CN113283465B - Diffusion tensor imaging data analysis method and device - Google Patents
Diffusion tensor imaging data analysis method and device Download PDFInfo
- Publication number
- CN113283465B CN113283465B CN202110359402.0A CN202110359402A CN113283465B CN 113283465 B CN113283465 B CN 113283465B CN 202110359402 A CN202110359402 A CN 202110359402A CN 113283465 B CN113283465 B CN 113283465B
- Authority
- CN
- China
- Prior art keywords
- network connection
- brain
- groups
- data
- whole brain
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000007405 data analysis Methods 0.000 title claims abstract description 28
- 238000002598 diffusion tensor imaging Methods 0.000 title claims description 8
- 210000004556 brain Anatomy 0.000 claims abstract description 154
- 239000011159 matrix material Substances 0.000 claims abstract description 99
- 239000013598 vector Substances 0.000 claims abstract description 57
- 210000004126 nerve fiber Anatomy 0.000 claims abstract description 41
- 239000000835 fiber Substances 0.000 claims abstract description 38
- 210000004885 white matter Anatomy 0.000 claims abstract description 29
- 238000007619 statistical method Methods 0.000 claims abstract description 25
- 238000007781 pre-processing Methods 0.000 claims abstract description 20
- 238000013145 classification model Methods 0.000 claims abstract description 13
- 238000003384 imaging method Methods 0.000 claims abstract description 10
- 238000012545 processing Methods 0.000 claims abstract description 7
- 238000000605 extraction Methods 0.000 claims description 14
- 238000005259 measurement Methods 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 10
- 238000010276 construction Methods 0.000 claims description 10
- 238000012937 correction Methods 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000009792 diffusion process Methods 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000009499 grossing Methods 0.000 claims description 3
- 238000013179 statistical model Methods 0.000 claims description 3
- 238000012353 t test Methods 0.000 claims description 3
- 238000005192 partition Methods 0.000 abstract description 20
- 238000004458 analytical method Methods 0.000 abstract description 6
- 238000004422 calculation algorithm Methods 0.000 description 7
- 230000035945 sensitivity Effects 0.000 description 6
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000011160 research Methods 0.000 description 5
- 238000012549 training Methods 0.000 description 5
- 238000005481 NMR spectroscopy Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000000692 Student's t-test Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 206010003805 Autism Diseases 0.000 description 2
- 208000020706 Autistic disease Diseases 0.000 description 2
- 206010039897 Sedation Diseases 0.000 description 2
- 210000000988 bone and bone Anatomy 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 230000009191 jumping Effects 0.000 description 2
- 230000036280 sedation Effects 0.000 description 2
- 208000014644 Brain disease Diseases 0.000 description 1
- 241000288113 Gallirallus australis Species 0.000 description 1
- 238000012307 MRI technique Methods 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000002552 dosage form Substances 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 238000001647 drug administration Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002599 functional magnetic resonance imaging Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000003340 mental effect Effects 0.000 description 1
- 230000006996 mental state Effects 0.000 description 1
- 230000000926 neurological effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000000932 sedative agent Substances 0.000 description 1
- 210000003625 skull Anatomy 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 230000002618 waking effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/0033—Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room
- A61B5/004—Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room adapted for image acquisition of a particular organ or body part
- A61B5/0042—Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room adapted for image acquisition of a particular organ or body part for the brain
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves
- A61B5/055—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
- G06T7/344—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving models
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2576/00—Medical imaging apparatus involving image processing or analysis
- A61B2576/02—Medical imaging apparatus involving image processing or analysis specially adapted for a particular organ or body part
- A61B2576/026—Medical imaging apparatus involving image processing or analysis specially adapted for a particular organ or body part for the brain
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Public Health (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Molecular Biology (AREA)
- Veterinary Medicine (AREA)
- General Physics & Mathematics (AREA)
- Radiology & Medical Imaging (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Heart & Thoracic Surgery (AREA)
- Biomedical Technology (AREA)
- Pathology (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Physiology (AREA)
- Signal Processing (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Psychiatry (AREA)
- General Engineering & Computer Science (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- High Energy & Nuclear Physics (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Neurology (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention discloses a data analysis method and device, and belongs to the technical field of data processing. The method of the invention comprises the following steps: when N whole-brain image data sets are received, preprocessing operation is carried out on the N whole-brain image data sets to obtain N whole-brain basic data sets; reconstructing a whole brain white matter nerve fiber bundle of each whole brain basic data by using a fiber tracking imaging strategy, and calculating indexes such as anisotropic fraction and the like in each whole brain basic data to obtain N whole brain white matter nerve fiber bundle data groups; registering with a plurality of pre-divided brain partitions, calculating a structural network connection matrix to obtain N structural network connection matrix groups, performing network statistical analysis on the N structural network connection matrix groups, and extracting features to obtain N feature value vector groups; and finally, classifying according to the N characteristic value vector groups, and establishing a classification model. The invention also discloses a device corresponding to the method. The invention can carry out global analysis on the brain data, and further can obtain the structural connection and the network among the brain partitions.
Description
Technical Field
The embodiment of the invention relates to the technical field of data processing, in particular to a data analysis method and device.
Background
With the rapid development of life science, the life science community pays more and more attention to the research on the most complex and precise organs of human body, namely brain. A conventional research method generally includes acquiring image data of a human brain by a magnetic resonance imaging technique and a functional magnetic resonance imaging technique, determining a brain partition to be researched based on the acquired image data, and calculating a relationship between the brain partition to be researched and other partitions, thereby obtaining research data of the brain partition.
Based on the traditional research mode, the existing brain data analysis method calculates the brain data by taking the determined brain partition to be researched as a core, so that the obtained research data is more focused on reflecting the characteristics of the brain partition to be researched, the obtained data result is relatively smooth, and further, the structure and the state of the whole brain researched by technicians are limited.
In view of the above, there is a need for a method for analyzing brain data comprehensively.
Disclosure of Invention
The embodiment of the invention provides a method and a device for analyzing diffusion tensor imaging data, which are used for solving the problem that the diffusion tensor imaging data obtained in the prior art are relatively unilateral.
In one aspect, an embodiment of the present invention provides a data analysis method, where the method includes:
a pretreatment step:
when N whole brain image data sets are received, performing preprocessing operation on each whole brain image data in the N whole brain image data sets to obtain N whole brain basic data sets; wherein, N is a positive integer greater than or equal to 2, and each whole brain image data group comprises a plurality of whole brain image data;
a structure network connection matrix construction step:
reconstructing a whole brain white matter nerve fiber bundle of each whole brain basic data in the N whole brain basic data groups, and calculating a designated index parameter of each whole brain basic data to obtain N whole brain white matter nerve fiber bundle data groups;
registering each full-white-matter nerve fiber bundle data in the N full-white-matter nerve fiber bundle data groups with a plurality of pre-divided brain areas, taking the mean value of designated index parameters between the brain areas as the structural network connection measurement between the brain areas, and calculating a structural network connection matrix to obtain N structural network connection matrix groups;
network statistical analysis:
performing network-based statistical analysis on the N structural network connection matrix groups, determining the network structure difference among the groups, and taking the structural network connection (namely network edge) corresponding to the network structure difference meeting the specified conditions as a characteristic object;
a characteristic extraction step:
based on the feature object, performing feature extraction on each structure network connection matrix in the N structure network connection matrix groups to generate feature value vectors to obtain N feature value vector groups;
a model establishing step:
and classifying according to the N characteristic connection value vector groups, and establishing a classification model.
In one possible implementation, the preprocessing step performs a preprocessing operation on each of the N whole brain image data sets (to acquire diffusion tensor imaging data), including:
performing format conversion on the whole brain image data, extracting image data without gradient weighting, and performing processing of deleting non-brain parts on the obtained image data, such as removing head bones;
and performing eddy current correction on the image data from which the non-brain part is deleted, and performing tensor fitting operation on the image data after the eddy current correction to obtain diffusion tensor imaging data which are used as whole brain basic data.
In one possible implementation manner, in the step of constructing the structural network connection matrix, the index parameter is specified as an anisotropy Fraction (FA).
Further, in the step of constructing the structural network connection matrix, when reconstructing the white matter nerve fiber tract of the whole brain of each of the N whole brain basic data sets according to the anisotropic Fraction (FA), a fiber tracking imaging mode may be adopted, which specifically includes:
(1) determining starting point x for fiber bundle tracking0＝(x0,y0,z0) Thereby obtaining a starting point x0A bit vector of (a), wherein x0Points (voxels) for which the FA value is greater than the set FA threshold;
(2) determining the current tracking direction: the current starting point x0Is given as (v) is given as the principal eigenvector of (v ═ vx,vy,vz) As the current tracking direction;
(3) determining an intercept point x ═ x, y, z of a next voxel of the fiber tracking track, that is, a next track point (current tracking position) of the fiber tracking track, where the coordinate of x is specifically:
s represents a specified arc length;
(4) determining whether a specified first tracking stop condition is met, if so, executing the step (7); otherwise, executing the step (5);
(5) obtaining a new principal eigenvector v 'based on the diffusion tensor of the interception point x, and executing the step (6) after determining the direction of the new principal eigenvector v';
wherein, the direction of the new principal eigenvector v' is: calculating the inner product of the new main characteristic vector v ' and the main characteristic vector v, if the inner product result is positive, keeping the sign of the new main characteristic vector v ', otherwise, changing the sign of the new main characteristic vector v ';
(6) determining whether a specified second tracking stop condition is met, if so, executing the step (7); otherwise, taking the interception point x as the current starting point x0Taking the new main feature vector v' as the current feature vector v to continue to execute the steps (2) to (6);
(7) i.e. the tracking of the fibres in the current tracking direction is ended and is resumed from the current starting point x0Initially, transform x0The sign of the main feature vector v is tracked reversely (steps (2) to (5)) is executed;
when the tracing in both directions is completed, based on the current starting point x0Obtaining a fiber coordinate chain according to the tracking results in the two directions, determining whether the length of the fiber coordinate chain is within a specified length range, if so, performing three-dimensional curve interpolation and smoothing on the fiber coordinate chain, and creating an attribute table of the fiber bundle;
wherein the first tracking stop condition includes:
current starting point x0Reach the border of the whole brain image;
or the FA value on the interception point x is smaller than the set FA threshold;
the second tracking stop condition is: the angle between the current dominant eigenvector v and the new dominant eigenvector v' is greater than the set angle threshold.
The fiber tracking processing of the whole image is completed based on the fiber tracking imaging mode, so that an index pointer matrix with the same size as the original image is created, wherein each voxel stores a pointer pointing to a fiber chain from the voxel, and the reconstruction of the full white matter nerve fiber bundle is completed.
In a possible implementation manner, in the step of network-based statistical analysis, the network-based statistical analysis is performed on N sets of network connection matrices of structures to determine differences in network structures among the sets, and taking brain regions corresponding to the network structure differences that satisfy specified conditions as feature points includes:
taking brain areas as network topology nodes, and taking structural network connection measurement between the brain areas as association measurement between the network topology nodes;
setting a matrix D as a statistical model of network statistical analysis, wherein the matrix D at least comprises three columns, which are respectively: individual identification, group attribute, and gender attribute; all the individual identifiers are set to be 1 so as to represent that all the structural network connection matrixes are obtained;
evaluating a size comparison result of association metric hypotheses between the N sets of structural network connection matrices by specifying a hypothesis to be tested and a p-value to be estimated (a parameter for deciding a hypothesis test result) in a one-sided t-test (Student's t test) by a one-dimensional contrast vector having elements of 0 and 1;
calculating to obtain a sparse matrix of each structural network connection matrix in the N structural network connection matrix groups based on a preset association measurement threshold;
based on the preset multiple comparison times, the nonparametric statistical analysis of clustering and multiple comparison on the topological space is carried out on the N structural network connection sparse matrix groups, and comprises the following steps: clustering edge connection on a topological space of the sparse matrix of the structural network connection through breadth search (DFS) and depth search (BFS) to obtain a plurality of sub-network connectors; the corrected p-value of each sub-network connector FWER (Family-wise error rate) is calculated, an array test of multiple comparisons is performed, and a certain number (denoted as X) of sub-network connectors are selected as network structure differences satisfying a predetermined condition based on the p-value.
In one possible implementation, the network structure difference satisfying the specified condition may be: and selecting a certain number of sub-network connectors from the sub-network connectors with the p values meeting the specified conditions.
On the other hand, the embodiment of the invention also provides a data analysis device, which comprises a preprocessing module, a structural network connection matrix construction module, a feature extraction module and a model establishment module;
the preprocessing module executes any one of the preprocessing steps of the embodiment of the invention to obtain N whole brain basic data sets;
a structural network connection matrix construction module, which executes any one of the structural network connection matrix construction steps of the embodiment of the invention on the N whole-brain basic data sets to obtain N structural network connection matrix sets;
the characteristic extraction module is used for taking the specified structural network connection as a characteristic object, extracting the characteristic of each structural network connection matrix in the N structural network connection matrix groups based on the characteristic object, generating a characteristic value vector and obtaining N characteristic value vector groups;
and the model establishing module is used for classifying according to the N characteristic value vector groups and establishing a classification model.
In another aspect, an embodiment of the present invention provides a computer device, where the computer device includes a processor and a memory, where the memory stores at least one computer program, and the at least one computer program is loaded and executed by the processor to implement any of the above data analysis methods.
In another aspect, an embodiment of the present invention provides a computer-readable storage medium, where at least one computer program is stored in the computer-readable storage medium, and the at least one computer program is loaded and executed by a processor to implement any of the data analysis methods described above.
The technical scheme provided by the embodiment of the invention at least has the following beneficial effects:
after the full brain white matter nerve fiber bundles are reconstructed, the full brain white matter nerve fiber bundles are registered with a plurality of divided brain areas, the structural network connection matrix of the full brain is obtained through calculation, the structural network connection matrix of each full brain is used as reference data, and network-based statistical analysis is carried out on a plurality of groups of full brain data, so that the brain data can be subjected to overall analysis, and the structural connection and the network among the brain subareas can be obtained.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
FIG. 1 is a flow chart of a method of data analysis provided by an embodiment of the present invention;
FIG. 2 is a flowchart of a method for reconstructing a white matter nerve fiber bundle in the whole brain in a data analysis method provided by an embodiment of the present invention;
FIG. 3 is an exemplary diagram of a characteristic curve provided by an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a data analysis apparatus according to an embodiment of the present invention.
Detailed Description
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
Referring to fig. 1, fig. 1 is a flowchart of a method of a data analysis method according to an embodiment of the present invention, and the data analysis method shown in fig. 1 can perform global analysis on brain data, so as to obtain structural connections and networks between brain partitions. The method comprises the following steps:
in step S1, a preprocessing operation is performed on each of the N whole brain image data sets.
In this embodiment, N is a positive integer greater than or equal to 2.
According to the embodiment of the invention, N groups of subjects with different symptoms provide whole brain data, and whole brain image data of each subject is acquired through a nuclear magnetic resonance device, wherein each group of subjects comprises a plurality of persons. Specifically, whole brain image data of each subject can be acquired by the nuclear magnetic resonance diffusion tensor imaging technology. It should be noted that, in order to ensure the accuracy of the data, each group of subjects maintains the same mental state, for example, being awake, sleeping naturally, or sleeping under sedation, when the image data is acquired. Wherein, when the image data is collected in the waking state, the eye opening or eye closing states of the tested object are kept consistent; when image data are collected in a natural sleep state, the data collection time is kept consistent after the user falls asleep; when image data acquisition is carried out in a sleeping state after sedation, each subject takes a sedative drug, a dosage form and a drug administration mode in a consistent manner, and the data acquisition time is kept in a consistent manner after falling asleep.
The device for executing the scheme is arranged in the computer, so that the whole brain image data of the tested object is acquired by the nuclear magnetic resonance equipment and then input into the computer, and the computer executes analysis operation on the whole brain image data.
Specifically, as the original data, not only the data format of the whole brain image data cannot be recognized by the computer, but also other attributes of the whole brain image data need to be adjusted, and based on this, the computer needs to preprocess each whole brain image data after receiving the whole brain image data input by the nuclear magnetic resonance equipment. Wherein, each whole brain image data is preprocessed, and the specific implementation is as follows:
first, image data in DICOM (Digital Imaging and Communications in Medicine) format is converted into NIFTI format. Secondly, image data (b0 image) without gradient weighting is extracted; then, a skull removal processing operation is performed on the b0 image data; further, performing eddy current correction on the image data after the de-heading bone treatment; and finally, carrying out tensor fitting operation on the corrected whole brain image data to obtain N whole brain basic data sets.
It should be noted that each implementation of the above pre-processing is well known to those skilled in the art, and the present invention is not described in detail herein.
Step S2, reconstructing the whole brain white matter nerve fiber bundles of each whole brain basic data.
Performing fiber tracking imaging according to an anisotropy Fraction (FA) from each of the N whole brain basis data sets, wherein the FA is calculated from the magnetic resonance gradient direction and the average gradient, and ranges between 0 and 1, using the formula:
wherein,
wherein λ is1，λ2，λ3For eigenvectors v representing gradient directions in diffusion tensor1、v2And v3Three characteristic values of (i.e.. lambda.)1，λ2And λ3The diffusion tensor in the voxel values constituting a voxel.
Specifically, referring to fig. 2, fig. 2 is a specific flowchart of the fiber-tracking imaging process involved in step S2 in fig. 1, i.e., reconstructing the pan-white matter nerve fiber bundles of each pan-brain basis data in the N pan-brain basis data sets using the fiber-tracking imaging strategy. The process comprises the following steps:
step S201, starting point x traced by fiber bundle0At the beginning, wherein,
x0＝(x0,y0,z0)
x0for points with FA values greater than the set FA threshold, (x)0,y0,z0) Represents a starting point x0Three-dimensional position coordinates of (a);
step S202, at x0Upper principal eigenvector v (max (λ)1,λ2,λ3) The corresponding feature vector) as the current direction, wherein,
v＝(vx,vy,vz)
vx,vy,vzrepresenting the components of v in three dimensions of a three-dimensional space;
step S203, parameterizing the fiber tracking track in the three-dimensional space by specifying the arc length S to obtain the current tracking position coordinate of the fiber:
the fiber tracking trajectory leaves the current voxel into the intercept point x of the next voxel, where,
x＝(x,y,z)
step S204, if the fiber tracking reaches the boundary of the whole brain image, jumping to step S209;
step S205, if the FA value on the intercept point x of the next voxel is smaller than the set FA threshold, the step S209 is skipped;
step S206, calculating the inner product of the new feature vector and the current feature vector v, if the inner product result is positive, keeping the sign of the new feature vector, and if the inner product result is negative, changing the sign of the new feature vector (namely negating);
wherein the new feature vector is: the dominant feature vector on x, which may be labeled v';
step S207, if the angle between the current feature vector v and the new feature vector is larger than a specified angle threshold, jumping to step S209; otherwise, go to step S208;
step S208, assigning x to x0Jumping to step S202;
step S209, when the tracking in the direction of the feature vector is finished, the starting point x is restarted0At first, x is converted0The sign of the characteristic vector v is tracked reversely;
step S210, if the starting point isx0After tracking in both directions is completed, the starting point x is combined0Obtaining a fiber coordinate chain according to the tracking results in the two directions;
step S211, determining whether the length of the fiber coordinate chain is greater than a minimum fiber length threshold value and less than a maximum fiber length threshold value, if so, performing three-dimensional curve interpolation and smoothing on the fiber coordinate chain, and creating an attribute table;
step S212, repeatedly applying the steps S201 to S211 to track until the whole image processing is completed;
step S213, creating an index pointer matrix with the same size as the original image, wherein each voxel stores a pointer pointing to the fiber chain from the voxel, and the full-white-matter fiber bundle reconstruction is completed.
Step S3, calculating the structural network connection between brain regions for each of the whole white matter fiber bundle data.
On the basis of the above description, N whole brain white matter nerve fiber bundle data sets can be correspondingly obtained from N whole brain matter nerve fiber bundle data sets, and according to each whole brain white matter nerve fiber bundle data set in the N whole brain matter nerve fiber bundle data sets, the data is registered with a plurality of pre-divided brain regions, and a structural network matrix is calculated to obtain N structural network connection matrix sets, which specifically includes:
dividing the whole brain into a plurality of brain partitions according to a preset standard template;
respectively acquiring corresponding FA basic data between corresponding brain partitions from each whole brain white matter nerve fiber bundle data;
calculating the average value of FA among a plurality of brain partitions pre-divided by each whole brain, and taking the average value of FA among the brain partitions as the structural network connection measurement among the brain partitions, namely the FA connection value among the plurality of brain partitions pre-divided by the whole brain to obtain a structural network connection matrix A,
wherein, the structural netElement fa in the network connection matrix ai,jDenotes the FA connection value between the brain partitions, i, j ═ 1,2, …, n, n denotes the number of the brain partitions.
And step S4, performing network statistical analysis on the N structural network connection matrix groups to realize feature extraction.
On the basis of the above steps, the network-based statistical analysis is performed on N structural network connection matrix groups, and X structural connection networks with significant inter-group differences are obtained by calculation, which specifically includes:
taking a coordinate set of a plurality of brain partitions (brain areas) divided by the whole brain under a standard space (MNI (mental Neurological institute) space) as a network topology node set, namely taking each brain partition as a node, thereby obtaining the coordinates of the network topology node based on the coordinates of the brain partitions and further obtaining the network topology node set;
taking each structural network connection matrix A in the N structural network connection matrix groups as a network connection matrix set element, wherein element values in the structural network connection matrix A are correlation metrics among network topology nodes;
the design matrix D is set as a statistical model based on statistical analysis of the network, wherein,
D＝[1 Gr Ge Ag]
the number of rows of the matrix D is the number of the structural network connection matrixes, and the matrixes correspond to the structural network connection matrixes in the N structural network connection matrix groups one by one; the matrix D can be composed of four columns, wherein the elements in the first column of the matrix are individual identifications and are all set to be 1 so as to obtain all structural network connection matrixes; the second column Gr is a group attribute; the third column Ge is gender attribute; the fourth column Ag is an age attribute. Wherein the age attribute is an optional column of the matrix D.
The size of the association metric (structural connection metric) hypothesis between the N sets of structural network connection matrices modeled is evaluated by specifying the hypothesis to be examined and the p-value to be estimated in the one-sided t-test by a one-dimensional contrast vector whose elements are 0 and 1.
And setting a threshold value for the correlation measurement in each structural network connection matrix in the N structural network connection matrix groups, and calculating to obtain a sparse matrix of each structural network connection matrix in the N structural network connection matrix groups. In the sparse matrix, only the correlation metric exceeding the set threshold is stored, so that N structural network connection sparse matrix groups are obtained.
Setting the times of multiple comparison, and carrying out clustering on the topological space and nonparametric statistical analysis of the multiple comparison on the N structural network connection sparse matrix groups, wherein the nonparametric statistical analysis comprises the following steps:
clustering of network connectors (structure network connection sparse matrix) in a topological space is carried out through breadth search and depth search, a plurality of sub-network connectors are obtained, namely edge connection clustering is carried out on the network connectors, and network connections corresponding to all edges included in each clustering result correspond to one sub-network connector.
Setting the number of times of multiple comparisons as n ', calculating a p value after correction of each sub-network connector FWER, and performing arrangement inspection of the multiple comparisons n', wherein the p value is an estimation of a proportion of the sub-network connectors which is larger than or equal to the maximum sub-network connector, namely whether the sub-network connector is the sub-network connector with the maximum size and whether a significant difference exists between groups (the difference is smaller than a certain difference threshold), thereby obtaining a plurality (marked as X) of structural connection networks with significant differences between groups.
And step S5, establishing a classification model.
On the basis of the steps, taking the structure connection in the X structure connection networks as the characteristics, performing characteristic extraction on each structure network connection matrix in the N structure network connection matrix groups to generate characteristic value vectors to obtain N characteristic value vector groups; and classifying according to the N characteristic value vector groups.
In the embodiment of the invention, the structural connection in the X structural connection networks is used as the characteristic, the N characteristic value vector groups are used as input samples, and the parameters of the classification model are determined by adopting algorithms such as a support vector machine and the like. Furthermore, in the subsequent treatment process, the result can be output through a classification model, and True Positive (TP)/recall (call), True Negative (TN), False Negative (FN) and False Positive (FP) can be counted through the result, and the accuracy (accuracy), sensitivity (sensitivity), specificity (specificity), precision (precision) and F-value (F-Measure) can be further calculated. Wherein,
it should be noted that, for the unbalanced condition of the data samples, the SMOTE algorithm (Synthetic minimum optimization Technique) is used to synthesize the over-sampling of the Minority class according to the input samples, and the method of randomly and simply copying the nearest neighbor samples is adopted to increase the data samples of the Minority class and add the data samples to the data set, so as to avoid the model over-fitting phenomenon caused by the unbalanced samples of the training set.
In addition, a working characteristic curve of the tested object can be drawn according to the true positive rate and the false positive rate, and the correlation between sensitivity and specificity is revealed by a mapping method, wherein the larger the area under the curve is, the higher the classification accuracy is.
The evaluation index for classifying the subjects among the multiple groups can be simplified into a plurality of groups of two-classification problems, namely, the evaluation numerical score after classifying each group of subjects and other groups (all other groups are taken as a whole), and the overall sensitivity, specificity and accuracy of the multiple groups of classifications are weighted averages of evaluation results of the two-classification problems.
In summary, in order to solve the problem that the data result obtained in the prior art is more comprehensive, in the embodiment of the present invention, after N whole brain image data sets are received and a preprocessing operation is performed on each whole brain image data in the N whole brain image data sets to obtain N whole brain basis data sets, a fiber tracking imaging algorithm may be used to reconstruct a whole brain white matter nerve fiber bundle of each whole brain basis data in the N whole brain basis data sets, and then indexes such as an anisotropic score of each whole brain basis data are calculated to obtain N whole brain white matter nerve fiber bundle data sets; then, registering each full-white-matter nerve fiber bundle data in the N full-white-matter nerve fiber bundle data groups with a plurality of pre-divided brain areas, respectively acquiring corresponding FA basic data between corresponding brain areas from each full-white-matter nerve fiber bundle data in the N full-white-matter nerve fiber bundle data groups, calculating an average value of FA between the brain areas as a structural connection metric, generating a structural network matrix, and acquiring N structural network connection matrix groups; further, performing network-based statistical analysis on the N structural network connection matrix groups to calculate X structural connection networks with significant inter-group differences; further, taking the structure connection in the X structure connection networks as a feature, performing feature extraction on each structure network connection matrix in the N structure network connection matrix groups to generate a feature value vector to obtain N feature value vector groups; and classifying according to the N characteristic value vector groups, and establishing a classification model. Therefore, according to the scheme, after the white matter nerve fiber bundles of the whole brain are reconstructed, the white matter nerve fiber bundles of the whole brain are registered with the plurality of divided brain areas, the structural network connection matrix of the whole brain is obtained through calculation, the structural network connection matrix of each whole brain is used as reference data, network-based statistical analysis is carried out on the multiple groups of whole brain data, therefore, the brain data can be subjected to overall analysis, and the structural connection and the network among the brain subareas can be obtained.
The following describes an implementation process of establishing a classification model according to an embodiment of the present invention with reference to an example.
Referring to fig. 3, fig. 3 is an exemplary graph of characteristic curves roc (receiver operating characteristic curve) provided by an embodiment of the present invention, wherein horizontal and vertical axes represent True Positive Rate (TPR) and False Positive Rate (FPR), respectively. In the example shown in fig. 3, a total of 121 brain images of children are acquired, wherein the asd (adaptive Spectrum recorder) group totally includes 95 brain images of autism children, the TDC group totally includes 26 brain images of normal children, and the image format conversion, the preprocessing and the fiber tracking imaging algorithm operations are sequentially performed on each of the two groups of brain images to reconstruct the full white matter nerve fiber bundles, so as to obtain two groups of data sets of full white matter nerve fiber bundles, where the setting parameters are as follows: the angle threshold is 60 °; the shortest threshold value of the fiber bundle is 30 mm; the longest threshold value of the fiber bundle is 300 mm; the number of the fiber bundles is 40000. And (3) superposing each whole white matter nerve fiber bundle in the two groups of whole white matter nerve fiber bundle data sets with a user-defined brain map (294 brain regions), and calculating a structural network connection matrix to obtain two groups of structural network connection matrix sets. Taking 294 brain areas as network nodes, taking structural network connection matrix elements as network edge measurement, taking two groups of structural network connection matrix groups as input, executing network statistics-based operation, obtaining 5 brain structure connection networks with significant difference between the two groups, and totaling 22 network connections, wherein the set parameters are as follows: fa threshold 3.0; the multiple comparison times are 1000 times. Taking 22 network connections as features, extracting the features of each structural network connection matrix in the two groups of structural network connection matrix groups to form a feature vector with the dimension of 1 multiplied by 22, and obtaining two groups of feature value vector groups. And establishing and training two classification models by taking the two groups of feature vector groups as training sets, and respectively setting labels according to the groups, wherein the label '1' represents an ASD group, and the label '2' represents a TDC group. The model is built mainly by adopting radial basis functions of a support vector machine, can be built and trained by using a weka toolbox based on java language and a lim-svm based on Matlab, and is verified by using a leave-one-out cross-validation algorithm and an additional test set, wherein the model is mainly set by the following parameters: penalty coefficient (cost) is 0.722; kernel parameter (gamma) 1.4044; the termination condition (eps) was 0.001. In addition, the model adopts the SMOTE algorithm to solve the problem of model overfitting caused by unbalanced training set samples. In this example, model cross-validation distinguished two groups of subjects with an accuracy of 91.62%, a sensitivity of 88.40%, a specificity of 94.80%, an accuracy of 94.40%, an F-Measure of 91.30%, and an area under the curve of the ROC curve of 0.9940; the accuracy of the model for distinguishing two groups of test objects in the test set is 90.5%; sensitivity was 91.7%, specificity was 88.9%, accuracy was 91.7%, and F-Measure was 91.7%. It follows that the model may assist a physician as a diagnostic aid for autism in children.
Therefore, according to the scheme, after the full brain white matter nerve fiber bundles are reconstructed, the full brain white matter nerve fiber bundles are registered with the plurality of divided brain areas, the structural network connection matrix of the full brain is obtained through calculation, the structural network connection matrix of each full brain is used as reference data, network-based statistical analysis is carried out on the multiple groups of full brain data, therefore, the brain data can be subjected to overall analysis on the basis of big data, and the structural connection and the network among the brain areas can be obtained. Taking the structural connection and the network as features, establishing a classification model by adopting machine learning algorithms such as a support vector machine, a random forest, a Bayesian classifier, a linear model and the like, and training, adjusting and verifying the model. The model can assist a physician in the work of assisting in diagnosing the relevant brain diseases.
Corresponding to the above implementation method, an embodiment of the present invention further provides a data analysis apparatus, referring to fig. 4, where fig. 4 is a schematic structural diagram of the data analysis apparatus provided in the embodiment of the present invention, and the apparatus includes a preprocessing module 11, a structural network connection matrix constructing module 12, a feature extracting module 13, and a model establishing module 14; the preprocessing module 11 is configured to, when N whole brain image data sets are received, perform preprocessing operation on each whole brain image data in the N whole brain image data sets to obtain N whole brain basic data sets; wherein, N is a positive integer greater than or equal to 2, and each whole brain image data group comprises a plurality of whole brain image data; the structural network connection matrix construction module 12 is configured to reconstruct the whole brain white matter nerve fiber bundle of each whole brain basic data in the N whole brain basic data sets, and calculate a designated index parameter of each whole brain basic data to obtain N whole brain white matter nerve fiber bundle data sets; registering each full-white-matter nerve fiber bundle data in the N full-white-matter nerve fiber bundle data groups with a plurality of pre-divided brain areas, taking the mean value of designated index parameters between the brain areas as the structural network connection measurement between the brain areas, and calculating a structural network connection matrix to obtain N structural network connection matrix groups; a feature extraction module 13, configured to use a specified structural network connection (i.e., a network edge) as a feature object, and perform feature extraction on each structural network connection matrix in the N structural network connection matrix groups based on the feature object, to generate a feature value vector, so as to obtain N feature value vector groups; and the model establishing module 14 is used for classifying according to the N characteristic connection value vector groups and establishing a classification model.
In a possible implementation manner, the specified structural network connection is obtained based on any of the above network statistical analysis steps of the embodiments of the present invention.
It should be noted that, when the apparatus provided in the foregoing embodiment implements the functions thereof, only the division of the functional modules is illustrated, and in practical applications, the above functions may be distributed by different functional modules according to needs, that is, the internal structure of the apparatus is divided into different functional modules to implement all or part of the functions described above. In addition, the apparatus and method embodiments provided by the above embodiments belong to the same concept, and specific implementation processes thereof are detailed in the method embodiments and are not described herein again.
In an exemplary embodiment, a computer device is also provided, the computer device comprising a processor and a memory, the memory having at least one computer program stored therein. The at least one computer program is loaded and executed by one or more processors to implement any of the data analysis methods described above.
In an exemplary embodiment, there is also provided a computer-readable storage medium having at least one computer program stored therein, the at least one computer program being loaded and executed by a processor of a computer device to implement any of the data analysis methods described above.
In one possible implementation, the computer-readable storage medium may be a Read-Only Memory (ROM), a Random Access Memory (RAM), a Compact Disc Read-Only Memory (CD-ROM), a magnetic tape, an optical data storage device, and the like.
Although the preferred embodiments of the present invention have been described, those skilled in the art will recognize the basic inventive preferred embodiments and all variations and modifications that fall within the scope of the present invention. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (10)
1. A method of data analysis, the method comprising:
a pretreatment step:
when N whole brain image data sets are received, performing preprocessing operation on each whole brain image data in the N whole brain image data sets to obtain N whole brain basic data sets; wherein, N is a positive integer greater than or equal to 2, and each whole brain image data group comprises a plurality of whole brain image data;
a structure network connection matrix construction step:
reconstructing a whole brain white matter nerve fiber bundle of each whole brain basic data in the N whole brain basic data groups, and calculating a designated index parameter of each whole brain basic data to obtain N whole brain white matter nerve fiber bundle data groups;
registering each full-white-matter nerve fiber bundle data in the N full-white-matter nerve fiber bundle data groups with a plurality of pre-divided brain areas, taking the mean value of designated index parameters between the brain areas as the structural network connection measurement between the brain areas, and calculating a structural network connection matrix to obtain N structural network connection matrix groups;
network statistical analysis:
performing network-based statistical analysis on the N structural network connection matrix groups to determine the network structure difference among the groups, and taking the structural network connection corresponding to the network structure difference meeting the specified conditions as a characteristic object;
a characteristic extraction step:
based on the feature object, performing feature extraction on each structure network connection matrix in the N structure network connection matrix groups to generate feature value vectors to obtain N feature value vector groups;
a model establishing step:
and classifying according to the N characteristic connection value vector groups, and establishing a classification model.
2. The data analysis method according to claim 1, wherein the preprocessing step, for each of the N whole brain image data sets, performs a preprocessing operation including:
carrying out format conversion on the whole brain image data, extracting image data without gradient weighting, and carrying out processing of deleting non-brain parts on the obtained image data;
and performing eddy current correction on the image data from which the non-brain part is deleted, and performing tensor fitting operation on the image data after the eddy current correction to obtain diffusion tensor imaging data which are used as whole brain basic data.
3. The data analysis method of claim 1, wherein in the structural network connection matrix construction step, the index parameter is designated as an anisotropy Fraction (FA).
4. The data analysis method as claimed in claim 1, wherein in the step of constructing the structural network connection matrix, when reconstructing the white matter nerve fiber bundles of the whole brain of each of the N whole brain basic data sets according to the FA, a fiber tracking imaging method is adopted, which specifically includes:
(1) determining starting point x for fiber bundle tracking0＝(x0,y0,z0) Thereby obtaining a starting point x0A feature vector of (2), wherein x0The FA value is larger than the set FA threshold value;
(2) determining the current tracking direction: the current starting point x0Is given as (v) is given as the principal eigenvector of (v ═ vx,vy,vz) As the current tracking direction;
(3) determining an intercept point x ═ x, y, z of a next voxel of the fiber tracking trajectory, where the coordinates of x are specifically:
s represents a specified arc length;
(4) determining whether a specified first tracking stop condition is met, if so, executing the step (7); otherwise, executing the step (5);
(5) obtaining a new principal eigenvector v 'based on the diffusion tensor of the interception point x, and executing the step (6) after determining the direction of the new principal eigenvector v';
wherein, the direction of the new principal eigenvector v' is: calculating the inner product of the new main characteristic vector v ' and the main characteristic vector v, if the inner product result is positive, keeping the sign of the new main characteristic vector v ', otherwise, changing the sign of the new main characteristic vector v ';
(6) determining whether a specified second tracking stop condition is met, if so, executing the step (7); otherwise, taking the interception point x as the current starting point x0Taking the new main feature vector v' as the current feature vector v to continue to execute the steps (2) to (6);
(7) i.e. the tracking of the fibres in the current tracking direction is ended and is resumed from the current starting point x0Initially, transform x0Main characteristic of (1)The sign of the vector v is tracked in the reverse direction (steps (2) to (5) are executed);
when the tracing in both directions is completed, based on the current starting point x0Obtaining a fiber coordinate chain according to the tracking results in the two directions, determining whether the length of the fiber coordinate chain is within a specified length range, if so, performing three-dimensional curve interpolation and smoothing on the fiber coordinate chain, and creating an attribute table of the fiber bundle;
wherein the first tracking stop condition includes:
current starting point x0Reach the border of the whole brain image;
or the FA value on the interception point x is smaller than the set FA threshold;
the second tracking stop condition is: the angle between the current dominant eigenvector v and the new dominant eigenvector v' is greater than the set angle threshold.
5. The data analysis method according to claim 1, wherein in the network statistical analysis step, the network statistical analysis is performed on N sets of the structural network connection matrix to determine differences in network structures among the sets, and the taking of the brain regions corresponding to the network structure differences satisfying the specified conditions as the feature points includes:
taking brain areas as network topology nodes, and taking structural network connection measurement between the brain areas as association measurement between the network topology nodes;
setting a matrix D as a statistical model of network statistical analysis, wherein the matrix D at least comprises three columns, which are respectively: individual identification, group attribute, and gender attribute; all the individual identifiers are set to be 1 so as to represent that all the structural network connection matrixes are obtained;
the method comprises the steps of specifying a hypothesis to be tested and a p value to be estimated in a one-dimensional t test through a one-dimensional contrast vector with elements of 0 and 1, and evaluating the size comparison result of correlation measurement hypotheses among N structural network connection matrix groups;
calculating to obtain a sparse matrix of each structural network connection matrix in the N structural network connection matrix groups based on a preset association measurement threshold;
based on the preset multiple comparison times, the nonparametric statistical analysis of clustering and multiple comparison on the topological space is carried out on the N structural network connection sparse matrix groups, and comprises the following steps: clustering edge connection on a topological space of the structure network connection sparse matrix through breadth search and depth search to obtain a plurality of sub-network connectors; and calculating the corrected p value of each sub-network connector FWER, performing multiple comparison arrangement tests, and selecting a certain number of sub-network connectors as the network structure difference meeting the specified conditions based on the p value.
6. The data analysis method according to claim 5, wherein in the network-based statistical analysis step, the network structure difference satisfying the specified condition is: and selecting a certain number of sub-network connectors from the sub-network connectors with p values meeting the specified conditions.
7. A data analysis device is characterized by comprising a preprocessing module, a structural network connection matrix construction module, a feature extraction module and a model establishment module;
wherein, the preprocessing module executes the preprocessing steps of claim 1 or 2 to obtain N whole brain basic data sets;
a structural network connection matrix construction module for performing the structural network connection matrix construction steps of claims 1, 3 or 4 on the N whole brain basis data sets to obtain N structural network connection matrix sets;
the characteristic extraction module is used for taking the specified structural network connection as a characteristic object, extracting the characteristic of each structural network connection matrix in the N structural network connection matrix groups based on the characteristic object, generating a characteristic value vector and obtaining N characteristic value vector groups;
and the model establishing module is used for classifying according to the N characteristic value vector groups and establishing a classification model.
8. The data analysis device of claim 7, wherein the specified structural network connection in the feature extraction module is obtained based on the network statistical analysis step of claim 1 or 5.
9. A computer device comprising a processor and a memory, the memory having stored therein at least one computer program that is loaded and executed by the processor to implement a data analysis method as claimed in any one of claims 1 to 6.
10. A computer-readable storage medium, in which at least one computer program is stored, which is loaded and executed by a processor to implement the data analysis method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110359402.0A CN113283465B (en) | 2021-04-02 | 2021-04-02 | Diffusion tensor imaging data analysis method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110359402.0A CN113283465B (en) | 2021-04-02 | 2021-04-02 | Diffusion tensor imaging data analysis method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113283465A CN113283465A (en) | 2021-08-20 |
| CN113283465B true CN113283465B (en) | 2022-04-29 |
Family
ID=77276329
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110359402.0A Active CN113283465B (en) | 2021-04-02 | 2021-04-02 | Diffusion tensor imaging data analysis method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113283465B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117172294B (en) * | 2023-11-02 | 2024-01-26 | 烟台大学 | Method, system, equipment and storage medium for constructing sparse brain network |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101763638A (en) * | 2009-12-14 | 2010-06-30 | 西北工业大学 | Method for classifying cerebral white matter fiber tracts in diffusion tensor nuclear magnetic resonance image |
| CN102332054A (en) * | 2011-08-25 | 2012-01-25 | 西北工业大学 | Method for identifying cerebral cortex surface marking region of three-dimensional cerebral magnetic resonance image |
| CN103020653A (en) * | 2012-12-07 | 2013-04-03 | 中国科学院自动化研究所 | Structure and function magnetic resonance image united classification method based on network analysis |
| CN107194918A (en) * | 2017-05-16 | 2017-09-22 | 北京大学 | Data analysing method and device |
| CN109102492A (en) * | 2018-07-02 | 2018-12-28 | 重庆邮电大学 | A kind of functional magnetic resonance imaging brain connection map construction method based on Cooperative Clustering |
| CN109308699A (en) * | 2018-10-29 | 2019-02-05 | 王新乐 | A kind of method of cranial nerve fiber image and its neurotransmitter fusion of imaging |
| CN110018432A (en) * | 2019-02-14 | 2019-07-16 | 清华大学 | A kind of Neuroimaging methods based on dispersion tensor, device and MR imaging apparatus |
| CN110689536A (en) * | 2019-09-30 | 2020-01-14 | 深圳大学 | Method and device for tracking grey matter and white matter of brain based on multi-modal magnetic resonance image |
| CN111340821A (en) * | 2020-02-20 | 2020-06-26 | 太原理工大学 | Deviation detection method of brain structure network based on module connection |
| CN111415324A (en) * | 2019-08-09 | 2020-07-14 | 复旦大学附属华山医院 | Classification and identification method of brain lesion image space distribution characteristics based on magnetic resonance imaging |
| CN111543949A (en) * | 2020-05-13 | 2020-08-18 | 北京航空航天大学 | Child ASD diagnosis device based on magnetoencephalogram and electroencephalogram |
| CN111627553A (en) * | 2020-05-26 | 2020-09-04 | 四川大学华西医院 | Method for constructing individualized prediction model of first-onset schizophrenia |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| NZ702800A (en) * | 2010-06-25 | 2017-03-31 | Shire Human Genetic Therapies | Methods and compositions for cns delivery of heparan n-sulfatase |
| WO2017153982A1 (en) * | 2016-03-06 | 2017-09-14 | Yeda Research And Development Co. Ltd. | Method for modulating myelination |
-
2021
- 2021-04-02 CN CN202110359402.0A patent/CN113283465B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101763638A (en) * | 2009-12-14 | 2010-06-30 | 西北工业大学 | Method for classifying cerebral white matter fiber tracts in diffusion tensor nuclear magnetic resonance image |
| CN102332054A (en) * | 2011-08-25 | 2012-01-25 | 西北工业大学 | Method for identifying cerebral cortex surface marking region of three-dimensional cerebral magnetic resonance image |
| CN103020653A (en) * | 2012-12-07 | 2013-04-03 | 中国科学院自动化研究所 | Structure and function magnetic resonance image united classification method based on network analysis |
| CN107194918A (en) * | 2017-05-16 | 2017-09-22 | 北京大学 | Data analysing method and device |
| CN109102492A (en) * | 2018-07-02 | 2018-12-28 | 重庆邮电大学 | A kind of functional magnetic resonance imaging brain connection map construction method based on Cooperative Clustering |
| CN109308699A (en) * | 2018-10-29 | 2019-02-05 | 王新乐 | A kind of method of cranial nerve fiber image and its neurotransmitter fusion of imaging |
| CN110018432A (en) * | 2019-02-14 | 2019-07-16 | 清华大学 | A kind of Neuroimaging methods based on dispersion tensor, device and MR imaging apparatus |
| CN111415324A (en) * | 2019-08-09 | 2020-07-14 | 复旦大学附属华山医院 | Classification and identification method of brain lesion image space distribution characteristics based on magnetic resonance imaging |
| CN110689536A (en) * | 2019-09-30 | 2020-01-14 | 深圳大学 | Method and device for tracking grey matter and white matter of brain based on multi-modal magnetic resonance image |
| CN111340821A (en) * | 2020-02-20 | 2020-06-26 | 太原理工大学 | Deviation detection method of brain structure network based on module connection |
| CN111543949A (en) * | 2020-05-13 | 2020-08-18 | 北京航空航天大学 | Child ASD diagnosis device based on magnetoencephalogram and electroencephalogram |
| CN111627553A (en) * | 2020-05-26 | 2020-09-04 | 四川大学华西医院 | Method for constructing individualized prediction model of first-onset schizophrenia |
Non-Patent Citations (4)
| Title |
|---|
| White Matter Abnormalities in Major Depression Biotypes Identified by Diffusion Tensor Imaging;Sugai Liang等;《Neurosciencs Bulletin》;20191001;第35卷(第5期);867–876 * |
| 弥散张量成像在检测颅内肿瘤与周围脑白质神经纤维束解剖关系中的临床应用价值;李虹等;《中国医科大学学报》;20201203;第49卷(第12期);1125-1129 * |
| 磁共振弥散张量成像技术在脑小血管病变与认知功能障碍的关系的研究进展;王瑾等;《世界最新医学信息文摘》;20180216;第18卷(第14期);35-37 * |
| 肠道菌群与孤独症发病关系的研究进展;谢宇非等;《生理科学进展》;20200225;第51卷(第1期);15-20 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113283465A (en) | 2021-08-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20220148191A1 (en) | Image segmentation method and apparatus and storage medium | |
| Sahu et al. | FINE_DENSEIGANET: Automatic medical image classification in chest CT scan using Hybrid Deep Learning Framework | |
| CN113616184B (en) | Brain network modeling and individual prediction method based on multi-mode magnetic resonance image | |
| WO2023077603A1 (en) | Prediction system, method and apparatus for abnormal brain connectivity, and readable storage medium | |
| Putra et al. | Enhanced skin condition prediction through machine learning using dynamic training and testing augmentation | |
| CN112735570B (en) | Image-driven brain atlas construction method, device, equipment and storage medium | |
| Wang et al. | Learning methods of convolutional neural network combined with image feature extraction in brain tumor detection | |
| US8861815B2 (en) | Systems and methods for modeling and processing functional magnetic resonance image data using full-brain vector auto-regressive model | |
| CN113255728B (en) | Depression classification method based on graph embedding and multi-mode brain network | |
| CN109325942A (en) | Eye fundus image Structural Techniques based on full convolutional neural networks | |
| CN111009321A (en) | Application method of machine learning classification model in juvenile autism auxiliary diagnosis | |
| CN113610808B (en) | Group brain map individuation method, system and equipment based on individual brain connection diagram | |
| CN112418337B (en) | Multi-feature fusion data classification method based on brain function hyper-network model | |
| CN112381178B (en) | Medical image classification method based on multi-loss feature learning | |
| WO2022121100A1 (en) | Darts network-based multi-modal medical image fusion method | |
| JP7191443B2 (en) | Target object attribute prediction method based on machine learning, related equipment and computer program | |
| CN111382807B (en) | Image processing method, image processing device, computer equipment and storage medium | |
| CN116110597B (en) | Digital twinning-based intelligent analysis method and device for patient disease categories | |
| WO2024216672A1 (en) | 3d segmentation method and system based on diffusion model and 3d transformer, device, and medium | |
| CN114038564B (en) | Noninvasive risk prediction method for diabetes | |
| Chen et al. | Generative adversarial U-Net for domain-free medical image augmentation | |
| WO2023108873A1 (en) | Brain network and brain addiction connection calculation method and apparatus | |
| CN115272295A (en) | Dynamic brain function network analysis method and system based on time domain-space domain combined state | |
| CN113786185A (en) | Static brain network feature extraction method and system based on convolutional neural network | |
| CN115330748A (en) | Brain image processing method, computer device, and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230706 Address after: Room 101-16, Floor 6, Building 9, Zone 2, Yard 31, Kechuang 13th Street, Beijing Economic-Technological Development Area, 102600 Patentee after: Beijing Beishan Medical Technology Co.,Ltd. Address before: 611731, No. 2006, West Avenue, Chengdu hi tech Zone (West District, Sichuan) Patentee before: University of Electronic Science and Technology of China Patentee before: Peking University |
|
| TR01 | Transfer of patent right |