Search

CN-120108692-B - Intelligent disease diagnosis device based on brain structure connection identifier and application thereof

CN120108692BCN 120108692 BCN120108692 BCN 120108692BCN-120108692-B

Abstract

The application provides a disease intelligent diagnosis device based on a brain structure connection identifier and application thereof, wherein the device comprises (1) a data acquisition module, (2) a data processing module, a data projection module and a Connectome Identifier extraction module, wherein the data processing module is used for reconstructing a spin distribution function of data and enabling the spin distribution function to be distributed in a standard space, (3) the data projection module is used for projecting the spin distribution function of a patient and a healthy individual into the standard space to obtain a Z value in the maximum direction in a voxel, and the Connectome Identifier extraction module is used for reducing the Z value to form a characteristic value in 1 dimension. The device of the application can be used for diagnosis of various brain diseases, lesion position and severity judgment.

Inventors

  • WANG ZHENMING
  • WEI PENGHU
  • LU JIE
  • ZHAO GUOGUANG
  • YIN ZHICHEN

Assignees

  • 首都医科大学宣武医院

Dates

Publication Date
20260505
Application Date
20250205

Claims (17)

  1. 1. A disease intelligent diagnostic device based on a brain structure connection identifier, the device comprising the following modules: (1) The data acquisition module is used for acquiring DSI data; (2) The data processing module is used for carrying out Q space differential isomorphic reconstruction on the DSI data obtained in the step (1), so that the individual space can be used for reconstructing a spin distribution function and simultaneously projecting to obtain the distribution of dispersion directions in each voxel in a standard space; (3) The data projection module is used for projecting spin distribution functions in 1 or more healthy individual standard spaces together and averaging to obtain the average value and standard deviation of the length of a dispersion vector in the direction with the largest dispersion in the voxel, projecting the spin distribution function of a patient individual into the standard space, taking the largest direction in the voxel of a single subject, subtracting the average value length mean of the population level, dividing the average value by the standard deviation length std to obtain the Z value of the largest direction in the voxel, wherein Z= (l-length mean )/ length std ; (4) Connectome Identifier extracting module, namely coding Z values in 3-dimensional matrix arrangement according to left-back-front order to form a 1-dimensional characteristic value, namely Connectome Identifier.
  2. 2. The apparatus of claim 1, the data acquisition module comprising an MRI instrument.
  3. 3. The device according to claim 1 or 2, wherein the data processing module performs the functions of processing the DSI data obtained in the step (1) by DSI-Studio software, recording DICOM format data in an original image, packaging the DICOM format data together with a b-value table into an SRC format file, reading the SRC file to a range for observing brain parenchyma in an influence space, selecting a threshold value capable of covering the brain parenchyma without including skull, and performing Q-space differential isomorphism reconstruction.
  4. 4. The apparatus of claim 1 wherein the data processing module performs functions comprising correcting magnetic field induced distortion using TOPUP before the start of the Q-space differential isomorphic reconstruction, and correcting head movement artifacts using Eddy current.
  5. 5. The device of claim 1, wherein the data processing module has a ratio of a diffusion sampling length of 1.1, a fitting method of selecting CDM, a space template of taking HCP1021, a direction distribution function interpolation parameter of 20 folds, a fiber processing number in a voxel of 10 and a thread number of 4.
  6. 6. The apparatus of claim 1, further comprising (5) Connectome Identifier an interpretation module for interpreting Connectome Identifier.
  7. 7. The apparatus of claim 1, the DSI data being DSI data of a patient.
  8. 8. The apparatus of claim 1, the DSI data being DSI data of a patient and DSI data of 1 or more healthy individuals.
  9. 9. The apparatus of claim 7, the DSI data is brain DSI data.
  10. 10. The device of claim 1, applied in the diagnosis of brain diseases.
  11. 11. The device of claim 1, applied to determine brain disease lesion location.
  12. 12. The device of claim 1, which is applied to distinguish brain disease severity.
  13. 13. The device of claim 10, wherein the brain disease is epilepsy, parkinson's disease, alzheimer's disease.
  14. 14. A storage medium having recorded thereon a program that performs the steps of: (1) Acquiring DSI data; (2) Performing data processing, namely performing Q space differential isomorphic reconstruction on DSI data, so that the individual space reconstructs a spin distribution function and simultaneously projects to obtain the distribution of dispersion directions in each voxel in a standard space; (3) The data projection comprises the steps of projecting spin distribution functions in a standard space of 1 or more healthy individuals together and averaging to obtain the average value and standard deviation of the length of a dispersion vector in the direction with the largest dispersion in a voxel, projecting the spin distribution functions of the individual patient into the standard space, taking the largest direction in the voxel of a single subject, subtracting the average value length mean of the population level, dividing the average value by the standard deviation length std to obtain the Z value of the largest direction in the voxel, wherein Z= (l-length mean )/ length std ; (4) Connectome Identifier extracting, namely coding Z values which are arranged in a 3D matrix according to the left-back-front sequence to form a 1-dimensional characteristic value, namely Connectome Identifier.
  15. 15. The method of claim 14 wherein the data processing step includes processing the acquired data with DSI-Studio software, recording DICOM-formatted data in the original image, packaging the data with b-value table into SRC-formatted file, reading the SRC file to a range affecting brain parenchyma in space, selecting a threshold that can cover brain parenchyma without incorporating skull, and performing Q-space differential isomorphism reconstruction.
  16. 16. The storage medium of claim 14 or 15, wherein the data processing comprises correcting magnetic field induced distortion using TOPUP before the start of the Q-space differential isomorphic reconstruction, and correcting head movement artifacts using Eddy current.
  17. 17. The storage medium of claim 14, wherein the ratio of the diffusion sampling length in the data processing process is 1.1, the fitting method selects CDM, the space template takes HCP1021, the direction distribution function interpolation parameter selects 20 folds, the number of fibers processed in the voxels selects 10, and the number of threads selects 4.

Description

Intelligent disease diagnosis device based on brain structure connection identifier and application thereof Technical Field The application belongs to the field of medical imaging data processing, and particularly provides a disease intelligent diagnosis device based on a brain structure connection identifier and application thereof. Background Diffusion tensor imaging (diffusion tensor imaging, DTI) techniques have been demonstrated in clinical applications by a number of studies, such as by autopsy experiments, to demonstrate the effectiveness of the medium-to-large white matter fiber tract tracking results obtained using DTI techniques. In subsequent studies, DTI was used for preoperative evaluation of neurosurgery, and the distribution and morphology of the main white matter structures can be delineated preoperatively, and common white matter fiber bundles reconstructed by DTI include cone bundles, visual conduction pathways, and linguistic conduction pathways. DTI can be further realized by presenting these structures for the purpose of preserving functionality for the patient, which has an important impact on the formulation of complex lesion surgical strategies. However, DTI technology has its limitations in that it cannot collect and analyze crossing fibers, cannot accurately map the start and stop points of white matter fibers, and fibers with false positives and false negatives often appear in the tracking process, which is determined by the technical principle of DTI. Technically, DTI can only collect the main limited dispersion direction in a voxel, namely the dispersion direction of water molecules after being wrapped by the coating, and is a relative value of the limited dispersion direction, not an absolute measurement value. Specifically, the tissue characteristics are measured by eigenvalues (λ1, λ2, and λ3). The axial dispersion capacity is represented by lambda 1, and reflects the water molecule dispersion capacity along the long axis of the fiber, and the radial eigenvalue is divided into two perpendicular eigenvalues lambda 2 and lambda 3. The partial anisotropy (fractional anisotropy, FA) is a diffusion scalar parameter commonly used in DTI technology, and reflects the degree of partial anisotropy, which is the ratio of the ability of water molecules to diffuse along a certain direction in the total diffusion ability, and the formula is: The value is between 0 and 1. In general, fibers with better myelination have higher FA values. The FA value decreases when the degree of myelination decreases or neurons are absent. In a tissue having dense fibers, the diffusivity in the direction perpendicular to the fibers is suppressed, the FA value is increased, and when the fiber structure is damaged by degenerative diseases or other diseases, the diffusivity in the direction perpendicular to the fibers is increased, and the FA value is decreased. Thus assessing brain FA values can assess structural changes in the tissue. Thus, DTI does not collect such information from the beginning of data acquisition for complex fiber structures such as intersections and angulations in voxels, and thus a large amount of variation in nerve tissue structures is mixed, and a large sample amount of data is required to reveal the network function state of the brain in response to histological features. The deterministic tracking developed on this basis does not reproduce the white matter fibrous anatomy found in the fibrous bundle anatomy and histological techniques. The same principle is that, for white matter fibers embedded in edematous tissue, DTI only considers the direction of voxel average, and thus cannot be sufficiently tracked. Thus, there is an urgent need for acquisition techniques that start with acquisition, i.e. fully take into account unrestricted diffusion directions within voxels. New fiber bundle profiling techniques such as high-angular-resolution diffusion imaging (HARDI) and DSI techniques, which have been increasingly developed in recent years, can be used to solve these problems. The diffusion spectrum imaging technique (diffusion spectrum imaging, DSI) is a diffusion-based white matter imaging technique, and compared with the traditional diffusion tensor imaging technique (diffusion tensor imaging, DTI), DSI has more imaging directions and higher b values, and can describe three-dimensional distribution of microscopic displacements of spins in voxels in units of voxels based on a probability density function (probability density function, PDF) framework, so that analysis of white matter fiber bundles of the brain is more accurate. White matter research is an important component in neuroscience research, and DTI technology makes in vivo white matter mapping possible for human beings, and can be widely used in the diagnosis and treatment process of brain diseases. The DTI technique employs a generalized Q-sampling imaging (GQI) algorithm, the core model of which is a GQI model, and the corres