Method and device for evaluating cerebrospinal fluid dynamic abnormalities in patients with SCA3

Through MRI image processing technology, including automatic segmentation and DTI analysis, the problem of cerebrospinal fluid dynamic abnormality assessment in SCA3 is solved, and the accurate quantification of the generation and flow of cerebrospinal fluid in SCA3 patients is achieved, and the disease mechanism is deeply understood, providing early diagnosis and treatment. New method.

CN118918130BActive Publication Date: 2024-12-03CHINA JAPAN FRIENDSHIP HOSPITAL
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411404814.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-10-10
Publication Date
2024-12-03
Estimated Expiration
2044-10-10

AI Technical Summary

Technical Problem

The prior art studies on the glial lymphatic system of spinal cerebellar ataxia type 3 (SCA3) are relatively limited, and it is difficult to accurately evaluate cerebrospinal fluid dynamic abnormalities, affecting the early diagnosis and treatment of the disease.

Method used

By acquiring high-resolution MRI images, using Freesurfer and FSL software to perform automatic segmentation and volume calculation of cerebrospinal fluid, combined with DTI image analysis, revealing choroid plexus dysfunction and brain lymphatic system activity patterns, comparing the volume and volume between healthy groups and SCA3 patients DTI-ALPS index differences.

Benefits of technology

It has achieved accurate quantification of the generation and flow of cerebrospinal fluid in SCA3 patients, a deep understanding of the pathogenesis of the disease, and provided new ideas and methods for the early diagnosis and treatment of the disease.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118918130B_ABST
    Figure CN118918130B_ABST
Patent Text Reader

Abstract

The present invention discloses a method and device for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients. By acquiring T1-weighted structural MRI images as the basic data for processing, CP automatic segmentation, CSF automatic segmentation, segmentation result review and correction, volume normalization processing, the quantitative difference of choroid plexus volume between healthy group HCs and SCA3 patient group is compared. The quantitative difference of cerebrospinal fluid volume between healthy group HCs and SCA3 patient group is compared. By acquiring DTI images of subjects, spatial normalization of fractional anisotropy images, defining regions of interest in different regions, extracting diffusion values ​​in different regions, calculating DTI-ALPS index in different regions, regional subdivision and gradient calculation, and using the two-sample t-test method to compare the difference in regional DTI-ALPS index between HCs and patients, the abnormalities of patients' cerebrospinal fluid, regional cerebral glymphatic system in the brain, and PVS flow of fluid in the brain can be revealed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to a method for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients, and a device for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients. Background Art

[0002] Spinocerebellar ataxia type 3 (also known as Machado-Joseph disease, abbreviated as SCA3 / MJD) is a common autosomal dominant cerebellar ataxia among neurodegenerative diseases. It usually manifests as motor and non-motor symptoms, such as pyramidal signs, extrapyramidal symptoms, cognitive decline, and rapid eye movement behavior disorder. The pathogenesis of the disease is related to protein misfolding, especially the expansion of the CAG repeat sequence in the coding region of the ATXN3 gene on chromosome 14, which leads to the production of abnormal polyglutamine proteins. This protein imbalance impairs cell function and triggers neuronal death, while enhancing pathological processes such as inflammatory response, mitochondrial dysfunction, and imbalance in the regulation of cellular stress response. In the pathological process of SCA3, maintaining the homeostasis of the brain environment is crucial for the clearance of abnormal proteins, because imbalance affects the cellular environment, triggers neuroinflammation, and drives the complex pathological evolution and progression of the disease. The glymphatic system, as an important waste removal mechanism of the brain, is critical for maintaining the health of the nervous system.

[0003] However, research on the glymphatic system in SCA3 is still quite limited. Summary of the invention

[0004] In order to overcome the defects of the prior art, the technical problem to be solved by the present invention is to provide a method for evaluating the dynamic abnormalities of cerebrospinal fluid in SCA3 patients, which can accurately and efficiently quantify the CP and CSF volumes of SCA3 patients, reveal the choroid plexus dysfunction that may exist in SCA3, reveal the abnormalities of PVS flow in SCA3 patients, reveal the abnormal patterns of regional cerebral glymphatic system activity in the brain of SCA3 patients and their correlation with the disease process, which not only helps to deeply understand the pathogenesis of SCA3, but also may provide new ideas and methods for the early diagnosis and treatment of the disease.

[0005] The technical solution of the present invention is: the method for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients comprises the following steps:

[0006] (1) Obtain high-resolution T1 images and DTI images as basic data for processing: Have the subject enter the scanning room of the MRI scanner and ensure that his head is fixed to reduce motion artifacts. Check that the acquired high-resolution T1 images and DTI image data are complete; convert the MRI image data into an image file in NIfTI format; use image processing methods to remove non-brain structures in the MRI image;

[0007] (2) Automatic CP segmentation: Using Freesurfer’s cortical reconstruction method, each voxel is classified into a specific anatomical structure from T1-weighted structural MRI images using prior anatomical knowledge and probabilistic models;

[0008] (3) Automatic CSF segmentation: FAST, mixed Gaussian model and offset field correction method in FSL software are used to segment brain images into different tissue types: CSF, gray matter and white matter. The maximum expectation EM algorithm is used to iteratively estimate the probability of each voxel belonging to each tissue type, and offset field correction is performed to deal with imaging inhomogeneities.

[0009] (4) Segmentation result review and correction: Manual review and correction of the CP area obtained by segmentation to ensure the accuracy and reliability of the segmentation results, including manual boundary editing, noise removal, and hole filling operations;

[0010] (5) Volume standardization: For CP volume calculation, FreeSurfer generates a file containing the segmentation results of all brain regions after segmentation. The file contains the labels of each anatomical structure, and the volume information of the choroid plexus region is extracted using FreeSurfer's statistical tools. For CSF volume calculation, a CSF probability image is used. Each voxel value of the image represents the probability that the voxel belongs to the CSF, and the image is standardized to reduce individual differences. The total intracranial volume (ICV) is used as a reference to calculate the ratio of CP, CSF and ICV.

[0011] (6) Comparison of volume differences between the healthy group and the SCA3 patient group: The two-sample t-test method was used to compare the differences in CP volume and CSF volume between HCs and SCA3 patients;

[0012] (7) Spatial standardization of fractional anisotropy FA images: First, the FA map and other diffusion index maps were calculated using the dtifit tool in the FSL software. Second, the FA map or other diffusion index maps of each subject were registered to the FA map of the JHU atlas template using the FLIRT tool in FSL. The resulting transformation matrix was applied to other diffusion index maps to ensure that all DTI base maps were located in a unified JHU atlas space with a spatial resolution of 1 × 1 × 1 mm³.

[0013] (8) Define ROIs in different regions: determine ROI location and reference coordinates; define ROI shape and size; create ROI;

[0014] (9) Extract diffusion values ​​of different regions: Use the fslstats tool of FSL to extract the average values ​​of Dxx,proj, Dxx,assoc, Dyy,proj, and Dzz,assoc from each ROI of the front, middle, and back regions. Dxx,proj is the diffusion coefficient of water molecules in the projected fiber region along the x direction; Dxx,assoc is the diffusion coefficient of water molecules in the contact fiber region along the x direction; Dyy,proj is the diffusion coefficient of water molecules in the projected fiber region along the y direction; Dzz,assoc is the diffusion coefficient of water molecules in the projected fiber region along the z direction;

[0015] (10) Calculation of DTI-ALPS index in different regions: The DTI-ALPS index is defined as the ratio of Dxx,proj and Dxx,assoc to Dyy,proj and Dzz,assoc, and its formula is:

[0016] ,

[0017] Calculate the DTI-ALPS index of each region: calculate the anterior DTI-ALPS, middle DTI-ALPS, and posterior DTI-ALPS separately;

[0018] (11) Regional subdivision and gradient calculation: DTI-ALPS gradients were calculated from a ROI of 3 × 3 × 3 mm³ at 1 mm intervals;

[0019] (12) The two-sample t-test method was used to compare the differences in regional DTI-ALPS index between HCs and SCA3 patients.

[0020] The present invention obtains T1-weighted structural MRI images as basic data for processing, automatically segments CP, automatically segments CSF, reviews and corrects segmentation results, and performs volume normalization processing to compare the quantitative differences in choroid plexus volume between the healthy group HCs and the SCA3 patient group. The choroid plexus is the main site of cerebrospinal fluid production, and the volume change of the choroid plexus can reflect the production capacity of cerebrospinal fluid. By comparing the quantitative differences in cerebrospinal fluid volume between the healthy group HCs and the SCA3 patient group, the change in cerebrospinal fluid volume can reflect the health status of the brain. If the cerebrospinal fluid volume increases or decreases abnormally, it may indicate pathological changes in the brain. By acquiring the DTI images of the subjects, spatial standardization of the fractional anisotropy images, defining different regional regions of interest, extracting the diffusion values ​​of different regions, calculating the DTI-ALPS index of different regions, regional subdivision and gradient calculation, and using the two-sample t-test method to compare the differences in regional DTI-ALPS index between HCs and SCA3 patients, the cerebrospinal fluid of SCA3 patients can reveal the regional cerebral lymphatic system of SCA3 patients and the abnormality of the PVS flow of brain fluid in SCA3 patients, as well as the correlation between the abnormal patterns of these activities and the disease process. Combining these two technologies, the present invention can not only accurately quantify the generation and flow of cerebrospinal fluid, but also help to deeply understand the pathogenesis of SCA3, and provide new ideas and methods for the early diagnosis and treatment of the disease.

[0021] Also provided is a device for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients, comprising:

[0022] An MRI image acquisition module configured to acquire T1-weighted structural MRI images as basic data for processing: checking the integrity of acquired T1-weighted structural MRI image data; converting the MRI image data into an image file in NIfTI format; and removing non-brain structures in the MRI image using an image processing method;

[0023] A CP segmentation module, which is configured to perform automatic CP segmentation: using Freesurfer's cortical reconstruction method, from T1-weighted structural MRI images, using prior anatomical knowledge and probabilistic models to classify each voxel into a specific anatomical structure;

[0024] CSF segmentation module, which is configured to perform automatic CSF segmentation: brain images are segmented into different tissue types: CSF, gray matter, and white matter using FAST, mixed Gaussian models, and offset field correction methods in FSL software, the probability of each voxel belonging to each tissue type is iteratively estimated by the maximum expectation EM algorithm, and offset field correction is performed to cope with inhomogeneities in imaging;

[0025] The review and correction module is configured to perform segmentation result review and correction: manually review and correct the CP area obtained by segmentation to ensure the accuracy and reliability of the segmentation results, including manual boundary editing, noise removal, and hole filling operations;

[0026] The volume processing module is configured to perform volume normalization processing: for CP volume calculation, FreeSurfer generates a file containing the segmentation results of all brain regions after segmentation is completed. The file contains the labels of various anatomical structures, and the volume information of the choroid plexus region is extracted using FreeSurfer's statistical tools; for CSF volume calculation, a CSF probability image is used, in which each voxel value represents the probability that the voxel belongs to the CSF, and the image is normalized to reduce individual differences; the total intracranial volume ICV is used as a reference to calculate the ratio of CP, CSF and ICV;

[0027] The first comparison module, which was configured to compare the volume differences between the healthy group and the SCA3 patient group: the differences in CP volume and CSF volume between HCs and SCA3 patients were compared using a two-sample t-test method;

[0028] A spatial normalization module configured to first calculate FA maps and other diffusion index maps using the dtifit tool in the FSL software; second, to register each subject's FA map or other diffusion index map to the FA map of the JHU atlas template using FSL's FLIRT tool, and to apply the resulting transformation matrix to the other diffusion index maps to ensure that all DTI base maps are in a unified JHU atlas space with a spatial resolution of 1 × 1 × 1 mm³;

[0029] Region of Interest module, which is configured to determine the ROI location, reference coordinates; define the ROI shape and size; create the ROI;

[0030] Extraction module, which is configured to use FSL's fslstats tool to extract the average values ​​of Dxx,proj, Dxx,assoc, Dyy,proj, and Dzz,assoc from each ROI in the front, middle, and back regions, where Dxx,proj is the water molecule diffusion coefficient in the projected fiber region along the x direction, Dxx,assoc is the water molecule diffusion coefficient in the contact fiber region along the x direction; Dyy,proj is the water molecule diffusion coefficient in the projected fiber region along the y direction; and Dzz,assoc is the water molecule diffusion coefficient in the projected fiber region along the z direction;

[0031] The DTI-ALPS index calculation module is configured to define the DTI-ALPS index as the ratio of Dxx,proj and Dxx,assoc to Dyy,proj and Dzz,assoc, and the formula is:

[0032] ,

[0033] Calculate the DTI-ALPS index of each region: calculate the anterior DTI-ALPS, middle DTI-ALPS, and posterior DTI-ALPS separately;

[0034] Segmentation and gradient calculation module, which is configured to calculate DTI-ALPS gradients from a ROI of size 3×3×3 mm³ at 1 mm intervals;

[0035] The second comparison module was configured to compare the differences in regional DTI-ALPS index between HCs and SCA3 patients using a two-sample t-test method. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A flow chart showing a method for evaluating abnormal cerebrospinal fluid dynamics in SCA3 patients according to the present invention. DETAILED DESCRIPTION

[0037] The choroid plexus (CP) is the main secretion site of cerebrospinal fluid (CSF) and plays an important role in maintaining the homeostasis of the brain environment. By quantifying the CP and CSF volumes of SCA3 patients, we can reveal the possible choroid plexus dysfunction in SCA3 and further understand the impact of the disease on the brain's waste removal mechanism. Using medical image processing technology, especially automated segmentation and volume calculation methods for MRI images, the CP and CSF volumes of SCA3 patients can be accurately and efficiently quantified. This will not only help to deeply understand the pathogenesis of SCA3, but may also provide new ideas and methods for the early diagnosis and treatment of the disease.

[0038] like Figure 1 As shown, this method for evaluating the dynamic abnormalities of cerebrospinal fluid in patients with SCA3 includes the following steps:

[0039] (1) Obtain high-resolution T1 images and DTI images as basic data for processing: Have the subject enter the scanning room of the MRI scanner and ensure that his head is fixed to reduce motion artifacts. Check that the acquired high-resolution T1 images and DTI image data are complete; convert the MRI image data into an image file in NIfTI format; use image processing methods to remove non-brain structures in the MRI image;

[0040] (2) Automatic CP segmentation: Using Freesurfer’s cortical reconstruction method, each voxel is classified into a specific anatomical structure from T1-weighted structural MRI images using prior anatomical knowledge and probabilistic models;

[0041] (3) Automatic CSF segmentation: FAST, mixed Gaussian model and offset field correction method in FSL software are used to segment brain images into different tissue types: CSF, gray matter and white matter. The maximum expectation EM algorithm is used to iteratively estimate the probability of each voxel belonging to each tissue type, and offset field correction is performed to deal with imaging inhomogeneities.

[0042] (4) Segmentation result review and correction: Manual review and correction of the CP area obtained by segmentation to ensure the accuracy and reliability of the segmentation results, including manual boundary editing, noise removal, and hole filling operations;

[0043] (5) Volume standardization: For CP volume calculation, FreeSurfer generates a file containing the segmentation results of all brain regions after segmentation. The file contains the labels of each anatomical structure, and the volume information of the choroid plexus region is extracted using FreeSurfer's statistical tools. For CSF volume calculation, a CSF probability image is used. Each voxel value of the image represents the probability that the voxel belongs to the CSF, and the image is standardized to reduce individual differences. The total intracranial volume (ICV) is used as a reference to calculate the ratio of CP, CSF and ICV.

[0044] (6) Comparison of volume differences between the healthy group and the SCA3 patient group: The two-sample t-test method was used to compare the differences in CP volume and CSF volume between HCs and SCA3 patients;

[0045] (7) Spatial standardization of fractional anisotropy FA images: First, the FA map and other diffusion index maps were calculated using the dtifit tool in the FSL software. Second, the FA map or other diffusion index maps of each subject were registered to the FA map of the JHU atlas template using the FLIRT tool in FSL. The resulting transformation matrix was applied to other diffusion index maps to ensure that all DTI base maps were located in a unified JHU atlas space with a spatial resolution of 1 × 1 × 1 mm³.

[0046] (8) Define ROIs in different regions: determine ROI location and reference coordinates; define ROI shape and size; create ROI;

[0047] (9) Extract diffusion values ​​of different regions: Use the fslstats tool of FSL to extract the average values ​​of Dxx,proj, Dxx,assoc, Dyy,proj, and Dzz,assoc from each ROI of the front, middle, and back regions. Dxx,proj is the diffusion coefficient of water molecules in the projected fiber region along the x direction; Dxx,assoc is the diffusion coefficient of water molecules in the contact fiber region along the x direction; Dyy,proj is the diffusion coefficient of water molecules in the projected fiber region along the y direction; Dzz,assoc is the diffusion coefficient of water molecules in the projected fiber region along the z direction;

[0048] (10) Calculation of DTI-ALPS index in different regions: The DTI-ALPS index is defined as the ratio of Dxx,proj and Dxx,assoc to Dyy,proj and Dzz,assoc, and its formula is:

[0049] ,

[0050] Calculate the DTI-ALPS index of each region: calculate the anterior DTI-ALPS, middle DTI-ALPS, and posterior DTI-ALPS separately;

[0051] (11) Regional subdivision and gradient calculation: DTI-ALPS gradients were calculated from a ROI of 3 × 3 × 3 mm³ at 1 mm intervals;

[0052] (12) The two-sample t-test method was used to compare the differences in regional DTI-ALPS index between HCs and SCA3 patients.

[0053] The present invention obtains T1-weighted structural MRI images as basic data for processing, automatically segments CP, automatically segments CSF, reviews and corrects segmentation results, and performs volume normalization processing to compare the quantitative differences in choroid plexus volume between the healthy group HCs and the SCA3 patient group. The choroid plexus is the main site of cerebrospinal fluid production, and the volume change of the choroid plexus can reflect the production capacity of cerebrospinal fluid. By comparing the quantitative differences in cerebrospinal fluid volume between the healthy group HCs and the SCA3 patient group, the change in cerebrospinal fluid volume can reflect the health status of the brain. If the cerebrospinal fluid volume increases or decreases abnormally, it may indicate pathological changes in the brain. By acquiring the DTI images of the subjects, spatial standardization of the fractional anisotropy images, defining different regional regions of interest, extracting the diffusion values ​​of different regions, calculating the DTI-ALPS index of different regions, regional subdivision and gradient calculation, and using the two-sample t-test method to compare the differences in regional DTI-ALPS index between HCs and SCA3 patients, the cerebrospinal fluid of SCA3 patients can reveal the regional cerebral lymphatic system of SCA3 patients and the abnormality of the PVS flow of brain fluid in SCA3 patients, as well as the correlation between the abnormal patterns of these activities and the disease process. Combining these two technologies, the present invention can not only accurately quantify the generation and flow of cerebrospinal fluid, but also help to deeply understand the pathogenesis of SCA3, and provide new ideas and methods for the early diagnosis and treatment of the disease.

[0054] Preferably, in step (1), the parameters are set as follows: TR=8028ms, TE=81.8ms, flip angle=90°, slice thickness=2mm, FOV=240×240mm², matrix=120×120, voxel size=2×2×2mm³; 8 images with b=0s / mm² and 64 images with b=1000s / mm² are acquired; the DTI data are imported into the FSL software, and the non-brain tissue part in the DTI data is removed using the BET tool in FSL; the DTI data is motion corrected using the eddy_correct tool of FSL to eliminate the influence of head movement on image quality; the image distortion caused by magnetic field inhomogeneity is eliminated using eddy current correction, and the distortion specific to the EPI sequence is corrected using EPI distortion correction; the DTI data is biased field corrected using the FUGUE tool of FSL to eliminate the signal intensity change caused by magnetic field inhomogeneity.

[0055] Preferably, in step (2), the choroid plexus is segmented by identifying and extracting the CP area through a probability model and an algorithm and template of image features, thereby segmenting the choroid plexus CP in the lateral ventricle.

[0056] Preferably, in step (4), a neuroimaging expert performs an independent quality assessment and final confirmation on the corrected CP segmentation result.

[0057] Preferably, in step (5), the FreeSurfer statistical tool reads the segmentation result file and calculates the number and volume of voxels corresponding to each label; the fslstats tool calculates the statistical information of the image, including the number and volume of voxels.

[0058] Preferably, in step (6), whether there is a statistically significant difference in CP volume and CSF volume between the two groups of people is evaluated by testing whether there is a significant difference in the means between the two independent samples; a two-sided t-test is performed, and potential covariates such as gender, age and education level are regressed.

[0059] Preferably, in step (7), the FA image is registered to the JHU atlas template, including: linear alignment and nonlinear alignment, using the FLIRT and FNIRT tools in FSL; initializing linear alignment: using FLIRT to perform preliminary linear alignment, roughly aligning the subject's FA image to the JHU template, initializing linear alignment principle: FLIRT uses 12 degrees of freedom affine transformation for alignment, including translation, rotation, scaling and shearing transformations, and it finds the optimal transformation parameters by maximizing the similarity between the two images to align the input image with the reference image; nonlinear alignment: using FNIRT to perform nonlinear alignment, finely aligning the FA image to the JHU template, FNIRT uses nonlinear transformation to accurately align each voxel of the input image to the reference image, and uses a regularized deformation field to describe the complex deformation between images. FNIRT achieves accurate alignment by maximizing the similarity of the registered images while minimizing the complexity of the deformation field.

[0060] Preferably, in step (8), the ROI position and reference coordinates are determined as follows: Select the z = 26 horizontal plane at the level of the lateral ventricle in the MNI space, and use the labels in the JHU white matter atlas to locate the areas of projection fibers and communication fibers; Define the ROI shape and size: Use the tools in FSL to define a spherical ROI with a diameter of 3 mm, and set the shape to a spherical area with a diameter of 3 mm, with a volume of a binary mask of 3×3×3 mm³; Use atlas labels and standard spatial coordinates to create ROIs, and place the ROIs in the anterior, middle, and posterior areas of the lateral ventricle respectively.

[0061] Preferably, in step (11), a total of 46 pairs of ROIs are generated, the DTI-ALPS of each pair of ROIs is calculated, all calculated DTI-ALPS indices and gradient results are summarized, and the rate of change of the DTI-ALPS index is calculated to form a comprehensive assessment of the activity of the lymphatic system; in step (12), a two-sample t-test is used to compare the inter-group differences in anterior DTI-ALPS, middle DTI-ALPS and posterior DTI-ALPS, and potential covariates such as gender, age and education level are regressed.

[0062] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes the steps of the above-mentioned embodiment method, and the storage medium can be: ROM / RAM, disk, CD, memory card, etc. Therefore, corresponding to the method of the present invention, the present invention also includes an evaluation device for dynamic abnormalities of cerebrospinal fluid in SCA3 patients, which is usually represented in the form of functional modules corresponding to the steps of the method. The device includes:

[0063] An MRI image acquisition module configured to acquire T1-weighted structural MRI images as basic data for processing: checking the integrity of acquired T1-weighted structural MRI image data; converting the MRI image data into an image file in NIfTI format; and removing non-brain structures in the MRI image using an image processing method;

[0064] A CP segmentation module, which is configured to perform automatic CP segmentation: using Freesurfer's cortical reconstruction method, from T1-weighted structural MRI images, using prior anatomical knowledge and probabilistic models to classify each voxel into a specific anatomical structure;

[0065] CSF segmentation module, which is configured to perform automatic CSF segmentation: brain images are segmented into different tissue types: CSF, gray matter, and white matter using FAST, mixed Gaussian models, and offset field correction methods in FSL software, the probability of each voxel belonging to each tissue type is iteratively estimated by the maximum expectation EM algorithm, and offset field correction is performed to cope with inhomogeneities in imaging;

[0066] The review and correction module is configured to perform segmentation result review and correction: manually review and correct the CP area obtained by segmentation to ensure the accuracy and reliability of the segmentation results, including manual boundary editing, noise removal, and hole filling operations;

[0067] The volume processing module is configured to perform volume normalization processing: for CP volume calculation, FreeSurfer generates a file containing the segmentation results of all brain regions after segmentation is completed. The file contains the labels of various anatomical structures, and the volume information of the choroid plexus region is extracted using FreeSurfer's statistical tools; for CSF volume calculation, a CSF probability image is used, in which each voxel value represents the probability that the voxel belongs to the CSF, and the image is normalized to reduce individual differences; the total intracranial volume ICV is used as a reference to calculate the ratio of CP, CSF and ICV;

[0068] The first comparison module, which is configured to compare the volume differences between the healthy group and the SCA3 patient group: the differences in CP volume and CSF volume between HCs and SCA3 patients are compared using a two-sample t-test method;

[0069] A spatial normalization module configured to first calculate FA maps and other diffusion index maps using the dtifit tool in the FSL software; second, to register each subject's FA map or other diffusion index map to the FA map of the JHU atlas template using FSL's FLIRT tool, and to apply the resulting transformation matrix to the other diffusion index maps to ensure that all DTI base maps are in a unified JHU atlas space with a spatial resolution of 1 × 1 × 1 mm³;

[0070] Region of Interest module, which is configured to determine the ROI location, reference coordinates; define the ROI shape and size; create the ROI;

[0071] Extraction module, which is configured to use FSL's fslstats tool to extract the average values ​​of Dxx,proj, Dxx,assoc, Dyy,proj, and Dzz,assoc from each ROI in the front, middle, and back regions, where Dxx,proj is the water molecule diffusion coefficient in the projected fiber region along the x direction, Dxx,assoc is the water molecule diffusion coefficient in the contact fiber region along the x direction; Dyy,proj is the water molecule diffusion coefficient in the projected fiber region along the y direction; and Dzz,assoc is the water molecule diffusion coefficient in the projected fiber region along the z direction;

[0072] The DTI-ALPS index calculation module is configured to define the DTI-ALPS index as the ratio of Dxx,proj and Dxx,assoc to Dyy,proj and Dzz,assoc, and the formula is:

[0073] ,

[0074] Calculate the DTI-ALPS index of each region: calculate the anterior DTI-ALPS, middle DTI-ALPS, and posterior DTI-ALPS separately;

[0075] Segmentation and gradient calculation module, which is configured to calculate DTI-ALPS gradients from a ROI of size 3×3×3 mm³ at 1 mm intervals;

[0076] The second comparison module was configured to compare the differences in regional DTI-ALPS index between HCs and SCA3 patients using a two-sample t-test method.

[0077] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any simple modification, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention are still within the protection scope of the technical solution of the present invention.

Claims

1. A method for evaluating the dynamic abnormality of cerebrospinal fluid in patients with SCA3, characterized by: It includes the following steps: (1) Obtain high-resolution T1 images and DTI images as basic data for processing: Have the subject enter the scanning room of the MRI scanner and ensure that his head is fixed to reduce motion artifacts. Check that the acquired high-resolution T1 images and DTI image data are complete; convert the MRI image data into an image file in NIfTI format; use image processing methods to remove non-brain structures in the MRI image; (2) Automatic CP segmentation: Using Freesurfer’s cortical reconstruction method, each voxel is classified into a specific anatomical structure from T1-weighted structural MRI images using prior anatomical knowledge and probabilistic models; (3) Automatic CSF segmentation: FAST, mixed Gaussian model and offset field correction method in FSL software are used to segment brain images into different tissue types: CSF, gray matter and white matter. The maximum expectation EM algorithm is used to iteratively estimate the probability of each voxel belonging to each tissue type, and offset field correction is performed to deal with imaging inhomogeneities. (4) Segmentation result review and correction: Manual review and correction of the CP area obtained by segmentation to ensure the accuracy and reliability of the segmentation results, including manual boundary editing, noise removal, and hole filling operations; (5) Volume standardization: For CP volume calculation, FreeSurfer generates a file containing the segmentation results of all brain regions after segmentation. The file contains the labels of each anatomical structure, and the volume information of the choroid plexus region is extracted using FreeSurfer's statistical tools. For CSF volume calculation, a CSF probability image is used. Each voxel value of the image represents the probability that the voxel belongs to the CSF, and the image is standardized to reduce individual differences. The total intracranial volume (ICV) is used as a reference to calculate the ratio of CP, CSF and ICV. (6) Comparison of volume differences between the healthy group and the SCA3 patient group: The two-sample t-test method was used to compare the differences in CP volume and CSF volume between HCs and SCA3 patients; (7) Spatial standardization of fractional anisotropy FA images: First, the FA map and other diffusion index maps were calculated using the dtifit tool in the FSL software. Second, the FA map or other diffusion index maps of each subject were registered to the FA map of the JHU atlas template using the FLIRT tool in FSL. The resulting transformation matrix was applied to other diffusion index maps to ensure that all DTI base maps were located in a unified JHU atlas space with a spatial resolution of 1 × 1 × 1 mm³. (8) Define ROIs in different regions: determine ROI location and reference coordinates; define ROI shape and size; create ROI; (9) Extract diffusion values ​​of different regions: Use the fslstats tool of FSL to extract the average values ​​of Dxx,proj, Dxx,assoc, Dyy,proj, and Dzz,assoc from each ROI of the front, middle, and back regions. Dxx,proj is the diffusion coefficient of water molecules in the projected fiber region along the x direction; Dxx,assoc is the diffusion coefficient of water molecules in the contact fiber region along the x direction; Dyy,proj is the diffusion coefficient of water molecules in the projected fiber region along the y direction; Dzz,assoc is the diffusion coefficient of water molecules in the projected fiber region along the z direction; (10) Calculation of DTI-ALPS index in different regions: The DTI-ALPS index is defined as the ratio of Dxx,proj and Dxx,assoc to Dyy,proj and Dzz,assoc, and its formula is: , Calculate the DTI-ALPS index of each region: calculate the anterior DTI-ALPS, middle DTI-ALPS, and posterior DTI-ALPS separately; (11) Regional subdivision and gradient calculation: DTI-ALPS gradients were calculated from a ROI of 3 × 3 × 3 mm³ at 1 mm intervals; (12) The two-sample t-test method was used to compare the differences in regional DTI-ALPS index between HCs and SCA3 patients.

2. The method for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients according to claim 1, characterized in that: In the step (1), the following parameters are set: TR=8028ms, TE=81.8ms, flip angle=90°, slice thickness=2mm, FOV=240×240mm², matrix=120×120, voxel size=2×2×2mm³; 8 images with b=0s / mm² and 64 images with b=1000s / mm² are acquired; the DTI data are imported into the FSL software, and the non-brain tissue part in the DTI data is removed using the BET tool in FSL; the DTI data is motion corrected using the eddy_correct tool of FSL to eliminate the influence of head movement on image quality; the image distortion caused by magnetic field inhomogeneity is eliminated using eddy current correction, and the distortion specific to the EPI sequence is corrected using EPI distortion correction; the DTI data is biased field corrected using the FUGUE tool of FSL to eliminate the signal intensity change caused by magnetic field inhomogeneity.

3. The method for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients according to claim 2, characterized in that: In the step (2), the choroid plexus is segmented by identifying and extracting the CP area through a probability model and an algorithm and template based on image features, thereby segmenting the choroid plexus CP in the lateral ventricle.

4. The method for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients according to claim 3, characterized in that: In step (4), neuroimaging experts conduct independent quality assessment and final confirmation of the corrected CP segmentation results.

5. The method for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients according to claim 4, characterized in that: In step (5), the FreeSurfer statistical tool reads the segmentation result file and calculates the number and volume of voxels corresponding to each label; the fslstats tool calculates the statistical information of the image, including the number and volume of voxels.

6. The method for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients according to claim 5, characterized in that: In step (6), by testing whether there is a significant difference in the means between the two independent samples, it is evaluated whether there is a statistically significant difference in CP volume and CSF volume between the two groups of people; Two-sided t tests were performed and regression was performed on potential covariates such as gender, age, and education level.

7. The method for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients according to claim 6, characterized in that: In the step (7), the FA image is registered to the JHU atlas template, including: linear alignment and nonlinear alignment, using the FLIRT and FNIRT tools in FSL; initial linear alignment: use FLIRT to perform preliminary linear alignment to roughly align the subject's FA image to the JHU template. The principle of initial linear alignment: FLIRT uses 12 degrees of freedom affine transformation for alignment, including translation, rotation, scaling and shearing transformations. It finds the optimal transformation parameters by maximizing the similarity between the two images to align the input image with the reference image; nonlinear alignment: use FNIRT to perform nonlinear alignment to finely align the FA image to the JHU template. FNIRT uses nonlinear transformation to accurately align each voxel of the input image to the reference image. A regularized deformation field is used to describe the complex deformation between images. FNIRT achieves accurate alignment by maximizing the similarity of the registered images while minimizing the complexity of the deformation field.

8. The method for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients according to claim 7, characterized in that: In step (8), the ROI position and reference coordinates are determined as follows: Select the z = 26 horizontal plane at the level of the lateral ventricle in the MNI space, and use the labels in the JHU white matter atlas to locate the areas of projection fibers and communication fibers; Define the ROI shape and size: Use the tools in FSL to define a spherical ROI with a diameter of 3 mm, set the shape to a spherical area with a diameter of 3 mm, and a binary mask with a volume of 3×3×3 mm³; Use the atlas labels and standard spatial coordinates to create ROIs, and place the ROIs in the anterior, middle, and posterior areas of the lateral ventricle.

9. The method for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients according to claim 8, characterized in that: In the step (11), a total of 46 pairs of ROIs are generated, and the DTI-ALPS of each pair of ROIs is calculated. All calculated DTI-ALPS indices and gradient results are summarized, and the rate of change of the DTI-ALPS index is calculated to form a comprehensive assessment of the activity of the lymphatic system. In the step (12), a two-sample t-test is used to compare the inter-group differences in anterior DTI-ALPS, middle DTI-ALPS, and posterior DTI-ALPS, and potential covariates such as gender, age, and education level are regressed.

10. A device for evaluating the dynamic abnormality of cerebrospinal fluid in SCA3 patients, characterized in that: It includes: An MRI image acquisition module configured to acquire T1-weighted structural MRI images as basic data for processing: checking that the acquired T1-weighted structural MRI image data is complete; Convert MRI image data into NIfTI format image files; use image processing methods to remove non-brain structures in MRI images; A CP segmentation module, which is configured to perform automatic CP segmentation: using Freesurfer's cortical reconstruction method, from T1-weighted structural MRI images, using prior anatomical knowledge and probabilistic models to classify each voxel into a specific anatomical structure; CSF segmentation module, which is configured to perform automatic CSF segmentation: brain images are segmented into different tissue types: CSF, gray matter, and white matter using FAST, mixed Gaussian models, and offset field correction methods in FSL software, the probability of each voxel belonging to each tissue type is iteratively estimated by the maximum expectation EM algorithm, and offset field correction is performed to cope with inhomogeneities in imaging; The review and correction module is configured to perform segmentation result review and correction: manually review and correct the CP area obtained by segmentation to ensure the accuracy and reliability of the segmentation results, including manual boundary editing, noise removal, and hole filling operations; The volume processing module is configured to perform volume normalization processing: for CP volume calculation, FreeSurfer generates a file containing the segmentation results of all brain regions after segmentation is completed. The file contains the labels of various anatomical structures, and the volume information of the choroid plexus region is extracted using FreeSurfer's statistical tools; for CSF volume calculation, a CSF probability image is used, in which each voxel value represents the probability that the voxel belongs to the CSF, and the image is normalized to reduce individual differences; the total intracranial volume ICV is used as a reference to calculate the ratio of CP, CSF and ICV; The first comparison module, which was configured to compare the volume differences between the healthy group and the SCA3 patient group: the differences in CP volume and CSF volume between HCs and SCA3 patients were compared using a two-sample t-test method; A spatial normalization module configured to first calculate FA maps and other diffusion index maps using the dtifit tool in the FSL software; second, to register each subject's FA map or other diffusion index map to the FA map of the JHU atlas template using FSL's FLIRT tool, and to apply the resulting transformation matrix to the other diffusion index maps to ensure that all DTI base maps are in a unified JHU atlas space with a spatial resolution of 1 × 1 × 1 mm³; Region of Interest module, which is configured to determine the ROI location, reference coordinates; define the ROI shape and size; create the ROI; Extraction module, which is configured to use FSL's fslstats tool to extract the average values ​​of Dxx,proj, Dxx,assoc, Dyy,proj, and Dzz,assoc from each ROI in the front, middle, and back regions, where Dxx,proj is the water molecule diffusion coefficient in the projected fiber region along the x direction, Dxx,assoc is the water molecule diffusion coefficient in the contact fiber region along the x direction; Dyy,proj is the water molecule diffusion coefficient in the projected fiber region along the y direction; and Dzz,assoc is the water molecule diffusion coefficient in the projected fiber region along the z direction; The DTI-ALPS index calculation module is configured to define the DTI-ALPS index as the ratio of Dxx,proj and Dxx,assoc to Dyy,proj and Dzz,assoc, and its formula is: , Calculate the DTI-ALPS index of each region: calculate the anterior DTI-ALPS, middle DTI-ALPS, and posterior DTI-ALPS separately; Segmentation and gradient calculation module, which is configured to calculate DTI-ALPS gradients from a ROI of size 3×3×3 mm³ at 1 mm intervals; The second comparison module was configured to compare the differences in regional DTI-ALPS index between HCs and SCA3 patients using a two-sample t-test method.

Citation Information

Patent Citations

  • Method and device for evaluating brain oxygen uptake fraction and brain oxygen metabolism rate based on QSM

    CN116740066A

  • Spinal cerebellar degeneration cerebral blood flow magnetic resonance imaging processing and visualization method and device

    CN117437133A