Method for constructing and analyzing susceptibility map of anti-magnetic components in grey matter brain region and application of method

Abstract

The invention discloses a grey matter brain region anti-magnetic component susceptibility map construction and analysis method and application thereof, which can be used for exploring a global distribution mode of cortical iron deposition of an AD patient on a voxel level and positioning a cortical region with susceptibility related to cognition. The method comprises the following steps: (1) carrying out MR data acquisition on a subject on a magnetic resonance scanner; (2) unwrapping the multi-echo phase image by using a phase unwrapping method based on Laplacian; performing brain stripping processing on the amplitude image of the first echo by using a BET algorithm built in FSL software and generating a binary brain mask image; removing a background field of the unwound phase diagram by using a V-SHARP algorithm in combination with the brain mask diagram; calculating a magnetic susceptibility map from the local field map by using an STAR-QSM algorithm; (3) calculating the cortex thickness by using the T1 weighted structure image, and performing post-processing on the 3D-FSPGR sequence structure image; and (4) carrying out voxel-based whole-brain QSM analysis and cortex thickness statistical analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to a method for constructing and analyzing a magnetic susceptibility map of diamagnetic components of a gray matter brain region, and applications of the method. Background Art

[0002] Alzheimer's disease (AD) is the most common neurodegenerative disorder in the elderly. Its primary clinical manifestation is progressive memory loss, ultimately leading to severe cognitive impairment and behavioral abnormalities. Currently, the pathological hallmarks of AD are widely considered to be the excessive accumulation of amyloid-β plaques (Aβ) and tau neurofibrillary tangles in the gray matter of the brain. Previous histochemical studies have demonstrated that iron deposition is associated with Aβ plaques and hyperphosphorylated tau neurofibrillary tangles and can promote the formation and progression of pathological Aβ and hyperphosphorylated tau. Furthermore, numerous studies have consistently observed iron deposition within insoluble Aβ plaques and tau neurofibrillary tangles in AD patients. Iron, a major metal element in the brain, plays a crucial role in axonal myelin formation, energy metabolism, and neurotransmitter synthesis. Iron levels in specific brain regions, such as the basal ganglia, increase with age. However, disruption of iron homeostasis may contribute to the development of various age-related neurodegenerative diseases. Previous studies have shown that iron overload-induced oxidative stress can cause neurotoxicity in neurons, leading to neuronal loss and degeneration. Therefore, existing evidence suggests that iron is involved in the pathogenesis of AD, but the specific mechanism remains unknown.

[0003] Quantitative Susceptibility Mapping (QSM) is a novel MR susceptibility technique that uses advanced post-processing and reconstruction algorithms to noninvasively quantify the magnetic susceptibility of living tissue. Although the magnetic susceptibility measured by QSM is nonspecific and can vary between different biophysical substances, such as iron, calcium, lipids, and myelin, paramagnetic iron is generally considered to be the primary source of magnetic susceptibility in gray matter, as the concentrations of other biophysical substances in the cortex are very low. Therefore, studies of neurodegenerative diseases often use QSM to assess cortical iron content. Many QSM studies have demonstrated excessive iron accumulation in multiple brain regions of patients with Alzheimer's disease (AD), primarily located in deep gray matter nuclei and cortical regions, and increased QSM values ​​are associated with cognitive decline in AD. However, the vast majority of QSM studies in AD are limited to region of interest (ROI)-based analyses, with ROI selection primarily focused on deep gray matter nuclei and a few crude cortical regions derived from brain atlases. Voxel-based whole-brain analysis is an efficient data-driven method that is not limited to a priori predefined brain regions and can therefore be used to broadly explore the spatial distribution of magnetic susceptibility changes in the brains of patients with AD. However, voxel-based whole-brain analysis has not yet been used to explore the spatial distribution of cortical iron deposition in patients with AD and its relationship with cognition.

[0004] Brain atrophy is one of the most prominent neuroimaging features of AD, particularly atrophy of the temporal lobe and hippocampus, indicating significant damage and loss of neurons and synapses in the cerebral cortex. Neuronal loss is considered a major cause of cognitive impairment and can reflect the progression of neurodegeneration. Therefore, understanding the relationship between cortical thickness and iron accumulation and its impact on cognition may provide new insights into the pathogenic mechanisms of iron accumulation in AD. However, to date, the relationship between cortical iron accumulation and cortical thickness in AD has not been systematically studied. Summary of the Invention

[0005] In order to overcome the defects of the existing technology, the technical problem to be solved by the present invention is to provide a method for constructing and analyzing the magnetic susceptibility map of the diamagnetic component of the gray matter brain area, which can explore the global distribution pattern of cortical iron deposition in AD patients at the voxel level and locate the cortical areas where magnetic susceptibility is related to cognition.

[0006] The technical solution of the present invention is: a method for constructing and analyzing a magnetic susceptibility map of diamagnetic components of gray matter brain regions, comprising the following steps:

[0007] (1) MRI data acquisition: The subjects underwent MR data acquisition on a magnetic resonance scanner;

[0008] (2) QSM reconstruction and post-processing: Use the Laplacian-based phase unwrapping method to unwrap the multi-echo phase image; use the built-in BET algorithm of FSL software to perform brain stripping on the amplitude map of the first echo and generate a binary brain mask map; use the V-SHARP algorithm combined with the brain mask map to remove the background field of the unwrapped phase map; use

[0009] The STAR-QSM algorithm calculates the magnetic susceptibility map from the local field map;

[0010] (3) Structural image post-processing: Use T1-weighted structural images to calculate cortical thickness.

[0011] 3D-FSPGR sequence structural images were post-processed;

[0012] (4) Statistical analysis: voxel-based whole-brain QSM analysis and cortical thickness statistical analysis were performed.

[0013] This study uses quantitative magnetic susceptibility imaging (QSM) to explore the spatial distribution pattern of cortical iron deposition in the whole brain of AD patients and its relationship with cognition and cortical thickness; to study the complex relationship between brain network topology characteristics, cortical atrophy and QSM signals in AD patients; and to further understand the mechanism of the selective susceptibility of specific neuronal groups to abnormal iron metabolism in AD by analyzing the relationship between abnormal QSM signals and gene transcription and cellular characteristics.

[0014] The invention also provides an application of a method for constructing and analyzing a magnetic susceptibility map of the diamagnetic component of gray matter brain regions, which is used to monitor the progression of AD and assess the severity of cognitive decline. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A flow chart of the method for constructing and analyzing the magnetic susceptibility map of the diamagnetic component of the gray matter brain region according to the present invention is shown. DETAILED DESCRIPTION

[0016] like Figure 1 As shown, the method for constructing and analyzing the magnetic susceptibility map of the diamagnetic component of the gray matter brain region includes the following steps:

[0017] (1) MRI data acquisition: The subjects underwent MR data acquisition on a magnetic resonance scanner;

[0018] (2) QSM reconstruction and post-processing: Use the Laplacian-based phase unwrapping method to unwrap the multi-echo phase image; use the built-in BET algorithm of FSL software to perform brain stripping on the amplitude map of the first echo and generate a binary brain mask map; use

[0019] The V-SHARP algorithm is combined with the brain mask to remove the background field of the unwrapped phase image;

[0020] Magnetic susceptibility maps are calculated from local field maps using the STAR-QSM algorithm;

[0021] (3) Structural image post-processing: Use T1-weighted structural images to calculate cortical thickness.

[0022] 3D-FSPGR sequence structural images were post-processed;

[0023] (4) Statistical analysis: voxel-based whole-brain QSM analysis and cortical thickness statistical analysis were performed.

[0024] This study uses quantitative magnetic susceptibility imaging (QSM) to explore the spatial distribution pattern of cortical iron deposition in the whole brain of AD patients and its relationship with cognition and cortical thickness; to study the complex relationship between brain network topology characteristics, cortical atrophy and QSM signals in AD patients; and to further understand the mechanism of the selective susceptibility of specific neuronal groups to abnormal iron metabolism in AD by analyzing the relationship between abnormal QSM signals and gene transcription and cellular characteristics.

[0025] Preferably, in step (1), the MRI scanning sequence includes: an axial 3D-mGRE sequence and a sagittal 3D-FSPGR sequence.

[0026] Preferably, in step (1), the MRI scanning sequence acquisition parameters are as follows:

[0027] 3D-mGRE sequence: flip angle FA = 12°; multiple echo times TE1 = 3.19 ms, ΔTE = 2.37 ms, TE8 = 19.77 ms; repetition time TR = 22.9 ms; bandwidth = 62.5

[0028] Hz / pixel; slice thickness = 1.0 mm; field of view = 256 mm × 256 mm; voxel size

[0029] =1×1×1mm 3 ; Scan time = 4.4 minutes;

[0030] 3D-FSPGR sequence: FA = 12°; TE = 2.9 ms; TR = 6.7 ms; bandwidth 31.25 Hz / pixel;

[0031] Slice thickness = 1.0 mm; FOV = 256 mm × 256 mm; Voxel size = 1 × 1 × 1 mm 3 ; Scan time = 4.17 minutes.

[0032] Preferably, the step (2) comprises the following sub-steps:

[0033] (2.1) The T1 structural images of the 3D-FSPGR sequence were bias-corrected, and then the T1 structural images of all subjects were fully iterated six times to construct a subject-specific T1 structural brain template. At the same time, the individual T1 structural images were spatially registered to the subject-specific T1 structural brain template;

[0034] (2.2) The amplitude map of the first echo is bias-corrected and rigidly registered to the T1 structural image in the individual space. Then, a third-order interpolation algorithm is used to map the QSM map to the subject-specific T1 structural brain template using the two-step deformation relationship described above, completing the spatial standardization of the QSM.

[0035] (2.3) All the spatially normalized QSM images of the subjects were averaged to obtain the subject-specific QSM brain template, and the 3D-FSPGR sequence structural images were segmented.

[0036] Obtain probability maps of gray matter, white matter, and cerebrospinal fluid.

[0037] Preferably, step (3) includes: head motion correction, volume segmentation, spherical expansion, surface reconstruction, spherical registration, smoothing, visual inspection, and manual correction, and the cortical thickness is defined as the vertical distance between the gray / white matter junction and the meningeal surface.

[0038] Preferably, in the voxel-based whole-brain QSM analysis of step (4), the absolute value of QSM is used, the standardized QSM image is smoothed using a three-dimensional Gaussian smoothing kernel with a standard deviation of 3 mm, and additional smoothing compensation is performed using a smoothed gray matter mask.

[0039] Preferably, in the statistical analysis of cortical thickness in step (4), a general linear model is constructed based on surface morphological analysis, and correction is performed for age and gender; multiple comparison correction is performed using a Z-based Monte Carlo simulation algorithm, and P < 0.01 at the vertex level and P < 0.05 at the cluster level are considered to be statistically significant; after multiple comparison correction, the average cortical thickness within statistically significant clusters is extracted from the brain surface images of all subjects; the surface labels of the significant clusters are converted into volume masks and mapped to the MNI152 space; and the average magnetic susceptibility values ​​of the significant cluster areas corresponding to all subjects are extracted.

[0040] Preferably, step (4) further comprises: using GraphPad Prism 8 software for statistical drawing, using the Shapiro-Wilk test to evaluate whether the data conform to the normal distribution, using the independent sample T test and the chi-square test to compare the inter-group differences in clinical characteristics, using the Mann-Whitney U test to evaluate the inter-group differences in the absolute value of QSM in the brain atrophy area, evaluating the correlation between cortical thickness and the absolute value of QSM in the atrophy area by controlling for age through partial correlation analysis, using the Benjamini-Hochberg false discovery rate method for multiple comparison correction, according to the SBM analysis results, there are 4 significantly atrophic brain area clusters, and the significance threshold after FDR correction is set to q<0.05.

[0041] Preferably, in step (4), a multivariate regression linear model is constructed for the entire cohort and the AD group based on the MMSE and MoCA scores, and in addition to the cortical thickness and the absolute value of QSM in the brain atrophy area, age, gender and education level are also included in the multivariate linear regression model.

[0042] By adopting the technical solution of the present invention, the following results are obtained:

[0043] 1. Demographic Information and Clinical Characteristics

[0044] Twenty AD patients and eight HC subjects were excluded due to poor imaging data quality due to motion artifacts and modality loss. A total of 30 AD patients (mean age: 68.5 ± 6.8 years, 21 females) and 26 HCs (mean age: 65.5 ± 8.1 years, 19 females) were included in this study. There were no significant differences in age (P = 0.136) or sex (P = 0.799) between the AD and HC groups; all subjects were right-handed. The mean years of education were higher in the AD group than in the HC group (P = 0.015). The MMSE score in the AD group was significantly lower than that in the HC group (P < 0.001). Demographic and clinical characteristics of all subjects are summarized in Table 1.

[0045] Table 1 Demographic information and clinical characteristics of the study cohort

[0046] AD group (n=30) HC group (n=26) P-value Gender (male / female) 9 / 21 7 / 19 0.799 Age (years) 68.5±6.8 65.5±8.1 0.136 Education level (years) 11.33±3.86 8.50±4.57 0.015 MMSE score 19.77±4.79 27.96±1.64 <0.001 MoCA Scoring 16.87±4.5 NA NA

[0047] Continuous variables are expressed as mean ± standard deviation. Abbreviations: AD, Alzheimer's disease; HC, healthy control; MMSE, Mini-Mental Exam Scale; MoCA, Montreal Cognitive Assessment; NA, not applicable.

[0048] 2. Cross-sectional Analysis of QSM Values ​​in the Whole Cortex

[0049] A voxel-wise whole-brain cross-sectional analysis revealed that compared with healthy controls, AD patients exhibited widespread increases in absolute magnetic susceptibility values ​​across the cortical ribbon, with significant clusters primarily located in the left cerebral cortex, reflecting a significant increase in iron content. Left-sided abnormalities were located in the frontal lobe, including the superior frontal gyrus, middle frontal gyrus, inferior frontal gyrus, frontal pole, and precentral gyrus; in the parietal lobe, including the precuneus, postcentral gyrus, supramarginal gyrus, and angular gyrus; in the temporal lobe, including the inferior temporal gyrus, middle temporal gyrus, fusiform gyrus, and parahippocampal gyrus; and in the occipital lobe, primarily the lateral occipital lobe and occipital pole, as well as the insula, putamen, and cingulate gyrus (P < 0.05, FWE correction). Significantly abnormal clusters were also found in the deep gray matter nuclei of the basal ganglia, including the left caudate nucleus and putamen. In the AD group, although the absolute magnetic susceptibility was significantly increased mainly in the left cerebral cortex, covering most of the left frontal, parietal, temporal, and occipital lobes, relatively sparse areas within the right cerebral cortex also showed abnormally increased absolute magnetic susceptibility, including the right precuneus, cingulate gyrus, lateral occipital lobe, occipital pole, insula, putamen, postcentral gyrus, and precentral gyrus (P < 0.05, FWE correction). Notably, the magnetic susceptibility of Crus II, VIIb, and VIIIa areas of the cerebellar cortex was also significantly increased in the AD group (P < 0.05, FWE correction). Finally, no significant clusters of decreased absolute QSM values ​​were found in the AD group. See Table 2 for details.

[0050] Table 2 Brain regions with significant differences in QSM between AD and HC groups

[0051]

[0052]

[0053] Continuous variables are expressed as mean ± standard deviation. Only significant clusters with more than 30 voxels are listed. Abbreviations: AD, Alzheimer's disease; HC, healthy controls; FWE, fractional error fraction; MNI, Montreal Neurological Institute; ppm, parts per million.

[0054] 3. Whole-cortical QSM regression analysis

[0055] Regression analysis of whole-brain QSM with cognitive scales revealed that increased absolute susceptibility in the right angular gyrus, supramarginal gyrus, lateral occipital lobe, parietal operculum, left frontal operculum, and inferior frontal gyrus was associated with decreased MMSE scores in AD patients (TFCE, FWE-corrected P < 0.05). Furthermore, susceptibility in the right angular gyrus, supramarginal gyrus, lateral occipital lobe, superior temporal gyrus, and left superior and inferior frontal gyri and frontal pole was negatively correlated with MoCA scores in AD patients (TFCE, FWE-corrected P < 0.05). Interestingly, the results of regression analysis of QSM with MoCA and MMSE scores showed similar spatial distribution patterns across brain regions, as shown in Table 3. However, unlike the cross-sectional analysis of whole-brain QSM between-group results, the cognitive-related QSM regression results were primarily localized in the right cerebral hemisphere. In contrast, at the whole-brain voxel level, no significant clusters were found with positive correlations between absolute susceptibility and MMSE or MoCA scores.

[0056] Table 3 Significant clusters in the regression analysis of whole-brain QSM, MMSE and MoCA scores

[0057]

[0058] Only significant clusters with a voxel count greater than 30 are listed. Abbreviations: AD, Alzheimer's disease; HC, healthy controls; FWE, fractional error rate; MNI, Montreal Neurological Institute; MMSE, Mini-Mental Exam Scale; MoCA, Montreal Cognitive Assessment.

[0059] 4. Brain Atrophy Measurement

[0060] Morphological analysis based on brain surface showed that the cortical thickness of multiple brain regions in the AD group was significantly reduced, including the left superior temporal gyrus (AD: 2.74 ± 0.28 mm [95% CI 2.63-2.84]; HC: 3.01 ± 0.16 mm [95% CI 2.94-3.07]), the right fusiform gyrus (AD: 2.76 ± 0.3 mm [95% CI 2.64-2.87]; HC: 3.11 ± 0.20 mm [95% CI 3.03-3.19]), the pars opercularis cortex (AD: 2.42 ± 0.17 mm [95% CI 2.36-2.49]; HC: 2.63 ± 0.13 mm [95% CI 2.58-2.67]), and the frontal pole (AD: 2.30 ± 0.23 mm [95% CI 2.21-2.38]; HC: 2.55±0.21 mm [95% CI 2.47-2.63]) (Monte Carlo Z-corrected P<0.05; Table 4). At the whole-brain vertex level, there were no regions with significantly increased cortical thickness in the AD group compared with the HC group.

[0061] Table 4 Clusters with significant differences in cortical thickness between the AD and HC groups

[0062]

[0063] Continuous variables are expressed as mean ± standard deviation. Abbreviations: AD, Alzheimer's disease; HC, healthy control; CWP, cluster-level P value; MNI, Montreal Neurological Institute.

[0064] 5. Relationship between QSM and cortical thickness

[0065] In further analysis within the regions of brain atrophy, the absolute magnetic susceptibility of the right frontal pole (AD: 0.034±0.007 ppm [95% CI 0.032-0.037]; HC: 0.030±0.005 ppm [95% CI 0.028-0.032]; P=0.016) and right frontal operculum (AD: 0.020±0.003 ppm [95% CI 0.018-0.021]; HC: 0.017±0.002 ppm [95% CI 0.017-0.018]; P=0.002) was significantly higher in the AD group than in the HC group (FDR-corrected P<0.05). However, there were no significant differences in absolute magnetic susceptibility between the two groups in the left superior temporal gyrus (AD: 0.027 ± 0.004 ppm [95% CI 0.026-0.029]; HC: 0.026 ± 0.003 ppm [95% CI 0.024-0.027]; P = 0.168) and the right fusiform gyrus (AD: 0.039 ± 0.006 ppm [95% CI 0.037-0.041]; HC: 0.036 ± 0.004 ppm [95% CI 0.034-0.037]; P = 0.057). Partial correlation analysis between QSM values ​​and cortical thickness showed that absolute magnetic susceptibility was significantly negatively correlated with cortical thickness in the right frontal operculum in the AD group (r = -0.510, P = 0.005, after FDR correction). In the entire study cohort, absolute magnetic susceptibility was negatively correlated with cortical thickness in the right fusiform gyrus (r = -0.436, P = 0.001) and right frontal operculum cortex (r = -0.521, P < 0.001) (FDR-corrected P < 0.05). However, in the right frontal pole and left superior temporal gyrus, absolute magnetic susceptibility was not significantly correlated with cortical thickness in the entire cohort (right frontal pole: r = -0.183, P = 0.182; left superior temporal gyrus: r = -0.191, P = 0.162) or in the AD group (right frontal pole: r = 0.013, P = 0.945; left superior temporal gyrus: r = -0.285, P = 0.134).

[0066] 6. Relationship between QSM, cortical thickness and cognition

[0067] In the MMSE multiple linear regression model for the entire cohort The cortical thickness of the left superior temporal gyrus (β=6.868; 95% CI: 2.073, 11.662; P=0.006) and the right fusiform gyrus (β=4.833; 95% CI: 1.098, 8.668; P=0.012) and the absolute magnetic susceptibility of the right frontal operculum (β=-642.135; 95% CI: -1048.664, -253.606; P=0.003) were independent factors affecting cognitive level (Table 5). The absolute magnetic susceptibility of the right frontal operculum was negatively correlated with the cognitive level of the entire group, with a high absolute standardized coefficient (absolute standardized coefficient) of 0.331.

[0068] Table 5 Multiple linear regression model of MMSE for the entire cohort

[0069]

[0070]

[0071] Abbreviations: MMSE, Mini-Mental State Examination; VIF: variance inflation factor; Adjusted R-squared

[0072] Furthermore, multiple linear regression analysis of the MMSE in the AD group (R2 adj = 0.599) revealed that left superior temporal gyrus cortical thickness (β = 7.163; 95% CI: 2.632, 11.694; P = 0.003), right frontal operculum absolute magnetic susceptibility (β = -1024.311; 95% CI: -1498.626, -549.996; P < 0.001), and left superior temporal gyrus magnetic susceptibility (β = 456.907; 95% CI: 104.445, 809.369; P = 0.013) were independent factors influencing cognitive ability (Table 6). Notably, right frontal operculum absolute magnetic susceptibility also showed a significant negative correlation with MMSE in AD patients, with the highest absolute standardized regression coefficient of 0.680. In contrast, left superior temporal gyrus absolute magnetic susceptibility showed a positive correlation with MMSE.

[0073] Table 6 Multiple linear regression model of MMSE in the AD group

[0074]

[0075] Abbreviations: MMSE, Mini-Mental State Examination; VIF: variance inflation factor; Adjusted R-squared

[0076] In the MoCA regression model of the AD group The thickness of the left superior temporal gyrus cortex (β=8.63; 95% CI=5.301, 11.959; P<0.001), education level (β=0.464; 95% CI=0.213, 0.716; P=0.001) and gender (β=2.845; 95% CI=0.818, 4.872; P=0.008) were significant predictors of cognitive ability, among which the absolute standardized regression coefficient of the thickness of the left superior temporal gyrus cortex was the highest, at 0.542 (Table 7).

[0077] Table 7 Multiple linear regression model of MoCA in the AD group

[0078]

[0079] Abbreviations: MoCA, Montreal Cognitive Assessment; VIF: variance inflation factor; Adjusted R-squared

[0080] Excessive iron deposition in the brain has been consistently found in AD patients. Cortical thinning, as the most critical structural imaging feature of AD, has been considered to be associated with cognitive impairment. However, the relationship between cortical iron content, cortical thickness and cognitive decline remains unclear. In this study, the spatial distribution pattern of magnetic susceptibility changes in the whole brain cortex was obtained, and the relationship between magnetic susceptibility and cognitive severity in AD patients was explored. In addition, the correlation between magnetic susceptibility and cortical thickness in these patients was further evaluated, as well as the contribution of magnetic susceptibility and cortical thickness to cognitive impairment. The main findings of the present invention are: (1) In AD patients, the magnetic susceptibility of the whole brain cortical belt is widely increased and asymmetrically distributed in the left cerebral cortex, deep gray matter nuclei and part of the cerebellar cortex; (2) The increase in magnetic susceptibility in the right parietal cortex and lateral occipital lobe of AD patients is associated with cognitive decline; (3) The increase in magnetic susceptibility is associated with a decrease in cortical thickness in the right frontal operculum cortex; (4) The magnetic susceptibility in the right frontal operculum atrophy area may be an independent influencing factor leading to the decrease in MMSE score.

[0081] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention are still within the scope of protection of the technical solution of the present invention.

Claims

1. A method for constructing and analyzing a magnetic susceptibility map of diamagnetic components in gray matter brain regions, characterized by: It includes the following steps: (1) MRI data acquisition: The subjects underwent MR data acquisition on a magnetic resonance scanner; (2) QSM reconstruction and post-processing: Use the Laplacian-based phase unwrapping method to unwrap the multi-echo phase image; use the built-in BET algorithm of FSL software to perform brain stripping on the amplitude map of the first echo and generate a binary brain mask map; use V-SHARP algorithm and combined with the brain mask to remove the background field of the unwrapped phase image; use The STAR-QSM algorithm calculates the magnetic susceptibility map from the local field map; (3) Structural image post-processing: Use T1-weighted structural images to calculate cortical thickness. 3D-FSPGR sequence structural images were post-processed; (4) Statistical analysis: voxel-based whole-brain QSM analysis and cortical thickness statistical analysis were performed.

2. The method for constructing and analyzing the diamagnetic susceptibility map of gray matter brain regions according to claim 1, characterized in that: In the step (1), the MRI scanning sequence includes: an axial 3D-mGRE sequence and a sagittal 3D-FSPGR sequence.

3. The method for constructing and analyzing the magnetic susceptibility map of the diamagnetic components of the gray matter brain region according to claim 2, Its characteristics are: In step (1), the MRI scanning sequence acquisition parameters are as follows: 3D-mGRE sequence: flip angle FA = 12°; multiple echo times TE1 = 3.19 ms, ΔTE = 2.37 ms, TE8 = 19.77 ms; repetition time TR = 22.9 ms; bandwidth = 62.5 Hz / pixel; slice thickness = 1.0 mm; field of view = 256 mm × 256 mm; voxel size = 1 × 1 × 1 mm 3 ; Scan time = 4.4 minutes; 3D-FSPGR sequence: FA = 12°; TE = 2.9 ms; TR = 6.7 ms; bandwidth 31.25 Hz / pixel; slice thickness = 1.0 mm; FOV = 256 mm × 256 mm; voxel size = 1 × 1 × 1 mm 3 ; Scan time = 4.17 minutes.

4. The method for constructing and analyzing the diamagnetic susceptibility map of gray matter brain regions according to claim 3, characterized in that: The step (2) comprises the following sub-steps: (2.1) The T1 structural images of the 3D-FSPGR sequence were bias-corrected, and then the T1 structural images of all subjects were fully iterated six times to construct a subject-specific T1 structural brain template. At the same time, the individual T1 structural images were spatially registered to the subject-specific T1 structural brain template; (2.2) The amplitude map of the first echo is bias-corrected and rigidly registered to the T1 structural image in the individual space. Then, a third-order interpolation algorithm is used to map the QSM map to the subject-specific T1 structural brain template using the two-step deformation relationship described above, completing the spatial standardization of the QSM. (2.3) The spatially normalized QSM images of all subjects were averaged to obtain a subject-specific QSM brain template. The 3D-FSPGR sequence structural images were segmented to obtain probability maps of gray matter, white matter, and cerebrospinal fluid.

5. The method for constructing and analyzing the diamagnetic susceptibility map of gray matter brain regions according to claim 4, characterized in that: The step (3) includes: head motion correction, volume segmentation, spherical expansion, surface reconstruction, spherical registration, smoothing, visual inspection, and manual correction. The cortical thickness is defined as the vertical distance between the gray / white matter junction and the meningeal surface.

6. The method for constructing and analyzing the diamagnetic susceptibility map of gray matter brain regions according to claim 5, characterized in that: In the voxel-based whole-brain QSM analysis of step (4), the absolute value of QSM is used, and the standardized QSM image is smoothed using a three-dimensional Gaussian smoothing kernel with a standard deviation of 3 mm, and an additional smoothing compensation is performed using a smoothed gray matter mask.

7. The method for constructing and analyzing the diamagnetic susceptibility map of gray matter brain regions according to claim 6, characterized in that: In the statistical analysis of cortical thickness in step (4), a general linear model was constructed based on surface morphological analysis, and corrections were made for age and gender; multiple comparison corrections were made using a Z-based Monte Carlo simulation algorithm, and values ​​of P < 0.01 at the vertex level and P < 0.05 at the cluster level were considered statistically significant; after multiple comparison corrections, the average cortical thickness within statistically significant clusters was extracted from the brain surface images of all subjects; the surface labels of the significant clusters were converted into volume masks and mapped to the MNI152 space; and the average magnetic susceptibility values ​​of the significant cluster areas corresponding to all subjects were extracted.

8. The method for constructing and analyzing the diamagnetic susceptibility map of gray matter brain regions according to claim 7, characterized in that: The step (4) further includes: using GraphPad Prism 8 software for statistical drawing, using the Shapiro-Wilk test to evaluate whether the data conform to the normal distribution, using the independent sample T test and the chi-square test to compare the inter-group differences in clinical characteristics, using the Mann-Whitney U test to evaluate the inter-group differences in the absolute value of QSM in the brain atrophy area, evaluating the correlation between cortical thickness and the absolute value of QSM in the atrophy area through partial correlation analysis controlling for age, using the Benjamini-Hochberg false discovery rate method to perform multiple comparison correction, according to the SBM analysis results, there are 4 significantly atrophic brain area clusters, and the significance threshold after FDR correction is set to q<0.

05.

9. The method for constructing and analyzing the diamagnetic susceptibility map of gray matter brain regions according to claim 8, characterized in that: In step (4), a multivariate linear regression model was constructed for the entire cohort and the AD group based on the MMSE and MoCA scores. In addition to the cortical thickness and the absolute value of QSM in the brain atrophy area, age, gender, and education level were also included in the multivariate linear regression model.

10. Use of the method for constructing and analyzing the diamagnetic susceptibility map of gray matter brain regions according to any one of claims 1 to 9, characterized in that: It is used to monitor the progression of AD and assess the severity of cognitive decline.

Images