A data processing method for evaluating blood-brain barrier permeability of mild and moderate white matter hyperintensity based on DCE-MRI image parameters

Abstract

The application discloses a data processing method for evaluating blood-brain barrier permeability of mild and moderate white matter hyperintensity based on DCE-MRI image parameters. trans The method comprises the following steps: acquiring DCE-MRI data; calculating volume transfer constant (K trans ) and plasma volume fraction (Vp) by using an Extended Tofts double-chamber model; performing Fazekas scoring based on T2-FLAIR image, and screening out mild and moderate lesion regions with Fazekas 0 / 1 score or 2 score; and taking K trans as an index of BBB permeability. The application excludes the severe lesion region with Fazekas 3 score, thereby avoiding the false negative bias caused by local low perfusion. Further, the perfusion correction is performed on Vp for periventricular WMH, and the risk prompt is performed on the state of type 2 diabetes associated with deep WMH. The application provides an accurate and quantifiable data processing scheme for evaluating blood-brain barrier damage of white matter hyperintensity, and can be integrated into a magnetic resonance post-processing system and used for assisting clinical research and drug efficacy evaluation.

Description

Technical Field

[0001] This invention relates to the field of medical image data processing technology, specifically to quantitative analysis of magnetic resonance images of cerebral small vessel diseases, and more specifically, to a data processing method for assessing the permeability of the blood-brain barrier (BBB) ​​in mild to moderate white matter hyperintensity (WMH) based on dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parameters. Background Technology

[0002] White matter hyperintensities (WMH) are a core imaging biomarker for cerebral small vessel disease (CSVD), appearing as high-signal foci in the periventricular or deep white matter regions on T2-FLAIR MRI sequences. WMH is closely associated with cognitive decline, gait abnormalities, and an increased risk of ischemic stroke. Damage to the blood-brain barrier (BBB) ​​is considered a key factor in the development and progression of WMH.

[0003] Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is currently the gold standard for in vivo quantitative detection of BBB permeability. Through analysis using the Extended Tofts two-chamber model, the volume transfer constant (K0) can be obtained. trans ) and plasma volume fraction (Vp). K trans The rate at which contrast agent leaks from inside the blood vessel to outside the blood vessel is a core parameter for assessing BBB permeability; Vp reflects local vascular density and perfusion level.

[0004] However, the existing technology for assessing BBB leakage in WMH has the following technical shortcomings:

[0005] First, traditional views hold that the severity of WMH (Fazekas score) is linearly positively correlated with BBB leakage. However, recent studies have found that the K region in severe WMH (Fazekas score 3) areas... trans The value actually decreased. Studies have shown that this is because the microvessels are sparse and local brain perfusion is significantly reduced in severely affected areas, leading to limited delivery of contrast agents to the lesion area, resulting in K+ ionization. trans Quantitative results are significantly underestimated by perfusion factors, leading to false negative bias. Existing methods do not stratify lesion severity, resulting in reduced assessment accuracy.

[0006] Second, the pathological mechanisms of peripheral white matter hyperintensity (PWMH) and deep white matter hyperintensity (DWMH) are clearly heterogeneous, with different driving factors for BBB leakage—PWMH is mainly regulated by age and local perfusion levels, while DWMH is closely related to type 2 diabetes. However, existing assessment methods do not differentiate between anatomical subtypes, nor do they introduce corresponding correction or correlation parameters, leading to inaccurate assessment results.

[0007] Therefore, there is an urgent need for a BBB permeability data processing method that can avoid perfusion interference from severe lesions, differentiate anatomical subtypes, and correlate with clinical risk factors. Summary of the Invention

[0008] To address the shortcomings of the existing technology, this invention aims to provide a data processing method for assessing blood-brain barrier permeability in mild to moderate white matter hyperintensity based on DCE-MRI image parameters, thereby solving the technical problem of inaccurate BBB leakage assessment caused by false negative interference from perfusion in severe lesions and neglect of anatomical subtype heterogeneity in the existing technology.

[0009] To achieve the above objectives, the present invention adopts the following technical solution:

[0010] A data processing method for assessing blood-brain barrier permeability in mild to moderate white matter hyperintensity based on DCE-MRI image parameters, characterized by the following steps:

[0011] Step S1: Acquire DCE-MRI data of the subject's brain;

[0012] Step S2: Based on the DCE-MRI data, calculate the volume transfer constant K of the white matter region using the Extended Tofts dual-chamber model. trans and plasma volume fraction Vp;

[0013] Step S3: Based on T2-FLAIR images, perform Fazekas 0-3 grading for periventricular white matter hyperintensity (PWMH) and deep white matter hyperintensity (DWMH), where 0 / 1 group represents no lesion or mild lesion, 2 group represents moderate lesion, and 3 group represents severe lesion.

[0014] Step S4: Delineate the region of interest (ROI) on the T2-FLAIR image. The ROIs in the Fazekas 2 and 3 groups are delineated in the core area of ​​the WMH lesion, and the ROIs in the Fazekas 0 / 1 groups are delineated in the normal white matter area of ​​the corresponding anatomical location.

[0015] Step S5: Using the K transAs a quantitative assessment indicator of blood-brain barrier (BBB) ​​permeability, the K value in the PWMH and DWMH regions of the Fazekas score of 2 was found to be... trans The value was significantly higher than that of the Fazekas 0 / 1 score WMH region, indicating increased BBB permeability; under the same Fazekas score stratification, the K value of PWMH was significantly higher. trans The values ​​are generally higher than DWMH;

[0016] Step S6: Using the Vp parameter, evaluate the local microvascular perfusion characteristics of the WMH region. The results show that, with Fazekas 0 / 1 score WMH as a reference, the Vp values ​​of PWMH and DWMH show a continuous decreasing trend as the Fazekas grade increases. The Vp values ​​of Fazekas 2 and 3 score PWMH are significantly lower than those of the 0 / 1 group, and the Vp value of Fazekas 3 score DWMH is significantly lower than that of the 0 / 1 group.

[0017] Step S7: Spearman rank correlation analysis was used to analyze the correlation between Ktrans, the core marker of BBB leakage, and Vp, the local perfusion evaluation index. The results showed that in the Fazekas 0 / 1 group, Ktrans in the PWMH region... trans It shows a moderate positive correlation with Vp, and the K region of DWMH trans It shows a weak positive correlation with Vp;

[0018] Step S8: Use a univariate analysis system to assess the balance of baseline risk factor distribution among WMH score groups of different anatomical sites and different lesion severity. Baseline indicators include gender, hypertension, type 2 diabetes, hyperlipidemia, coronary heart disease, smoking history, and drinking history to exclude confounding variables from interfering with the heterogeneity analysis results of WMH BBB leakage and perfusion characteristics.

[0019] Step S9: Establish a multiple linear regression model using the BBB leakage core quantitative parameter K. trans With age, sex, vascular risk factors, lifestyle habits, Vp (ventricular puncture rate), and lesion severity as the dependent variable, a multiple linear regression analysis was conducted. The results showed that there were differentiated independent regulatory patterns of BBB leakage in PWMH and DWMH. Age was an independent negative influencing factor for BBB leakage in the PWMH region, while Vp and Fazekas 2 score were independent positive influencing factors. Sex, hypertension, diabetes, hyperlipidemia, coronary heart disease, smoking, and alcohol consumption had no significant independent effects on peripheral white matter BBB leakage.

[0020] Regression analysis of DWMH showed that type 2 diabetes and Fazekas 2 lesion grade were independent positive factors influencing BBB leakage in the DWMH region. The effect of local perfusion level Vp on BBB leakage in DWMH was close to the statistical critical value. Age, sex, hypertension, hyperlipidemia, coronary heart disease, smoking, and alcohol consumption had no significant independent effect on deep white matter BBB leakage.

[0021] Step S10: Conclusion Analysis

[0022] Firstly, BBB leakage in WMH does not increase linearly with the severity of the lesion, but rather exhibits a significant non-linear change characteristic. That is, the BBB permeability of Fazekas score 2 is significantly higher than that of areas with no / mild lesions, while the BBB permeability of Fazekas score 3 shows a significant decreasing trend.

[0023] Secondly, under the same Fazekas classification, the degree of BBB leakage in PWMH is generally higher than that in DWMH, suggesting that there is clear anatomical heterogeneity in BBB damage in WMH.

[0024] Third, PWMH and DWMH exhibit different regulatory modes of BBB leakage. PWMH BBB leakage is mainly regulated by local brain perfusion levels, while DWMH BBB leakage is independently driven by type 2 diabetes.

[0025] Furthermore, the Fazekas scoring criteria in step S3 are as follows:

[0026] 0 / 1 points: No lesions or punctate / cap-shaped lesions with a diameter <5mm;

[0027] 2 points: For PWMH, it is a smooth halo-like confluent change, involving the circumference of the lateral ventricle body >25%; for DWMH, it is a focal small-area confluent change.

[0028] Furthermore, in step S4, the area of ​​a single ROI is controlled to be between 50 and 100 mm², and it avoids cerebral blood vessels and sulci.

[0029] Furthermore, the DCE-MRI data acquisition parameters are as follows: flip angle 10°, TR / TE 2.51 / 0.92 ms, time resolution 3 seconds, a total of 90 phases acquired, the contrast agent is gadobutrol, dose 0.1 mmol / kg, injection rate 3 mL / s.

[0030] Furthermore, the determination of "significantly higher than" is based on a nonparametric test with a statistical significance threshold of P<0.05.

[0031] The beneficial effects of this invention include: avoiding false negative bias in perfusion: by limiting the analysis to mild to moderate WMH areas with Fazekas scores of 0-2, severe WMH (Fazekas score 3) lesions are actively excluded. Experimental verification shows that severe WMH areas, due to sparse microvessels and localized hypoperfusion, have limited contrast agent delivery, and their K... trans The value decreased significantly compared to the 0 / 1 group, but this decrease was positively correlated with the reduction in Vp value, indicating a measurement artifact rather than a true BBB repair. This invention significantly improves the accuracy of BBB leakage assessment by excluding such areas.

[0032] Anatomical subtype-specific processing: This invention distinguishes between PWMH and DWMH and provides differentiated data processing strategies for each—introducing Vp perfusion correction for PWMH and type 2 diabetes status association for DWMH. Experimental data confirm that PWMH's K... trans It showed a significant positive correlation with Vp, and the BBB leakage of DWMH was independently driven by type 2 diabetes. This subtype-specific treatment overcomes the "one-size-fits-all" shortcomings of existing methods.

[0033] Provides quantifiable intermediate results: This invention outputs K trans Values, Vp values, BBB permeability indicators, or comprehensive predictive scores are all objective and quantifiable intermediate results that do not directly provide a disease diagnosis conclusion and comply with the requirements of the Patent Law for data processing methods.

[0034] Clearly defined intervention window: This invention compares the K values ​​of different Fazekas classifications. trans The results showed that BBB leakage in the Fazekas 2-point region was significantly higher than in the 0 / 1-point region, suggesting that this stage is the pathologically active period of BBB damage. This information can provide a reference for subsequent clinical decisions. Attached Figure Description

[0035] Figure 1 This is an overall flowchart of the method of the present invention.

[0036] Figure 2 K for different Fazekas grades and anatomical sites of WMH trans Box plot of value distribution.

[0037] Figure 3 Box plots of Vp values ​​for WMH at different Fazekas grades and anatomical sites show that perfusion levels decrease with increasing severity.

[0038] Figure 4 For different Fazekas classifications, the PWMH and DWMH regions K trans Scatter plot of correlation with Vp. Detailed Implementation

[0039] The present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0040] Example 1 (BBB permeability assessment of PWMH)

[0041] 1. Data Collection

[0042] A total of 75 valid participants were included in the study. The age of the enrolled population was not normally distributed, with a median age of 67 years and an interquartile range of 58–75 years. Among them, 43 were male, accounting for 57.33% of the total study population. Regarding vascular risk factors, 39 participants (52.00%) had hypertension, 20 had type 2 diabetes (26.67%), 27 had hyperlipidemia (36.00%), and 17 had coronary heart disease (22.67%). Regarding lifestyle history, 25 participants (33.33%) had a history of smoking, and 18 participants (24.00%) had a history of alcohol consumption.

[0043] The United Imaging 3.0T MRI scanner (UMR780) and a 24-channel head and neck coil were used. Subjects were placed in a supine position with their heads fixed with a sponge rubber pad.

[0044] DCE-MRI sequence parameters: multiple flip angle scans (3°, 6°, 9°, 12°, 15°), matrix 112×100, FOV 230×200 mm, TR / TE 4.11 / 1.84 ms, excitation times 1, 20 slices, slice thickness 6 mm; multi-phase dynamic enhancement sequence: flip angle 10°, TR / TE 2.51 / 0.92 ms, temporal resolution 3 seconds, a total of 90 phases acquired, total scan time 253 seconds. The contrast agent was gadobutrol, injected via high-pressure intravenous injection (rate 3 mL / s) at the end of phase 5, dose 0.1 mmol / kg, followed by a bolus injection of 20 mL of 0.9% sodium chloride solution at the same rate.

[0045] 2. Fazekas score and lesion screening

[0046] Based on T2-FLAIR images, the United Imaging Brain Small Vessel Disease Intelligent Analysis System, combined with physician review, was used to perform Fazekas scoring on PWMH (Pulmonary Small Vessel Hemorrhage). Scoring criteria: 0 / 1 point – focal punctate, cap-shaped, or pencil-line-like mild changes without significant fusion; 2 points – smooth halo-like fused changes involving >25% of the lateral ventricle body circumference. PWMH regions with Fazekas scores of 0 / 1 or 2 were selected for analysis. Fazekas score 3: Lesions show extensive, irregular, confluent changes with high signal extending into deep white matter, and the affected area in the circumference of the lateral ventricle body is further expanded. Fazekas scoring for DWMH: Fazekas 0 / 1: Lesions are scattered, punctate high signal areas <5 mm in diameter, without confluence; DWMH Fazekas 2: Lesions show focal, small-scale confluent changes; DWMH Fazekas 3: Lesions show large areas of confluent high signal, or bridging of periventricular white matter and deep white matter lesions.

[0047] Fazekas grading distribution characteristics in WMH: In PWMH, 30 cases (40.00%) had a Fazekas score of 0 / 1, 32 cases (42.67%) had a score of 2, and 13 cases (17.33%) had a score of 3; In DWMH, 50 cases (66.67%) had a Fazekas score of 0 / 1, 16 cases (21.33%) had a score of 2, and 9 cases (12.00%) had a score of 3.

[0048] 3. Data Preprocessing and Parameter Calculation

[0049] An Extended Tofts dual-chamber model was adopted. First, motion correction of the dynamic sequence images was performed using a rigid body registration algorithm. An arterial input function was constructed using the M1 segment of the middle cerebral artery, with the superior sagittal sinus as the venous output reference.

[0050] On T2-FLAIR sequences, regions of interest (ROIs) were delineated blinded by trained analysts. For patients with Fazekas score of 2 and PWMH, the ROI was delineated in the core area of ​​the lesion; for patients with Fazekas score of 0 / 1, the ROI was delineated in the normal white matter area corresponding to the anatomical location. The ROI area was controlled to be 50-100 mm², avoiding cerebral blood vessels and sulci. Ki values ​​of the corresponding anatomical regions were calculated bilaterally. trans The arithmetic mean of the Vp values.

[0051] 4. Statistical Analysis Methods

[0052] GraphPad Prism 9 software was used for data statistical analysis and chart creation. Qualitative data were described as number of cases (percentage) n (%), and comparisons between groups were performed using the χ² test or Fisher's exact test. Quantitative data, after normality testing, were expressed as mean ± standard deviation for normally distributed data, and comparisons between groups were performed using the two independent samples t-test. Non-normally distributed data were described as median (interquartile range) [M (P25, P75)], and comparisons between groups were performed using the Mann-Whitney U test. Spearman rank correlation analysis was used to analyze the correlation between BBB leakage indicators and local perfusion indicators. Stepwise multiple linear regression analysis was used to explore the independent influencing factors of different anatomical subtypes of WMH BBB leakage. All tests were two-tailed, with a significance level of α=0.05, and P < 0.05 was considered statistically significant.

[0053] 5. BBB permeability assessment and perfusion correction

[0054] With K trans The value is used as an indicator of BBB transparency. Through a nonparametric Mann-Whitney U test, the Fazekas score of 2 indicates PWMH, and the K value of DWMH indicates... trans The values ​​were all significantly higher than those in the 0 / 1 group. Meanwhile, Fazekas scored 3 points for PWMH and DWMH K. trans The value was significantly lower than that in the 0 / 1 group, and did not show a continuous upward trend with the increase of lesion severity; under the same Fazekas score stratification, the K value of PWMH was significantly lower. trans The values ​​are generally higher than DWMH, with the intergroup K values ​​for levels 0 / 1 and 2 WMH being particularly high. trans All differences in values ​​were statistically significant (all P < 0.05), but there was no statistically significant difference between groups for grade 3 WMH (P > 0.05).

[0055] Simultaneously, Vp value was calculated to correct for local perfusion level on K. transThe quantitative results showed interference, with dual heterogeneity in local perfusion levels of WMH based on Fazekas grading and anatomical region, as illustrated in Figure 3. Using Fazekas 0 / 1 WMH as a reference, the Vp value of PWMH showed a continuous decreasing trend with increasing Fazekas grading. Specifically, the Vp values ​​of Fazekas 2 and 3 PWMH were significantly lower than those of the 0 / 1 group (P=0.027, P<0.0001). Similarly, the Vp value of DWMH also showed a decreasing trend with increasing Fazekas grading, except for Fazekas 3 DWMH, where the Vp value was significantly lower than that of the 0 / 1 group (P=0.021). Under the same Fazekas rating stratification, the Vp value of PWMH in level 0 / 1 WMH was significantly higher than that of DWMH (P<0.0001); the Vp value of PWMH in level 2 WMH was slightly higher than that of DWMH, and the difference was at the statistical critical level (P=0.054); there was no statistically significant difference in the Vp value between PWMH and DWMH in level 3 WMH (P=0.883).

[0056] Spearman rank correlation analysis was used to analyze K, the core biomarker of BBB leakage. trans Correlation analysis was performed with the local perfusion evaluation index Vp. The results are shown in Figure 4. Except for the Fazekas 3-point PWMH group, the K values ​​of the other subgroups were... trans The values ​​and Vp values ​​both show a positive correlation trend. In the Fazekas 0 / 1 subgroup, the PWMH region K... trans It shows a moderate positive correlation with Vp (r=0.444, P=0.014), DWMH region K trans It showed a weak positive correlation with Vp (r=0.288, P=0.043), and both associations were statistically significant; in the lesion area of ​​the Fazekas 2 subgroup, K trans It maintains a positive correlation with Vp, but the correlation does not reach a statistically significant level.

[0057] Based on the above correlation, the present invention further performs perfusion correction:

[0058] Based on the BBB core quantitative parameter K transUsing demographic characteristics (age, sex), vascular risk factors (hypertension, diabetes, hyperlipidemia, coronary heart disease), lifestyle habits (smoking, alcohol consumption), perfusion index (Vp), and lesion severity (Fazekas score 2) as independent variables, multiple linear regression analysis was conducted to identify the independent regulatory factors of BBB leakage in different anatomical sites of WMH. The results showed that there were differentiated independent regulatory patterns of BBB leakage between PWMH and DWMH. Age was an independent negative influencing factor for BBB leakage in the PWMH region, while local perfusion index Vp and Fazekas score lesion grade were independent positive influencing factors for BBB leakage in the PWMH region. Sex, hypertension, diabetes, hyperlipidemia, coronary heart disease, smoking, and alcohol consumption had no statistically significant effect on peripheral white matter BBB leakage (all P>0.05). Regression analysis of DWMH showed that type 2 diabetes and Fazekas score lesion grade were independent regulatory factors for BBB leakage in the DWMH region. Independent positive influencing factors of leakage: The effect of local perfusion level Vp on BBB leakage in DWMH is close to the statistical critical value. Age, sex, hypertension, hyperlipidemia, coronary heart disease, smoking, and alcohol consumption have no significant independent effects on deep white matter BBB leakage (all P>0.05).

[0059] 5. Output Results

[0060] Output at least one of the following pieces of information: First, BBB leakage in WMH does not increase linearly with the severity of the lesion (Fazekas score), but rather exhibits a significant non-linear change characteristic, that is, the BBB permeability of moderate WMH (Fazekas score 2) is significantly higher than that of areas with no / mild lesions, while the BBB permeability of severe WMH (Fazekas score 3) shows a significant decreasing trend; Second, under the same Fazekas grade, the degree of BBB leakage in PWMH is generally higher than that in DWMH, suggesting that there is a clear anatomical heterogeneity in BBB damage in WMH; Third, there are different regulatory patterns for BBB leakage in PWMH and DWMH. BBB leakage in PWMH is mainly regulated by the level of local brain perfusion, while BBB leakage in DWMH is independently driven by type 2 diabetes.

[0061] Industrial applicability

[0062] The method of this invention can be integrated into the post-processing workstation of a magnetic resonance imaging (MRI) device or into standalone medical image analysis software, operating automatically as a data processing module. Its inputs are standard DCE-MRI raw data and T2-FLAIR images, and its output is quantitative BBB permeability indicators and related information. This method does not rely on the subjective judgment of physicians, has high repeatability and automation potential, and can be widely applied to the quantitative imaging assessment of cerebral small vessel diseases, the detection of surrogate endpoints in drug clinical trials, and the standardized analysis of population cohort studies.

[0063] The technical solutions provided by the embodiments of the present invention have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of the embodiments of the present invention. The descriptions of the embodiments above are only for helping to understand the principles of the embodiments of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the embodiments of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A data processing method for assessing blood-brain barrier permeability in mild to moderate white matter hyperintensity based on DCE-MRI image parameters, characterized in that, Includes the following steps: Step S1: Acquire DCE-MRI data of the subject's brain; Step S2: Based on the DCE-MRI data, calculate the volume transfer constant K of the white matter region using the Extended Tofts dual-chamber model. trans and plasma volume fraction Vp; Step S3: Based on T2-FLAIR images, perform Fazekas 0-3 grading for periventricular white matter hyperintensity (PWMH) and deep white matter hyperintensity (DWMH), where 0 / 1 group represents no lesion or mild lesion, 2 group represents moderate lesion, and 3 group represents severe lesion. Step S4: Delineate the region of interest (ROI) on the T2-FLAIR image. The ROIs in the Fazekas 2 and 3 groups are delineated in the core area of ​​the WMH lesion, and the ROIs in the Fazekas 0 / 1 groups are delineated in the normal white matter area of ​​the corresponding anatomical location. Step S5: Using the K trans As a quantitative assessment indicator of blood-brain barrier (BBB) ​​permeability, the results showed that a Fazekas score of 2 indicated K values ​​in the PWMH and DWMH regions. trans The value was significantly higher than that of the Fazekas 0 / 1 score WMH region, indicating increased BBB permeability; under the same Fazekas score stratification, the K value of PWMH was significantly higher. trans The values ​​are generally higher than DWMH; Step S6: Using the Vp parameter, evaluate the local microvascular perfusion characteristics of the WMH region. The results show that, with Fazekas 0 / 1 score WMH as a reference, the Vp values ​​of PWMH and DWMH show a continuous decreasing trend as the Fazekas grade increases. The Vp values ​​of Fazekas 2 and 3 score PWMH are significantly lower than those of the 0 / 1 group, and the Vp value of Fazekas 3 score DWMH is significantly lower than that of the 0 / 1 group. Step S7: Using Spearman rank correlation analysis, analyze the core biomarker K of BBB leakage. trans Correlation analysis was performed with the local perfusion evaluation index Vp. The results showed that in the Fazekas 0 / 1 group, the PWMH region K... trans It shows a moderate positive correlation with Vp, and the K region of DWMH trans It shows a weak positive correlation with Vp; Step S8: Use a univariate analysis system to assess the balance of baseline risk factor distribution among WMH score groups of different anatomical sites and different lesion severity. Baseline indicators include gender, hypertension, type 2 diabetes, hyperlipidemia, coronary heart disease, smoking history, and drinking history to exclude confounding variables from interfering with the heterogeneity analysis results of WMH BBB leakage and perfusion characteristics. Step S9: Establish a multiple linear regression model using the BBB leakage core quantitative parameter K. trans With age, sex, vascular risk factors, lifestyle habits, Vp (ventricular puncture rate), and lesion severity as the dependent variable, a multiple linear regression analysis was conducted. The results showed that there were differentiated independent regulatory patterns of BBB leakage in PWMH and DWMH. Age was an independent negative influencing factor for BBB leakage in the PWMH region, while Vp and Fazekas 2 score were independent positive influencing factors. Sex, hypertension, diabetes, hyperlipidemia, coronary heart disease, smoking, and alcohol consumption had no significant independent effects on peripheral white matter BBB leakage. Regression analysis of DWMH showed that type 2 diabetes and Fazekas 2 lesion grade were independent positive factors influencing BBB leakage in the DWMH region. The effect of local perfusion level Vp on BBB leakage in DWMH was close to the statistical critical value. Age, sex, hypertension, hyperlipidemia, coronary heart disease, smoking, and alcohol consumption had no significant independent effect on deep white matter BBB leakage. Step S10: Conclusion Analysis Firstly, BBB leakage in WMH does not increase linearly with the severity of the lesion, but rather exhibits a significant non-linear change characteristic. That is, the BBB permeability of Fazekas score 2 is significantly higher than that of areas with no / mild lesions, while the BBB permeability of Fazekas score 3 shows a significant decreasing trend. Secondly, under the same Fazekas classification, the degree of BBB leakage in PWMH is generally higher than that in DWMH, suggesting that there is clear anatomical heterogeneity in the BBB damage of WMH. Third, PWMH and DWMH exhibit different regulatory modes of BBB leakage. PWMH BBB leakage is mainly regulated by local brain perfusion levels, while DWMH BBB leakage is independently driven by type 2 diabetes.

2. The method according to claim 1, characterized in that, The Fazekas scoring criteria in step S3 are as follows: 0 / 1 points: No lesions or punctate / cap-shaped lesions with a diameter <5mm; 2 points: For PWMH, it is a smooth halo-like confluent change, involving the circumference of the lateral ventricle body >25%; for DWMH, it is a focal small-area confluent change.

3. The method according to claim 1, characterized in that, In step S4, the area of ​​a single ROI is controlled to be 50-100 mm², and cerebral blood vessels and sulci are avoided.

4. The method according to claim 1, characterized in that, The DCE-MRI data acquisition parameters were: flip angle 10°, TR / TE 2.51 / 0.92 ms, time resolution 3 seconds, a total of 90 phases were acquired, the contrast agent was gadobutrol, the dose was 0.1 mmol / kg, and the injection rate was 3 mL / s.

5. The method according to claim 1, characterized in that, The determination of "significantly higher than" is based on a nonparametric test, with a statistical significance threshold of P<0.05.

Images