive patients.

[0101] Generation of Treatment Plan: According to the predicted risk level, the system will recommend corresponding treatment measures for the patient. For example, for high-risk patients, the system may recommend further imaging examinations, drug treatments, lifestyle interventions, etc. For low-risk patients, the system may recommend regular monitoring and health management.

[0102] The feature extraction and prediction model module can extract radiomics features from DWI-VEMRI and DTI-ALPS image data and accurately evaluate the risk level of cerebrovascular injury in hypertensive patients using machine learning algorithms. Through these prediction results, doctors can provide precise treatment plans for patients to ensure the effectiveness and personalization of treatment. At the same time, this module can also comprehensively analyze the health status of cerebral blood vessels by calculating features such as glymphatic system function index, vascular elasticity, and blood flow velocity, and detect potential risks of cerebrovascular injury in advance.

[0103] The data visualization and reporting module is used to generate a quantitative report on cerebrovascular injury and display the analysis results of the prediction model in the form of charts.

[0104] Please refer to the appendix Figure 5, The main function of the data visualization and reporting module is to integrate the quantitative results from the virtual elastography analysis module and the microcirculation analysis module, generate a quantitative report on cerebrovascular injury, and display the analysis results of the prediction model in the form of charts. In this way, the module can help doctors intuitively understand the cerebrovascular health status of patients, and thus provide support for clinical decision-making. This module can not only provide detailed quantitative results on cerebrovascular injury, but also generate personalized treatment plans and follow-up monitoring plans for patients.

[0105] The implementation of this module includes subunits such as the display of quantitative analysis results, the visualization of prediction results, and the generation of personalized reports. Through these functions, the system can generate a comprehensive assessment report on cerebrovascular injury and provide accurate health management and treatment plans.

[0106] In this embodiment, the specific implementation of the data visualization and reporting module is divided into the following subunits, and the functions of each unit play an important role in the entire system.

[0107] Generally, the function of the quantitative analysis result display unit is to integrate the quantitative results of the virtual elastography analysis module and the microcirculation analysis module to generate quantitative data on cerebrovascular injury.

[0108] This unit combines data such as the cerebrovascular elastic modulus (E) and blood flow velocity (vvv) in the virtual elastography analysis module with data such as the glymphatic system function index (LLL) in the microcirculation analysis module to generate comprehensive cerebrovascular health data.

[0109] Elastic modulus E: The cerebrovascular elastic modulus calculated by the virtual elastography analysis module through calculating the diffusion tensor has the following formula: where D xx , D yy , D zz are the three principal axis components of the diffusion tensor, respectively representing the diffusion coefficients of the blood vessel in the X, Y, and Z directions. The elastic modulus E is an important indicator reflecting the hardness of the blood vessel. A higher elastic modulus means blood vessel sclerosis, which may be related to hypertension or other vascular diseases.

[0110] Blood flow velocity v: The blood flow velocity is calculated through the virtual elastic tensor model, and the formula is: where Δx represents the displacement of blood flow, and Δt is the time interval. The blood flow velocity provides an important dynamic indicator for the health status of the cerebrovascular. A lower blood flow velocity may indicate blood vessel obstruction or insufficient perfusion.

[0111] Lymphatic system function index L: The lymphatic system function index L is calculated by DTI-ALPS technology and reflects the health status of cerebral microcirculation. The calculation formula of this index is: where C ij is the connectivity between the i-th fiber bundle and the j-th fiber bundle, and N is the total number of fiber bundles. A lower L value usually means a decline in the function of cerebral vascular microcirculation, which is particularly common in patients with hypertension.

[0112] This unit converts the above quantitative data into charts. Common display methods include: Bar chart: Used to show the comparison of different patients in terms of indicators such as elastic modulus, blood flow velocity, and lymphatic system function index. Through the bar chart, doctors can easily understand the health status of patients, especially the comparison with the normal range.

[0113] Line chart: Used to display the changing trends of blood flow velocity and elastic modulus over time, helping doctors observe the changes in the vascular health status of patients.

[0114] Scatter plot: Used to analyze the relationship between different radiomics features, such as the relationship between blood flow velocity and elastic modulus, to determine whether there are abnormalities.

[0115] These charts can help doctors quickly identify potential health problems and further support the diagnosis process.

[0116] As an option, the prediction result visualization unit is used to display the analysis results of the prediction model in the form of charts to help doctors intuitively understand the health status of patients.

[0117] The prediction result visualization unit generates different charts to display the prediction results according to the cerebrovascular injury risk levels output by the machine learning model: Risk level distribution chart: Displays the distribution of the patient group in different risk levels (such as low risk, medium risk, high risk). Through this chart, doctors can quickly understand the risk distribution of the patient group.

[0118] ROC curve: The ROC (Receiver Operating Characteristic) curve shows the performance of the model at different thresholds and helps evaluate the effectiveness of the prediction model by comparing the false positive rate and the true positive rate.

[0119] Accuracy chart and precision-recall chart: Display the prediction accuracy, precision, and recall of the model on different datasets, providing doctors with the model evaluation results.

[0120] These charts and visualizations can enable doctors to more intuitively understand the cerebrovascular injury risk levels of hypertensive patients and take corresponding measures.

[0121] The customized report generation unit generates personalized treatment reports based on the quantitative analysis results, the output of the prediction model, and the patient's clinical data.

[0122] Report content: Specific indicators of cerebrovascular injury: The report details relevant indicators such as the elastic modulus of cerebral blood vessels, blood flow velocity, and glymphatic system function index, and provides a comparison with normal values. For example, patients with a lower blood flow velocity may require further examinations or treatments.

[0123] Prediction results: According to the prediction model, the report will show the risk level of the patient's cerebrovascular injury and provide a description of the clinical significance of this level. Low-risk patients may not require intervention, while high-risk patients may need to take immediate intervention measures.

[0124] Recommended treatment plan: Based on the patient's risk level and the degree of vascular injury, the report will recommend a personalized treatment plan. For example, drug treatment, lifestyle intervention, surgical treatment, etc. For high-risk patients, more frequent follow-ups or stricter health management plans may be recommended.

[0125] Subsequent monitoring plan: The report will provide a subsequent monitoring plan based on the risk assessment to ensure that the patient can receive timely examinations and evaluations during the treatment process.

[0126] The reports generated by this module will be presented in a standardized medical report format and support multiple output formats (such as PDF, Excel). The report content will include forms such as charts and images to ensure that doctors can comprehensively understand the patient's health status and make decisions quickly.

[0127] The data visualization and reporting module helps doctors make quick judgments and adopt appropriate treatment plans by converting complex quantitative data into easy-to-understand charts and reports. By integrating the results of the virtual elastography analysis and microcirculation analysis modules, the module can provide a comprehensive assessment of cerebrovascular injury, providing strong data support for clinical decision-making. The generation of personalized reports can help doctors develop customized treatment plans for each patient and guide subsequent health management and monitoring.

[0128] The early monitoring and evaluation method for hypertensive brain injury of multimodal MRI data described below can be mutually referred to the early monitoring and evaluation system for hypertensive brain injury of multimodal MRI data described above.

[0129] Please refer to the appendix Figure 6 , the early monitoring and evaluation method for hypertensive brain injury of multimodal MRI data, applied to the above-mentioned early monitoring and evaluation system for hypertensive brain injury of multimodal MRI data, includes the following steps: S1. Collect multimodal MRI data of hypertensive patients, including DWI-VEMRI sequences and DTI-ALPS sequences; S2. Denoise, register, and standardize the collected image data to generate high-quality standardized image data; S3. Calculate the virtual elastic tensor based on DWI-VEMRI data, and then calculate the elastic modulus, blood flow velocity, and perfusion index of cerebral blood vessels, and perform quantitative analysis of cerebral vascular injury; S4. Use DTI-ALPS data for fiber tractography, analyze the diffusion direction of water molecules in the brain, and calculate the glymphatic system function index; S5. Extract radiomics features from DWI-VEMRI and DTI-ALPS data, and construct a prediction model for hypertensive-related cerebral vascular injury through machine learning algorithms; S6. Generate a quantitative report, display the analysis results of cerebral vascular injury, and provide personalized suggestions based on the prediction model. The method of this embodiment can be used to implement the above system embodiment, and its principle and technical effects are similar, which will not be elaborated here.

[0130] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An early monitoring and evaluation system for hypertensive brain injury of multimodal MRI data, characterized in that, Including: A data acquisition module, which is used to collect multimodal image data of hypertensive patients through an MRI device, and the data includes DWI-VEMRI sequences and DTI-ALPS sequences; A data preprocessing module, which is used to perform denoising, registration, and normalization processing on the collected DWI-VEMRI and DTI-ALPS image data to generate standardized high-quality image data; A virtual elastography analysis module, which is used to calculate the virtual elastic tensor of cerebral blood vessels based on DWI-VEMRI data, and obtain the elastic modulus of cerebral blood vessels through this tensor, and the elastic modulus is used to evaluate the hardness or elastic properties of cerebral blood vessels; A microcirculation analysis module, which is used to perform fiber bundle tracking analysis through DTI-ALPS technology, calculate the glymphatic system function index, and evaluate the changes in cerebral microcirculation caused by hypertension; A feature extraction and prediction model module, which is used to extract radiomics features from the DWI-VEMRI and DTI-ALPS image data, and combine machine learning algorithms to construct a prediction model for cerebrovascular injury related to hypertension; A data visualization and reporting module, which is used to generate a quantitative report of cerebrovascular injury and display the analysis results of the prediction model in the form of charts.

2. The early monitoring and evaluation system for hypertensive brain injury of multimodal MRI data according to claim 1, wherein The virtual elastography analysis module includes: A DWI data acquisition unit, which is used to acquire high-resolution DWI-VEMRI image data and collect diffusion-weighted image data in different directions; An elastic tensor calculation unit, which is based on the acquired DWI data and uses a diffusion tensor imaging model to calculate the virtual elastic tensor of the vascular region; An elastic modulus calculation unit, which calculates the elastic modulus of cerebral blood vessels according to the obtained virtual elastic tensor, so as to evaluate the hardness or elastic properties of cerebral blood vessels; A blood flow velocity calculation unit, which calculates the cerebral blood flow velocity through a virtual elastic tensor model and estimates the blood flow velocity in cerebral blood vessels; A cerebral blood vessel perfusion analysis unit, which calculates the perfusion index of cerebral blood vessels and provides a comprehensive evaluation of cerebral blood vessel function by comparing the differences between the healthy group and hypertensive patients.

3. The early monitoring and evaluation system for hypertensive brain injury of multimodal MRI data according to claim 2, characterized in that, The formula for calculating the virtual elastic tensor of the vascular region in the elastic modulus calculation unit is as follows: where E represents the elastic modulus of the cerebral blood vessels, and D xx , D yy , D zz are the three main components of the diffusion tensor, respectively.

4. The early monitoring and evaluation system for hypertensive brain injury of multimodal MRI data according to claim 1, characterized in that, The microcirculation analysis module includes: A water molecule diffusion tracking unit, which uses DTI-ALPS technology to track the diffusion direction of water molecules in the brain, obtains the diffusion trajectory of water molecules along the fiber bundle, and analyzes the structural connectivity of white matter fibers in the brain; A glymphatic system function analysis unit, which calculates the function index of the glymphatic system based on the water molecule diffusion tracking data; A microcirculation evaluation unit, which evaluates the cerebrovascular microcirculation function of hypertensive patients according to the obtained glymphatic system function index and compares it with the healthy control group.

5. The early monitoring and evaluation system for hypertensive brain injury of multimodal MRI data according to claim 4, characterized in that, The formula for calculating the function index of the glymphatic system in the glymphatic system function analysis unit is as follows: Among them, C ij is the connection degree between the i-th fiber bundle and the j-th fiber bundle, N is the total number of fiber bundles, and L represents the lymphatic system function index.

6. The early monitoring and evaluation system for hypertensive brain injury of multimodal MRI data according to claim 1, characterized in that, The feature extraction and prediction model module includes: An image feature extraction unit, which extracts relevant radiomics features from DWI-VEMRI and DTI-ALPS image data, such as texture features, morphological features, vascular elasticity, blood flow velocity, and glymphatic system function index; A machine learning model construction unit, which uses a support vector machine, a random forest or a deep learning algorithm to construct a prediction model based on radiomics features to predict the risk level of cerebrovascular injury; A personalized treatment recommendation generation unit, which provides personalized treatment recommendations for patients according to the prediction results combined with the patients' clinical data.

7. The early monitoring and evaluation system for hypertensive brain injury of multimodal MRI data according to claim 1, wherein The data visualization and reporting module includes: A quantitative analysis result display unit, which integrates the quantitative results of the virtual elastography analysis module and the microcirculation analysis module to generate quantitative data on cerebrovascular injury; A prediction result visualization unit, which displays the analysis results of the prediction model in the form of charts to help doctors intuitively understand the health status of patients; A customized report generation unit, which generates a personalized report for patients, including specific indicators of cerebrovascular injury, prediction results, recommended treatment plans and subsequent monitoring plans.

8. Early monitoring and evaluation method for hypertensive brain injury of multimodal MRI data, characterized in that, An early monitoring and evaluation system for hypertensive brain injury applied to the multi-modal MRI data described in any one of claims 1-7 above, comprising the following steps: S1. Collect multi-modal MRI data of hypertensive patients, including DWI-VEMRI sequences and DTI-ALPS sequences; S2. Denoise, register and standardize the collected image data to generate high-quality standardized image data; S3. Calculate the virtual elastic tensor based on DWI-VEMRI data, and then calculate the elastic modulus, blood flow velocity and perfusion index of cerebral blood vessels, and perform quantitative analysis of cerebrovascular injury; S4. Use DTI-ALPS data for fiber tractography, analyze the diffusion direction of water molecules in the brain and calculate the glymphatic system function index; S5. Extract radiomics features from DWI-VEMRI and DTI-ALPS data, and construct a prediction model for hypertension-related cerebrovascular injury through machine learning algorithms; S6. Generate a quantitative report, display the analysis results of cerebrovascular injury, and provide personalized recommendations based on the prediction model.

Images