Method and device for evaluating flow of cerebrospinal fluid along space around blood vessel

Abstract

The invention discloses a method and a device for evaluating flow of cerebrospinal fluid along a surrounding space of a blood vessel. The method comprises the following steps: acquiring an fMRI image of a subject, acquiring a T1 weighted anatomical image of the subject, acquiring global BOLD signals of a brain and a cerebellum, performing CSF segmentation, capturing an inflow effect of CSF, calculating a cross-correlation function between the global BOLD signals and CSF signals, analyzing a negative derivative of the BOLD signals and the cross-correlation function of the CSF signals, and performing statistical analysis. Therefore, the change of the glue lymph function of a spinal cerebellar ataxia type 3 patient can be revealed, and the effect between a BOLD signal and CSF kinetics in metabolite removal is provided.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and particularly to a method for evaluating the flow of cerebrospinal fluid along the perivascular space, and a device for evaluating the flow of cerebrospinal fluid along the perivascular space. 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. Patients with SCA3 usually present with motor and non-motor symptoms, such as pyramidal signs, extrapyramidal symptoms, cognitive decline, and rapid eye movement behavior disorder. The pathogenesis of this disease is related to protein misfolding, specifically caused by the expansion of CAG repeat sequences in the coding region of the ATXN3 gene on chromosome 14, which produces abnormal polyglutamine proteins. This disorder disrupts the balance between protein production and clearance, thereby impairing cell function and triggering neuronal death. Maintaining the homeostasis of the brain environment is crucial for clearing these abnormal proteins, as its imbalance can affect the cellular environment, trigger neuroinflammation, and drive the complex pathological evolution and progression of the disease. In addition, the key pathological processes of SCA3 also include enhanced inflammatory response, accumulation of metabolic waste due to mitochondrial dysfunction, and imbalance in the regulation of cellular stress response. Considering these pathological processes, the waste clearance mechanism of the brain, especially through the glymphatic system, is very critical for maintaining nervous system health. However, current research on the glymphatic system in SCA3 is still limited. Therefore, it is particularly important to deeply understand the pathophysiology of SCA3. This study aims to explore the functional status of the glymphatic system in patients with SCA3 and investigate its impact on the development and progression of the disease. Summary of the Invention

[0003] 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 flow of cerebrospinal fluid along the perivascular space, which can reveal the changes in glymphatic function in patients with spinocerebellar ataxia type 3 and provide the role of BOLD signal and CSF dynamics in metabolite clearance.

[0004] The technical solution of the present invention is as follows: This method for evaluating the flow of cerebrospinal fluid along the perivascular space includes the following steps:

[0005] (1) Obtain the fMRI images of the subject: Preprocess the fMRI images;

[0006] (2) Obtain the T1-weighted anatomical images of the subject: Register the rs-fMRI images with high-resolution T1-weighted structural MRI images and register them into the MNI-152 space;

[0007] (3) Obtain the global BOLD signals of the brain and cerebellum: The global BOLD signals are obtained from the cortical gray matter regions of the brain and cerebellum, which are clearly demarcated in the preprocessed functional images in individual space by the Automated Anatomical Labeling 2 atlas;

[0008] (4) Perform CSF segmentation: Place the CSF mask in the bottom slices of the rs-fMRI images, which contain the cerebellum of all subjects, and confirm by visual inspection to capture the inflow effect of CSF;

[0009] (5) Capture the inflow effect of CSF: When capturing the inflow effect of CSF, define the region of interest related to cerebrospinal fluid. To extract the signals relevant for analysis, the fMRI data is band-pass filtered to eliminate high-frequency and low-frequency noises unrelated to the CSF inflow effect;

[0010] (6) Calculate the cross-correlation function between the global BOLD signal and the CSF signal: The extracted BOLD and CSF signals are detrended and further filtered to eliminate artifacts related to long-term changes and residual noise. The BOLD and CSF signals are temporally smoothed, and the cross-correlation function between the global BOLD signal and the CSF signal is calculated using the spm signal processing tool;

[0011] (7) Analyze the cross-correlation function of the negative derivative of the BOLD signal and the CSF signal: Based on the calculated cross-correlation function, analyze the correlation pattern between the BOLD signal and the CSF signal, including examining the peak position, correlation strength, and time delay, as well as their relationship with specific brain functions or activity patterns;

[0012] (8) Conduct statistical analysis, including two-sample t-tests, to compare the gBOLD-CSF coupling metrics between the healthy control group and the SCA3 group, and perform two-sided t-tests.

[0013] The present invention obtains the fMRI images of the subjects, obtains the T1-weighted anatomical images of the subjects, obtains the global BOLD signals of the brain and cerebellum, performs CSF segmentation, captures the inflow effect of CSF, calculates the cross-correlation function between the global BOLD signal and the CSF signal, analyzes the cross-correlation function of the negative derivative of the BOLD signal and the CSF signal, and conducts statistical analysis, thereby being able to reveal the changes in glymphatic function in patients with spinocerebellar ataxia type 3 and providing the role of the BOLD signal and CSF dynamics in metabolite clearance.

[0014] There is also provided a device for evaluating the flow of cerebrospinal fluid along the perivascular space, which includes:

[0015] an fMRI image acquisition module configured to acquire the fMRI images of the subjects: preprocess the fMRI images;

[0016] A T1 image acquisition module configured to acquire T1-weighted anatomical images of a subject:

[0017] register the rs-fMRI images with high-resolution T1-weighted structural MRI images and register them to

[0018] the MNI-152 space;

[0019] A BOLD signal acquisition module configured to acquire global BOLD signals of the brain and cerebellum:

[0020] acquire global BOLD signals from the cortical gray matter regions of the brain and cerebellum, which are clearly demarcated in the preprocessed functional images in individual space by the Automated Anatomical Labeling 2 atlas;

[0021] A CSF segmentation module configured to perform CSF segmentation: place a CSF mask in the bottom slices of the rs-fMRI images, which contain the cerebellum of all subjects, and confirm by visual inspection to capture the inflow effect of CSF;

[0022] An inflow effect capture module configured to capture the inflow effect of CSF: when capturing the inflow effect of CSF, define the region of interest related to cerebrospinal fluid, and perform band-pass filtering on the fMRI data to eliminate high-frequency and low-frequency noises unrelated to the inflow effect of CSF in order to extract signals relevant for analysis;

[0023] A calculation module configured to calculate the cross-correlation function between the global BOLD signal and the CSF signal: perform detrending and further filtering on the extracted BOLD and CSF signals to eliminate artifacts related to long-term changes and residual noise, perform temporal smoothing on the BOLD and CSF signals, and use the spm signal processing tool to calculate the cross-correlation function between the global BOLD signal and the CSF signal;

[0024] An analysis module configured to analyze the cross-correlation function between the negative derivative of the BOLD signal and the CSF signal: according to the calculated cross-correlation function, analyze the correlation pattern between the BOLD signal and the CSF signal, including examining the peak position, correlation strength, and time delay, as well as their relationship with specific brain functions or activity patterns;

[0025] A statistical analysis module configured to perform statistical analysis, including a two-sample t-test to compare the gBOLD-CSF coupling indices of a healthy control group and an SCA3 group, and perform a two-sided t-test. Description of the Drawings

[0026] Figure 1 Shows a flowchart of a method for evaluating the flow of cerebrospinal fluid along the perivascular space according to the present invention. Detailed implementation mode

[0027] As Figure 1 shown, this method for evaluating the flow of cerebrospinal fluid along the perivascular space includes the following steps:

[0028] (1) Obtain the fMRI images of the subject: Preprocess the fMRI images;

[0029] (2) Obtain the T1-weighted anatomical images of the subject: Register the rs-fMRI images with high-resolution T1-weighted structural MRI images and register them into the MNI-152 space;

[0030] (3) Obtain the global BOLD signals of the brain and cerebellum: Obtain the global BOLD signals from the cortical gray matter regions of the brain and cerebellum, and these regions are clearly demarcated in the preprocessed functional images in individual space by the Automated Anatomical Labeling 2 atlas;

[0031] (4) Perform CSF segmentation: Place the CSF mask in the bottom slices of the rs-fMRI images, and these slices contain the cerebellum of all subjects and are confirmed by visual inspection to capture the inflow effect of CSF;

[0032] (5) Capture the inflow effect of CSF: When capturing the inflow effect of CSF, define the region of interest related to cerebrospinal fluid. In order to extract the signals related to the analysis, perform band-pass filtering on the fMRI data to eliminate the high-frequency and low-frequency noises unrelated to the inflow effect of CSF;

[0033] (6) Calculate the cross-correlation function between the global BOLD signal and the CSF signal: Perform detrending and further filtering on the extracted BOLD and CSF signals to eliminate the artifacts related to long-term changes and residual noises, perform temporal smoothing on the BOLD and CSF signals, and use the spm signal processing tool to calculate the cross-correlation function between the global BOLD signal and the CSF signal;

[0034] (7) Analyze the cross-correlation function between the negative derivative of the BOLD signal and the CSF signal: According to the calculated cross-correlation function, analyze the correlation pattern between the BOLD signal and the CSF signal, including checking the peak position, correlation intensity, and time delay, as well as their relationship with specific brain functions or activity patterns;

[0035] (8) Perform statistical analysis, including two-sample t-tests, to compare the gBOLD-CSF coupling indexes of the healthy control group and the SCA3 group, and perform two-sided t-tests.

[0036] The present invention acquires the fMRI images of a subject, acquires the T1-weighted anatomical images of the subject, acquires the global BOLD signals of the brain and cerebellum, performs CSF segmentation, captures the inflow effect of CSF, calculates the cross-correlation function between the global BOLD signal and the CSF signal, analyzes the cross-correlation function between the negative derivative of the BOLD signal and the CSF signal, and performs statistical analysis, so as to be able to reveal the changes in glymphatic function in patients with spinocerebellar ataxia type 3 and provide the role of the BOLD signal and CSF dynamics in metabolite clearance.

[0037] Preferably, the method further includes step (9) clinical correlation analysis: extracting the severity score SARA score of the SCA clinical motor disorder symptoms of the subject, calculating the correlation between the DTI-ALPS indices in the anterior, middle, and posterior regions and the SARA score, and using a multiple linear regression model to control the influence of covariates of age, gender, and educational level.

[0038] Preferably, in the step (1), the preprocessing includes slice time correction, motion correction, skull stripping, spatial smoothing using a 4-mm FWHM kernel, 0.01-0.1 Hz band-pass filtering, and removal of linear and quadratic time trends. Regression of the global signal and CSF signal, as well as regression of motion parameters, are omitted in the preprocessing.

[0039] Preferably, in the step (2), high-resolution T1-weighted structural MRI images of the subject are acquired, and the images are used to depict the detailed anatomical structure of the brain. These T1-weighted structural MRI images are registered with the rs-fMRI images of the subject to ensure their spatial consistency. The fMRI images are rigidly transformed with the anatomical images to eliminate spatial misalignment caused by the head movement of the subject or different scanning parameters. To be compatible with the standard neuroimaging analysis framework, these registered images are further registered to the MNI-152 space.

[0040] Preferably, in the step (3), to ensure the perfect alignment of the cortical gray matter mask with the functional space, the cortical gray matter mask is transformed from the MNI-152 standard space to the individual functional space of each subject. The CSF signal and the gray matter BOLD signal are extracted from the defined ROI. The time derivative of the BOLD signal is extracted and processed with a negative sign and threshold setting as needed.

[0041] Preferably, in the step (4), the CSF mask is a binary image used to identify the regions in the image that belong to the cerebrospinal fluid.

[0042] Preferably, in step (5), the ROI is defined based on the distribution and flow characteristics of cerebrospinal fluid in the brain to ensure accurate extraction of signals relevant for analysis. In band-pass filtering, an appropriate frequency range of 0.2 - 4 Hz is selected to filter the fMRI data.

[0043] Preferably, in step (6), after obtaining the cross-correlation function between the negative derivative dx / dtBOLD of the BOLD signal and the cerebrospinal fluid (CSF) signal, the correlation pattern between the two is further explored in depth by observing the waveform of the cross-correlation function to determine its peak position, which reflects the time point with the strongest correlation between the negative derivative of the BOLD signal and the CSF signal.

[0044] Preferably, in step (7), the correlation strength is calculated and evaluated. The correlation strength is the amplitude of the cross-correlation function near the peak position. The magnitude of the correlation strength reflects the strength of the association between the negative derivative of the BOLD signal and the CSF signal, thereby revealing the tightness of their dependence relationship; the time delay is the lag or lead time of the change in the negative derivative of the BOLD signal relative to the change in the CSF signal.

[0045] Preferably, in step (8), the two-sample t-test is used to check whether there are significant differences in the means between two independent samples to find out whether there are statistically significant differences in the gBOLD-CSF coupling index between two groups of people; when performing a two-sided t-test, covariates of gender, age, and education level are regressed.

[0046] Those of ordinary skill in the art can understand that all or part of the steps in implementing the above method embodiments can be completed by instructing relevant hardware through a program. The program can be stored in a computer-readable storage medium. When the program is executed, it includes the steps of the above method embodiments, and the storage medium can be: ROM / RAM, magnetic disk, optical disc, memory card, etc. Therefore, corresponding to the method of the present invention, the present invention also includes an apparatus for evaluating the flow of cerebrospinal fluid along the perivascular space, which is usually represented in the form of functional modules corresponding to the steps of the method. The apparatus includes:

[0047] An fMRI image acquisition module configured to acquire fMRI images of a subject and preprocess the fMRI images;

[0048] A T1 image acquisition module configured to acquire T1-weighted anatomical images of the subject and register the

[0049] rs-fMRI images with high-resolution T1-weighted structural MRI images and register them to the

[0050] MNI-152 space;

[0051] A BOLD signal acquisition module configured to acquire global BOLD signals of the brain and cerebellum:

[0052] Acquire global BOLD signals from the cortical gray matter regions of the brain and cerebellum, which are clearly demarcated in the preprocessed functional images in individual space by the Automated Anatomical Labeling 2 atlas;

[0053] A CSF segmentation module configured to perform CSF segmentation: Place a CSF mask in the bottom slices of the rs-fMRI images, which contain the cerebellum of all subjects and are confirmed by visual inspection to capture the inflow effect of CSF;

[0054] An inflow effect capture module configured to capture the inflow effect of CSF: When capturing the inflow effect of CSF, define the region of interest related to cerebrospinal fluid, and perform band-pass filtering on the fMRI data to eliminate high-frequency and low-frequency noises unrelated to the inflow effect of CSF in order to extract signals relevant for analysis;

[0055] A calculation module configured to calculate the cross-correlation function between the global BOLD signal and the CSF signal: Detrend and further filter the extracted BOLD and CSF signals to eliminate artifacts related to long-term changes and residual noise, perform temporal smoothing on the BOLD and CSF signals, and use the spm signal processing tool to calculate the cross-correlation function between the global BOLD signal and the CSF signal;

[0056] An analysis module configured to analyze the cross-correlation function between the negative derivative of the BOLD signal and the CSF signal: According to the calculated cross-correlation function, analyze the correlation pattern between the BOLD signal and the CSF signal, including examining the peak position, correlation strength, and time delay, as well as their relationship with specific brain functions or activity patterns;

[0057] A statistical analysis module configured to perform statistical analysis, including a two-sample t-test to compare the gBOLD-CSF coupling metrics between the healthy control group and the SCA3 group, and perform a two-sided t-test.

[0058] The above are only the preferred embodiments of the present invention and do not impose any form of limitation on the present invention. Any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A method for evaluating the flow of cerebrospinal fluid along perivascular spaces, characterized in that: It includes the following steps: (1) Obtaining fMRI images of the subject: preprocessing the fMRI images; (2) Obtain T1-weighted anatomical images of the subjects: rs-fMRI images are registered with high-resolution T1-weighted structural MRI images and registered to the MNI-152 space; (3) Obtaining global BOLD signals of the cerebrum and cerebellum: Obtaining global BOLD signals from cortical gray matter areas of the cerebrum and cerebellum, which were clearly delineated in preprocessed functional images in individual space using the automated anatomical labeling 2 atlas; (4) Perform CSF segmentation: Place a CSF mask in the bottom slice of the rs-fMRI images that contain the cerebellum of all subjects and confirm by visual inspection to capture the CSF inflow effect; (5) Capturing the CSF inflow effect: When capturing the CSF inflow effect, the region of interest related to the cerebrospinal fluid is defined. In order to extract the signals relevant to the analysis, the fMRI data are band-pass filtered to eliminate high-frequency and low-frequency noise that is not related to the CSF inflow effect; (6) Calculate the cross-correlation function between the global BOLD signal and the CSF signal: detrend and further filter the extracted BOLD and CSF signals to eliminate artifacts related to long-term changes and residual noise, perform temporal smoothing on the BOLD and CSF signals, and use the SPM signal processing tool to calculate the cross-correlation function between the global BOLD signal and the CSF signal; (7) Analyze the cross-correlation function between the negative derivative of the BOLD signal and the CSF signal: Based on the calculated cross-correlation function, analyze the correlation pattern between the BOLD signal and the CSF signal, including examining the peak position, correlation strength and time delay, and their relationship with specific brain functions or activity patterns; (8) Statistical analysis was performed, including two-sample t -tests, and two-sided t -tests were performed to compare the gBOLD-CSF coupling indices between the healthy control group and the SCA3 group.

2. The method for evaluating the flow of cerebrospinal fluid along the perivascular space according to claim 1, characterized in that: The method also includes step (9) clinical correlation analysis: extracting the SARA score, which is the severity score of the SCA clinical movement disorder symptoms of the subject, calculating the correlation between the DTI-ALPS index of the front, middle and back regions and the SARA score, and using a multivariate linear regression model to control the influence of covariates such as age, gender and education level.

3. The method for evaluating the flow of cerebrospinal fluid along the perivascular space according to claim 2, characterized in that: In the step (2), a high-resolution T1-weighted structural MRI image is obtained from the subject, and the image is used to depict the detailed anatomical structure of the brain. These T1-weighted structural MRI images are registered with the subject's rs-fMRI image to ensure that they are spatially consistent. The fMRI image and the anatomical image are rigidly transformed to eliminate spatial misalignment caused by the subject's head movement or different scanning parameters. In order to be compatible with the standard neuroimaging analysis framework, these registered images are further registered to the MNI-152 space.

4. The method for evaluating the flow of cerebrospinal fluid along the perivascular space according to claim 3, characterized in that: In step (3), in order to ensure perfect alignment between the cortical gray matter mask and the functional space, the cortical gray matter mask is converted from the MNI-152 standard space to the individual functional space of each subject, and the CSF signal and gray matter BOLD signal are extracted from the defined ROI. The time derivative of the BOLD signal is extracted, and the negative sign is processed and the threshold is set as needed.

5. The method for evaluating the flow of cerebrospinal fluid along the perivascular space according to claim 4, characterized in that: In step (4), the CSF mask is a binary image used to identify the area in the image that belongs to the cerebrospinal fluid.

6. The method for evaluating the flow of cerebrospinal fluid along the perivascular space according to claim 5, characterized in that: In step (5), the ROI is defined based on the distribution and flow characteristics of cerebrospinal fluid in the brain to ensure that the signals related to the analysis can be accurately extracted, and an appropriate frequency range of 0.2-4 Hz is selected in the bandpass filter to filter the fMRI data.

7. The method for evaluating the flow of cerebrospinal fluid along the perivascular space according to claim 6, characterized in that: In the step (6), after obtaining the cross-correlation function between the negative derivative dx / dtBOLD of the BOLD signal and the cerebrospinal fluid CSF signal, the correlation pattern between the two is further explored, and the waveform of the cross-correlation function is observed to determine its peak position, which reflects the time point when the correlation between the negative derivative of the BOLD signal and the CSF signal is the strongest.

8. The method for evaluating the flow of cerebrospinal fluid along the perivascular space according to claim 7, characterized in that: In the step (7), the correlation strength is calculated and evaluated. The correlation strength is the amplitude of the cross-correlation function near the peak position. The magnitude of the correlation strength reflects the strength of the correlation between the negative derivative of the BOLD signal and the CSF signal, thereby revealing the closeness of their dependence. The time delay is the lag or lead time of the change of the negative derivative of the BOLD signal relative to the change of the CSF signal.

9. The method for evaluating the flow of cerebrospinal fluid along the perivascular space according to claim 8, characterized in that: In step (8), the two-sample t-test is performed to test whether there is a significant difference in the means between the two independent samples, so as to understand whether there is a statistically significant difference between the two groups of people in the gBOLD-CSF coupling index; in the bilateral t-test, the covariates of gender, age and education level are regressed.

10. A device for evaluating the flow of cerebrospinal fluid along perivascular spaces, characterized in that: It includes: an fMRI image acquisition module configured to acquire an fMRI image of a subject: preprocess the fMRI image; a T1 image acquisition module configured to acquire T1-weighted anatomical images of the subject: registering the rs-fMRI images with the high-resolution T1-weighted structural MRI images and registering them to the MNI-152 space; a BOLD signal acquisition module configured to acquire global BOLD signals of the cerebrum and cerebellum: global BOLD signals were acquired from cortical gray matter regions of the cerebrum and cerebellum that were clearly delineated in preprocessed functional images in individual space using the automated anatomical labeling 2 atlas; a CSF segmentation module configured to perform CSF segmentation: CSF masks were placed in the bottom slices of the rs-fMRI images that contained the cerebellum of all subjects and confirmed by visual inspection to capture CSF inflow effects; An inflow effect capture module, which is configured to capture the inflow effect of CSF: when capturing the CSF inflow effect, a region of interest related to cerebrospinal fluid is defined, and in order to extract the signal relevant to the analysis, the fMRI data is band-pass filtered to eliminate high-frequency and low-frequency noises that are not related to the CSF inflow effect; a computation module configured to compute a cross-correlation function between a global BOLD signal and a CSF signal: detrending and further filtering the extracted BOLD and CSF signals to remove artifacts associated with long-term changes and residual noise, performing temporal smoothing on the BOLD and CSF signals, and computing a cross-correlation function between the global BOLD signal and the CSF signal using an SPM signal processing tool; an analysis module configured to analyze a cross-correlation function between a negative derivative of a BOLD signal and a CSF signal: analyzing a correlation pattern between the BOLD signal and the CSF signal based on the calculated cross-correlation function, including examining peak positions, correlation strengths and time delays, and their relationships with specific brain functions or activity patterns; The statistical analysis module was configured to perform statistical analyses, including a two-sample t-test to compare the gBOLD-CSF coupling index between the healthy control group and the SCA3 group, and a two-sided t-test.

Images