CN115272295A - Dynamic brain function network analysis method and system based on time domain-space domain combined state - Google Patents
Dynamic brain function network analysis method and system based on time domain-space domain combined state Download PDFInfo
- Publication number
- CN115272295A CN115272295A CN202211078611.9A CN202211078611A CN115272295A CN 115272295 A CN115272295 A CN 115272295A CN 202211078611 A CN202211078611 A CN 202211078611A CN 115272295 A CN115272295 A CN 115272295A
- Authority
- CN
- China
- Prior art keywords
- time
- brain function
- dynamic
- state
- independent components
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30016—Brain
Abstract
The invention belongs to the technical field of medical image analysis, and provides a dynamic brain function network analysis method and system based on a time domain-space domain combined state. In the time dimension, intercepting a preprocessed four-dimensional resting state functional magnetic resonance image data segment according to a preset sliding window, and calculating Pearson correlation coefficients of time sequence signals between any two brain regions in the window segment to obtain a dynamic brain function connection matrix; extracting the dynamic brain function connection matrix by using an independent component analysis method to obtain independent components corresponding to individuals and time sequences corresponding to the independent components; based on the individual corresponding independent components and the time sequences corresponding to the independent components, a time domain-space domain fusion discrimination model is constructed by utilizing an efficient forward search strategy and a classifier, a specific brain dynamic network related to the specific neuropsychiatric disease is identified, and a decision value is output, so that the quantitative measurement of the dynamic brain functional network on the individual level is realized.
Description
Technical Field
The invention belongs to the technical field of medical image analysis, and particularly relates to a dynamic brain function network analysis method and system based on a time domain-space domain combined state.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
At present, the functional magnetic resonance imaging technology is widely used for basic research and clinical application of brain functions of neuropsychiatric diseases, and the technology can record brain function activities in a four-dimensional space, namely, can represent one-dimensional time function activities of voxels in a three-dimensional space. The traditional brain function network analysis method assumes that a brain network is in a steady state, namely, a brain function network mode does not change along with time, extracts a steady state brain function network based on an interactive relation between brain signal measurement brain intervals in acquisition time, and mainly focuses on a spatial specific mode and a model structure of brain functions in the process of analyzing the steady state brain network. However, the steady state brain function network is not sufficient to reflect the complex time-varying nature of the brain system, and dynamic brain network analysis is considered to be useful in revealing the iconographic features of the disease and the corresponding brain change mechanisms.
The most common method for estimating the dynamic brain network based on the resting state functional magnetic resonance image is a sliding time window algorithm. A large number of researches use the dynamic function connection matrix obtained by the algorithm to describe and compare and verify the network characteristics of the neuropsychiatric disease patients, including clustering or dynamic causal analysis of the brain dynamic network space connection state and the like. Although a large amount of research analyzes the relation between the time-varying activities of the brain and the functional connection thereof, an inter-group statistical method is generally used or discriminant analysis is performed based on network spatial attributes, quantitative indexes and biomarkers for effectively and comprehensively depicting a dynamic network are lacked, and individual level analysis and individualized diagnosis and treatment are difficult to realize. Furthermore, the black box decision nature and large sample requirements of deep neural networks make them limited in the decision-making task of neuropsychiatric diseases.
Disclosure of Invention
In order to solve the technical problems in the background art, the invention provides a dynamic brain function network analysis method and system based on a time domain-space domain combined state, which can identify brain network biomarkers related to neuropsychiatric diseases, realize quantitative analysis of a group dynamic brain function network and open up a new way for individual diagnosis and treatment of neuropsychiatric diseases.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention provides a dynamic brain function network analysis method based on a time domain-space domain combined state.
The dynamic brain function network analysis method based on the time domain-space domain combined state comprises the following steps:
acquiring four-dimensional resting state functional magnetic resonance image data of a specific neuropsychiatric disease patient and a normal control group, and preprocessing the four-dimensional resting state functional magnetic resonance image data;
in the time dimension, intercepting a preprocessed four-dimensional resting state functional magnetic resonance image data segment according to a preset sliding window, and calculating a Pearson correlation coefficient of a time sequence signal between any two brain areas in the window segment to obtain a dynamic brain function connection matrix;
extracting the dynamic brain function connection matrix by using an independent component analysis method to obtain independent components corresponding to individuals and time sequences corresponding to the independent components;
based on the individual corresponding independent components and the time sequences corresponding to the independent components, a time domain-space domain fusion discrimination model is constructed by utilizing an efficient forward search strategy and a classifier, a specific brain dynamic network related to the specific neuropsychiatric disease is identified, and a decision value is output, so that the quantitative measurement of the dynamic brain function network on the individual level is realized.
The second aspect of the invention provides a dynamic brain function network analysis system based on a time domain-space domain combined state.
A dynamic brain function network analysis system based on a time domain-space domain combined state comprises:
a data acquisition and pre-processing module configured to: acquiring four-dimensional resting state functional magnetic resonance image data of a specific neuropsychiatric disease patient and a normal control group, and preprocessing the four-dimensional resting state functional magnetic resonance image data;
a dynamic brain function connection matrix construction module configured to: in the time dimension, intercepting a preprocessed four-dimensional resting state functional magnetic resonance image data segment according to a preset sliding window, and calculating a Pearson correlation coefficient of a time sequence signal between any two brain areas in the window segment to obtain a dynamic brain function connection matrix;
a time-space joint state module configured to: extracting the dynamic brain function connection matrix by using an independent component analysis method to obtain independent components corresponding to individuals and time sequences corresponding to the independent components;
an analysis module configured to: based on the individual corresponding independent components and the time sequences corresponding to the independent components, a time domain-space domain fusion discrimination model is constructed by utilizing an efficient forward search strategy and a classifier, a specific brain dynamic network related to the specific neuropsychiatric disease is identified, and a decision value is output, so that the quantitative measurement of the dynamic brain functional network on the individual level is realized.
A third aspect of the invention provides a computer-readable storage medium.
A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the steps of the dynamic brain function network analysis method based on a spatio-temporal union state as described in the first aspect above.
A fourth aspect of the invention provides a computer apparatus.
A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the steps of the time-space domain joint state based dynamic brain function network analysis method according to the first aspect.
Compared with the prior art, the invention has the beneficial effects that:
the method firstly collects the resting state functional magnetic resonance image data of participants of different groups (a certain neuropsychiatric disease patient and a normal control group), and preprocesses the collected image data. And then, intercepting 60-second segments by using an overlapped sliding window in a time dimension, extracting four-dimensional data of each segment, and performing Pearson correlation calculation on the four-dimensional data by using a brain atlas template to obtain a functional connection matrix. And constructing a dynamic brain function connection matrix according to the function connection matrix corresponding to the reserved real time (the moment corresponding to the segment window). And extracting a functional connection state (space domain information) with individual specificity and spatial correspondence and a corresponding time fluctuation sequence (time domain information) of the functional connection state from the dynamic brain function connection matrix by using an independent component analysis method. And finally, constructing a time domain-space domain fusion classifier in the Riemannian manifold space by adopting a multivariate mode discrimination model, identifying a specific brain dynamic network related to the specific neuropsychiatric disease, and outputting a decision value to realize quantitative measurement of the dynamic brain functional network on an individual level. The invention can identify the brain network biomarkers related to the neuropsychiatric diseases, realize the quantitative analysis of the group dynamic brain function network and open up a new way for the individualized diagnosis and treatment of the neuropsychiatric diseases.
The invention carries out quantitative analysis based on the time domain information and the space domain information of the dynamic brain function network, can comprehensively and effectively depict the change of the dynamic brain function network, carries out multivariate discriminant analysis on the brain network in a united state in a Riemann manifold space, identifies the specific brain dynamic network with the most discriminant significance, quantificationally depicts the brain function dynamic network characteristic of an individual level, and realizes the space-time observation of the brain function network. When the relevant physiological mechanism of the neuropsychiatric diseases is researched, the specific indexes output by the model can be used for auxiliary diagnosis of the diseases, so that early diagnosis and early treatment of the relevant diseases are promoted.
Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
FIG. 1 is a flow chart illustrating the construction of individual brain function network pattern independent components according to the existing approach in accordance with an embodiment of the present invention;
FIG. 2 is a flowchart of a dynamic brain function network discrimination method based on a time domain-space domain joint state according to an embodiment of the present invention;
fig. 3 is a flowchart illustrating a feature search analysis according to an embodiment of the present invention.
Detailed Description
The invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
It is noted that the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of methods and systems according to various embodiments of the present disclosure. It should be noted that each block in the flowchart or block diagrams may represent a module, segment, or portion of code, the module, segment, or portion of code may comprise one or more executable instructions for implementing the logical functions specified in the various embodiments. It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
Example one
As shown in fig. 1, the present embodiment provides a dynamic brain function network analysis method based on a time domain-space domain joint state, and the present embodiment is illustrated by applying the method to a server, it is understood that the method may also be applied to a terminal, and may also be applied to a system including a terminal and a server, and is implemented by interaction between the terminal and the server. The server may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a network server, cloud communication, middleware service, a domain name service, a security service CDN, a big data and artificial intelligence platform, and the like. The terminal may be, but is not limited to, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and the like. The terminal and the server may be directly or indirectly connected through wired or wireless communication, and the application is not limited herein. In this embodiment, the method includes the steps of:
acquiring four-dimensional resting state functional magnetic resonance image data of a specific neuropsychiatric disease patient and a normal control group, and preprocessing the four-dimensional resting state functional magnetic resonance image data;
in the time dimension, intercepting a preprocessed four-dimensional resting state functional magnetic resonance image data segment according to a preset sliding window, and calculating a Pearson correlation coefficient of a time sequence signal between any two brain areas in the window segment to obtain a dynamic brain function connection matrix;
extracting the dynamic brain function connection matrix by using an independent component analysis method to obtain independent components corresponding to individuals and time sequences corresponding to the independent components;
based on the individual corresponding independent components and the time sequences corresponding to the independent components, a time domain-space domain fusion discrimination model is constructed by utilizing an efficient forward search strategy and a classifier, a specific brain dynamic network related to the specific neuropsychiatric disease is identified, and a decision value is output, so that the quantitative measurement of the dynamic brain functional network on the individual level is realized.
The specific scheme of the embodiment can be realized by adopting the following steps:
1. four-dimensional resting state functional magnetic resonance images are acquired for participants of different groups (patients with certain neuropsychiatric diseases and normal control groups), and the acquisition time is usually not less than 8 minutes.
FIG. 1 is a flow chart of the construction of individual brain function network pattern independent components according to existing approaches.
2. And preprocessing the acquired data. The preprocessing of the resting state functional magnetic resonance data mainly comprises time sequence correction, head movement correction, registration, space standardization, space smoothing, filtering and the like.
3. And dividing the preprocessed fMRI data into N brain areas according to a brain map template, and extracting the average time sequence of voxels contained in each brain area. The average time series is sliced into segments according to a preset sliding window size W (typically 60 seconds), and calculating Pearson correlation coefficient r of time sequence signals between any two brain regions in the window segment, and converting r value into z value by using Fisher' z transformation.
Assuming that T is the average time series length of the original data, a dynamic brain function connection matrix P is obtained after sliding window calculation, where P is a three-dimensional matrix of N × L (L = T/W) representing a time-series network of brain functions, where N × N is a function connection matrix in a window, and L is the number of layers of the time-series network.
4. All brain functional network pattern components of the subject were obtained by group Independent Component Analysis (ICA). Firstly, all individual dynamic brain function connection matrixes/brain function time sequence networks are connected in series in a time domain to obtain three-dimensional data with longer time; then, independent component analysis is used for the three-dimensional time sequence network data which are connected in series to generate group independent components; and finally mapping the group independent components to a single individual through an inverse reconstruction step to obtain the independent components corresponding to the individual and the corresponding time sequence thereof, wherein the independent components correspond to the spatial function connection state (spatial domain information) of the dynamic brain function network, and the time sequence corresponds to the time fluctuation (time domain information) of the function connection state. In order to explore a finer brain network pattern, the number of independent components may be set to 50 or more, or the number of components may be estimated from data.
FIG. 2 is a flow chart of a dynamic brain function network discrimination method based on a time domain-space domain combined state according to the present invention.
5. A time domain-space domain fusion discrimination model is constructed by utilizing an efficient forward selection search mechanism and a Support Vector Machine (SVM) model, and a specific brain dynamic network of a specific disease is identified. From the multivariate data representation perspective, the ICA can get the functional connectivity status as a basis function Zhang Chengzi space, which can characterize dynamic brain functional network pattern (dFNP).
Suppose independent component IC = { IC i I =1, … k }, dFNP can be defined as:
intra-spatial IC is characterized by brain network connectivity status, which is the Spatial State (SS) of the individual; IC in the time domain is characterized by the time series to which IC corresponds, which is the time domain state (TS) of an individual. IC can thus be defined as the joint time-space state:
ic i ＝{(q si ,q ti )|q si ∈R ss ,q ti ∈R ts }
wherein R is ss 、R ts Respectively representing spatial and temporal state spaces.
Because the time domain-space domain is a non-Euclidean space, the distance between dFNPs is defined in discriminant analysis in a space fusion or state fusion mode. Spatial fusion measures SS and TS separately in two riemann spaces, followed by nuclear fusion. The state fusion is to project the SS and the TS into a Riemann space for measurement in a mapping and merging mode. Thus, the fnp distance after fusion can be defined as:
wherein f, g and h are distance functions in different spaces respectively.
The invention maps SS into Grassmann manifold to measure geodesic distance, TS needs to further calculate time cooperativity matrix of IC (partial correlation matrix of any two IC time series), and maps it with Symmetrical Positive Definite (SPD) manifold to measure geodesic distance. Distance measurement matrixes among all individuals in different manifolds are respectively embedded into a common kernel function, such as a radial basis kernel function and a sigmoid kernel function, and are used as self-defined kernels in an SVM model to perform linear fusion, and fusion proportion coefficients can be selected through cross validation.
Based on the fused kernel function, an SVM classifier between the specific disease and the normal control group can be established. Since different brain networks (independent components) have different pattern expression capabilities, classifiers constructed based on different combinations of brain networks have different classification performances. Therefore, according to the efficient forward search strategy and the classifier performance thereof, the brain network combination with the most discriminating significance is selected from all independent components.
The efficient forward search strategy facilitates faster results with a large number of independent components and a high degree of computation, and fig. 3 is a flow chart of feature search analysis according to the present invention, the process of which is:
1) Assuming that there were M subjects in the dataset, ICA yielded corresponding k independent components. For each independent component ic i (i =1, … k), the set of M tested corresponding independent components
Classifier C can be built i Based on dFNP can be obtained through cross validation i ＝span(ic i ) The time domain-space domain combined state is established, and a Riemann flow kernel fusion-based SVM classifier and classification performance estimation (indexes such as the working characteristic curve area and the classification accuracy of a subject can be adopted to evaluateClassifier performance).
2) According to the estimated value of the classifier performance, the classifier C is paired according to the numerical value i Sorting in descending order to obtain sorted independent components ic i ' (i =1, … k), reflects the classification performance of each independent component on the data set in the time-space domain.
3) Sequentially mixing ic with i ' the first n (n =1, … k) ordered independent components are used as dFNP to construct the Riemann-based manifold-kernel fusion SVM classifier C n ', and estimate classification performance. From C n In the method, dFNP corresponding to the classifier with the best classification performance is selected as the dynamic function network mode with the most distinguishing significance. Thus, only 2k-1 classifiers are required to be established to obtain the specific brain network combination with the most discriminating significance.
6. And (3) constructing a time domain-airspace fusion discrimination model by utilizing the identified specific brain function network combination, and outputting a posterior probability value after classification to realize quantitative measurement of the dynamic brain function network state. And (3) carrying out correlation analysis on the quantified characteristic value and a clinically widely applied behavioural scale, discussing whether the intrinsic brain network is correlated with the apparent behavioural scale, and further analyzing the significance and the influence of the intrinsic brain network.
The manifold space described above takes the grassmann and SPD manifolds as examples, and the discriminant model takes the SVM as an example to describe the method of the present invention, but the present invention may also construct other machine learning models for dynamic brain function network analysis in other manifolds.
Example two
The embodiment provides a dynamic brain function network analysis system based on a time domain-space domain combined state.
A dynamic brain function network analysis system based on a time domain-space domain combined state comprises:
a data acquisition and pre-processing module configured to: acquiring four-dimensional resting state functional magnetic resonance image data of a specific neuropsychiatric disease patient and a normal control group, and preprocessing the four-dimensional resting state functional magnetic resonance image data;
a dynamic brain function connection matrix construction module configured to: in the time dimension, intercepting a preprocessed four-dimensional resting state functional magnetic resonance image data segment according to a preset sliding window, and calculating a Pearson correlation coefficient of a time sequence signal between any two brain areas in the window segment to obtain a dynamic brain function connection matrix;
a time-space joint state module configured to: extracting the dynamic brain function connection matrix by using an independent component analysis method to obtain independent components corresponding to individuals and time sequences corresponding to the independent components;
an analysis module configured to: based on the individual corresponding independent components and the time sequences corresponding to the independent components, a time domain-space domain fusion discrimination model is constructed by utilizing an efficient forward search strategy and a classifier, a specific brain dynamic network related to the specific neuropsychiatric disease is identified, and a decision value is output, so that the quantitative measurement of the dynamic brain functional network on the individual level is realized.
It should be noted here that the data acquisition and preprocessing module, the dynamic brain function connection matrix construction module, the time domain-space domain joint state module and the analysis module are the same as the example and application scenario realized by the steps in the first embodiment, but are not limited to the disclosure in the first embodiment. It should be noted that the modules described above as part of a system may be implemented in a computer system such as a set of computer executable instructions.
EXAMPLE III
The present embodiment provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps in the dynamic brain function network analysis method based on a temporal-spatial domain joint state as described in the first embodiment.
Example four
The embodiment provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the program to implement the steps in the time-space domain joint state-based dynamic brain function network analysis method according to the embodiment.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. The dynamic brain function network analysis method based on the time domain-space domain combined state is characterized by comprising the following steps of:
acquiring four-dimensional resting state functional magnetic resonance image data of a specific neuropsychiatric disease patient and a normal control group, and preprocessing the four-dimensional resting state functional magnetic resonance image data;
in the time dimension, intercepting a preprocessed four-dimensional resting state functional magnetic resonance image data segment according to a preset sliding window, and calculating a Pearson correlation coefficient of a time sequence signal between any two brain areas in the window segment to obtain a dynamic brain function connection matrix;
extracting the dynamic brain function connection matrix by using an independent component analysis method to obtain independent components corresponding to individuals and time sequences corresponding to the independent components;
based on the individual corresponding independent components and the time sequences corresponding to the independent components, a time domain-space domain fusion discrimination model is constructed by utilizing an efficient forward search strategy and a classifier, a specific brain dynamic network related to the specific neuropsychiatric disease is identified, and a decision value is output, so that the quantitative measurement of the dynamic brain functional network on the individual level is realized.
2. The time-space domain joint state based dynamic brain function network analysis method according to claim 1, wherein said preprocessing comprises: timing correction, head motion correction, registration, spatial normalization, spatial smoothing, and filtering.
3. The dynamic brain function network analysis method based on the time domain-space domain combined state as claimed in claim 1, wherein the step of intercepting the preprocessed four-dimensional resting state function magnetic resonance image data segments according to a preset sliding window in the time dimension specifically comprises:
dividing the preprocessed four-dimensional resting state functional magnetic resonance image data into a plurality of brain areas according to a brain map template, extracting an average time sequence of voxels contained in each brain area, and segmenting the average time sequence into segments according to a preset sliding window.
4. The dynamic brain function network analysis method based on time domain-space domain combined state according to claim 1, wherein the extracting the dynamic brain function connection matrix by using the independent component analysis method to obtain the individual corresponding independent component and the time sequence corresponding to the independent component specifically comprises:
all the individual dynamic brain function connection matrixes are connected in series on a time domain to obtain three-dimensional time sequence network data;
generating group independent components by using an independent component analysis method for the three-dimensional time series network data;
and mapping the group independent components to a single individual through an inverse reconstruction step so as to obtain the independent components corresponding to the individual and the time sequence corresponding to the independent components, wherein the independent components correspond to the spatial function connection state of the dynamic brain function network, and the time sequence corresponds to the time fluctuation of the spatial function connection state.
5. The dynamic brain function network analysis method based on time-space domain combined state according to claim 1, wherein the process of using the classifier specifically comprises:
from the view of multivariate data representation, independent components corresponding to individuals and time sequences corresponding to the independent components are obtained by an independent component analysis method and are established into a time domain-space domain combined state as a basis function Zhang Chengzi space, and the subspace represents a dynamic brain function network mode dFNP;
suppose independent component IC = { IC i I =1,. K }, then dFNP is defined as:
IC is defined as the joint time-space state:
ic i ＝{(q si ，q ti )|q si ∈R ss ，q ti ∈R ts }
wherein R is ss 、R ts Respectively representing space domain and time domain state spaces;
defining the distance between dFNP in discriminant analysis and adopting a space fusion or state fusion mode; measuring the space domain state and the time domain state of the individuals in two Riemann popular spaces respectively by space fusion, and then carrying out kernel fusion; the state fusion is to project the individual space domain state and the individual time domain state into a Riemann popular space for measurement in a mapping and merging mode; thus, the dFNP distance after fusion is defined as:
wherein f, g and h are distance functions in different spaces respectively;
mapping the individual space domain state into a Grassmann manifold to measure the geodesic distance, wherein the individual time domain state needs to further calculate a time cooperativity matrix of the IC, and mapping the time cooperativity matrix and the symmetrical positive definite manifold to measure the geodesic distance; respectively embedding distance measurement matrixes among all individuals in different manifolds into a kernel function, and performing linear fusion as a custom kernel in a classifier;
based on the fused kernel function, a classifier between the patients with the specific neuropsychiatric disease and the normal control group is established.
6. The time-space domain joint state-based dynamic brain function network analysis method according to claim 5, wherein the efficient forward search strategy specifically comprises:
1) Assuming that there are M tested data sets, the independent component analysis method yields corresponding k independent components, for each independent component ic i (i = 1.. K), set of independent components corresponding to M subjects
Establishing a classifier C i Is based on dFNP through cross validation i ＝span(ic i ) The time domain-space domain combined state is obtained, and a Riemann flow kernel fusion classifier and estimation classification performance are constructed;
2) According to the estimated value of the classifier performance, the classifier C is paired according to the numerical value i Sorting in descending order to obtain sorted independent components ic i ‘(i＝1，...k)；
3) Sequentially mixing ic with i The first n (n =1,. K) ordered independent components of' are used as dynamic brain function modes to construct a SVM (support vector machine) classifier C based on Riemann flow-form nuclear fusion n ', and estimating classification performance; from C n And selecting the dynamic brain function mode corresponding to the classifier with the best classification performance as the dynamic brain function network mode with the most distinguishing significance.
7. The time-space domain joint state-based dynamic brain function network analyzing method according to claim 1, wherein the outputting the decision value to realize the quantitative measurement of the dynamic brain function network at an individual level specifically comprises: and constructing a time domain-space domain fusion discrimination model by using the identified specific brain function network combination, and outputting a classification posterior probability value to realize quantitative measurement of the dynamic brain function network state.
8. A dynamic brain function network analysis system based on a time domain-space domain combined state is characterized by comprising:
a data acquisition and pre-processing module configured to: acquiring four-dimensional resting state functional magnetic resonance image data of a specific neuropsychiatric disease patient and a normal control group, and preprocessing the four-dimensional resting state functional magnetic resonance image data;
a dynamic brain function connection matrix construction module configured to: in the time dimension, intercepting a preprocessed four-dimensional resting state functional magnetic resonance image data segment according to a preset sliding window, and calculating a Pearson correlation coefficient of a time sequence signal between any two brain areas in the window segment to obtain a dynamic brain function connection matrix;
a time-space joint state module configured to: extracting the dynamic brain function connection matrix by using an independent component analysis method to obtain independent components corresponding to individuals and time sequences corresponding to the independent components;
an analysis module configured to: based on the individual corresponding independent components and the time sequences corresponding to the independent components, a time domain-space domain fusion discrimination model is constructed by utilizing an efficient forward search strategy and a classifier, a specific brain dynamic network related to the specific neuropsychiatric disease is identified, and a decision value is output, so that the quantitative measurement of the dynamic brain functional network on the individual level is realized.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method for dynamic brain function network analysis based on spatio-temporal union states as defined in any one of claims 1 to 7.
10. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method for dynamic brain function network analysis based on spatio-temporal union states as claimed in any one of claims 1 to 7 when executing the program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211078611.9A CN115272295A (en) | 2022-09-05 | 2022-09-05 | Dynamic brain function network analysis method and system based on time domain-space domain combined state |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211078611.9A CN115272295A (en) | 2022-09-05 | 2022-09-05 | Dynamic brain function network analysis method and system based on time domain-space domain combined state |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115272295A true CN115272295A (en) | 2022-11-01 |
Family
ID=83754100
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211078611.9A Pending CN115272295A (en) | 2022-09-05 | 2022-09-05 | Dynamic brain function network analysis method and system based on time domain-space domain combined state |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115272295A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116898401A (en) * | 2023-07-17 | 2023-10-20 | 燕山大学 | Autism spectrum disorder subtype classification method and device based on combined recursion quantification |
| CN117045228A (en) * | 2023-09-27 | 2023-11-14 | 燕山大学 | Dynamic brain network gender difference analysis method and device for autism spectrum disorder |
| CN117408322A (en) * | 2023-12-15 | 2024-01-16 | 北京邮电大学 | Brain function and brain structure mapping model training method, device and mapping method |
| CN117408322B (en) * | 2023-12-15 | 2024-04-16 | 北京邮电大学 | Brain function and brain structure mapping model training method, device and mapping method |
-
2022
- 2022-09-05 CN CN202211078611.9A patent/CN115272295A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116898401A (en) * | 2023-07-17 | 2023-10-20 | 燕山大学 | Autism spectrum disorder subtype classification method and device based on combined recursion quantification |
| CN116898401B (en) * | 2023-07-17 | 2024-02-02 | 燕山大学 | Autism spectrum disorder subtype classification method and device based on combined recursion quantification |
| CN117045228A (en) * | 2023-09-27 | 2023-11-14 | 燕山大学 | Dynamic brain network gender difference analysis method and device for autism spectrum disorder |
| CN117408322A (en) * | 2023-12-15 | 2024-01-16 | 北京邮电大学 | Brain function and brain structure mapping model training method, device and mapping method |
| CN117408322B (en) * | 2023-12-15 | 2024-04-16 | 北京邮电大学 | Brain function and brain structure mapping model training method, device and mapping method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Single slice based detection for Alzheimer’s disease via wavelet entropy and multilayer perceptron trained by biogeography-based optimization | |
| Khodatars et al. | Deep learning for neuroimaging-based diagnosis and rehabilitation of autism spectrum disorder: a review | |
| US20220122250A1 (en) | Brain feature prediction using geometric deep learning on graph representations of medical image data | |
| KR102125127B1 (en) | Method of brain disorder diagnosis via deep learning | |
| WO2023077603A1 (en) | Prediction system, method and apparatus for abnormal brain connectivity, and readable storage medium | |
| CN110188836B (en) | Brain function network classification method based on variational self-encoder | |
| CN113616184B (en) | Brain network modeling and individual prediction method based on multi-mode magnetic resonance image | |
| CN110236543B (en) | Alzheimer disease multi-classification diagnosis system based on deep learning | |
| CN109447183A (en) | Model training method, device, equipment and medium | |
| CN109344889B (en) | Brain disease classification apparatus, user terminal, and computer-readable storage medium | |
| US11693071B2 (en) | Systems and methods for mapping neuronal circuitry and clinical applications thereof | |
| CN115272295A (en) | Dynamic brain function network analysis method and system based on time domain-space domain combined state | |
| CN112735570B (en) | Image-driven brain atlas construction method, device, equipment and storage medium | |
| CN111863244B (en) | Functional connection mental disease classification method and system based on sparse pooling graph convolution | |
| CN116503680B (en) | Brain image structured analysis and brain disease classification system based on brain atlas | |
| CN113688862B (en) | Brain image classification method based on semi-supervised federal learning and terminal equipment | |
| Liu et al. | An enhanced multi-modal brain graph network for classifying neuropsychiatric disorders | |
| Qiang et al. | A deep learning method for autism spectrum disorder identification based on interactions of hierarchical brain networks | |
| CN117218453B (en) | Incomplete multi-mode medical image learning method | |
| Mousavian et al. | Depression detection using atlas from fMRI images | |
| CN114266738A (en) | Longitudinal analysis method and system for mild brain injury magnetic resonance image data | |
| Fisher et al. | Blind Subgrouping of Task-based fMRI | |
| Alharthi et al. | Do it the transformer way: A comprehensive review of brain and vision transformers for autism spectrum disorder diagnosis and classification | |
| Ara et al. | Novel approach of brain tumor segmentation using convolutional neural network hybridized with water cycle algorithm | |
| Tang et al. | Learning Asynchronous Common and Individual Functional Brain Network for AD Diagnosis |
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 |