Information processing method and device for auxiliary analysis of brain dysfunction, equipment and medium
By using brain region-based modeling units, combined with threshold screening and deep learning models, the problem of insufficient information fusion in rs-fMRI data classification was solved, enabling accurate and interpretable analysis of brain dysfunction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HAINAN UNIV
- Filing Date
- 2026-05-07
- Publication Date
- 2026-06-23
AI Technical Summary
Existing technologies struggle to simultaneously integrate brain region features, connectivity relationships, and spatial prior information in rs-fMRI data classification, resulting in limited model representation of complex brain network structures, insufficient stability, and a lack of interpretability.
Using brain regions as the basic modeling unit, feature data is generated through preprocessing to determine the importance and relevance of brain regions. Graph structure encoding is constructed by combining threshold screening, and enhanced features are generated using a deep learning model for classification.
This achievement enables interpretable rs-fMRI data classification at the brain region level, improving the accuracy and stability of classification and providing a scientific basis for medical image analysis.
Smart Images
Figure CN122266729A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data analysis technology, and in particular to information processing methods, devices, equipment and media for auxiliary analysis of brain dysfunction. Background Technology
[0002] Functional magnetic resonance imaging (fMRI) can acquire brain oxygenation level signals over time without introducing exogenous intervention, thereby revealing the functional activity characteristics of different brain regions and their interactions. By analyzing rs-fMRI (resting-state fMRI) data, brain functional networks can be further constructed, providing important evidence for the auxiliary research and assessment of brain functional abnormalities such as Alzheimer's disease and mild cognitive impairment. Therefore, how to effectively model and accurately classify rs-fMRI data has always been an important research direction in the field of intelligent medical image analysis.
[0003] In existing technologies, automatic classification methods for rs-fMRI data typically use brain region time-series features or brain functional connectivity matrices as the analysis objects. One type of method focuses on modeling the time series of individual brain regions, which can characterize local functional changes but struggles to fully reflect the collaborative relationships between brain regions. Another type of method introduces functional connectivity networks to describe the inter-brain region association structures, but often treats connectivity information as a holistic feature, ignoring the differences in the relative importance of different brain regions within the network. Furthermore, while some deep learning-based models have achieved certain improvements in classification performance, their modeling process is mostly focused on global feature representation, lacking a clear characterization of the contributions of brain region-level features, making it difficult to provide interpretable results for medical analysis.
[0004] Furthermore, brain functional networks not only contain information on functional connectivity strength, but also the topological properties of brain regions within the network and the spatial structural relationships between brain regions. Existing methods, when modeling rs-fMRI data, often fail to simultaneously integrate brain region features, connectivity relationships, and spatial prior information within a unified framework. This results in limited model representation of complex brain network structures and insufficient stability under conditions of sparse connectivity or noise interference.
[0005] Therefore, there is an urgent need for an rs-fMRI data classification method that can use brain regions as basic modeling units and output interpretable results at the brain region level while incorporating brain network structure information, so as to meet the dual requirements of accuracy and interpretability in medical image analysis. Summary of the Invention
[0006] In view of this, the purpose of this invention is to provide an information processing method, apparatus, device, and medium for auxiliary analysis of brain dysfunction, capable of using brain regions as basic modeling units and outputting interpretable rs-fMRI data classification results at the brain region level while incorporating brain network structure information, thereby meeting the dual requirements of accuracy and interpretability in medical image analysis. The specific solution is as follows: In a first aspect, this application provides an information processing method for auxiliary analysis of brain dysfunction, applied to a computer device, comprising: Initial functional magnetic resonance (fMRI) data is acquired and preprocessed to obtain target fMRI data; wherein, the fMRI data is time series data. Based on the target functional magnetic resonance imaging data, feature data corresponding to each brain region is generated, and the target importance corresponding to each brain region and the target correlation between each brain region are determined based on the feature data. Based on the comparison results of the relevance of each target with the preset threshold, the target spatial code corresponding to each brain region is determined, and the corresponding target graph structure code is constructed based on the target importance, target relevance and target spatial code of each brain region; Based on the target map structure encoding and the feature data, corresponding enhancement features are generated, and the classification results corresponding to the target functional magnetic resonance imaging data are determined based on the enhancement features, so as to generate and output the corresponding brain dysfunction analysis results using the classification results.
[0007] Optionally, the step of acquiring initial functional magnetic resonance imaging (fMRI) data and preprocessing the initial fMRI data to obtain target fMRI data includes: The initial functional magnetic resonance data were acquired using a target functional magnetic resonance imaging (fMRI) device under resting conditions or preset experimental conditions. The initial functional magnetic resonance imaging (fMRI) data is cropped, normalized, or standardized to obtain the target fMRI data.
[0008] Optionally, generating feature data corresponding to each brain region based on the target functional magnetic resonance imaging (fMRI) data includes: Based on a preset brain region partitioning template, the target functional magnetic resonance imaging (fMRI) data is mapped to each brain region to obtain the target fMRI data corresponding to each brain region. Based on the target functional magnetic resonance imaging data corresponding to each brain region, the feature data corresponding to each brain region is generated; the feature data corresponds one-to-one with the brain regions in the preset brain region partitioning template.
[0009] Optionally, determining the target importance corresponding to each brain region and the target correlation between each brain region based on the feature data includes: The relative importance of each brain region in the brain functional network is determined based on the feature data, and the target importance corresponding to each brain region is determined based on the relative importance of each brain region. The functional connectivity strength between the brain regions is determined based on the feature data, and the target correlation between the brain regions is determined based on the functional connectivity strength between the brain regions.
[0010] Optionally, the process of determining the target spatial code corresponding to each brain region based on the comparison results of the relevance of each target with a preset threshold, and constructing the corresponding target graph structure code based on the target importance, target relevance, and target spatial code of each brain region, includes: For any two brain regions, if the target correlation between the two brain regions is greater than or equal to the preset threshold, then the target spatial encoding corresponding to the two brain regions is determined, and the target graph structure encoding corresponding to the two brain regions is constructed based on the target importance, the target correlation and the target spatial encoding corresponding to the two brain regions.
[0011] Optionally, the process of determining the target spatial code corresponding to each brain region based on the comparison results of the relevance of each target with a preset threshold, and constructing the corresponding target graph structure code based on the target importance, the target relevance, and the target spatial code of each brain region, further includes: For any two brain regions, if the target correlation between the two brain regions is less than the preset threshold, then the target graph structure encoding corresponding to the two brain regions is directly constructed based on the target importance and target correlation of the two brain regions.
[0012] Optionally, the step of generating corresponding enhanced features based on the structure encoding of each target map and the feature data, and determining the classification result corresponding to the target functional magnetic resonance data based on the enhanced features, includes: A preset deep learning model is used to model the structure encoding of each target graph and the feature data to generate the enhanced features, and the classification result is generated based on the enhanced features; the classification result is a preset normal state, a preset cognitive impairment state, or a preset Alzheimer's disease state. Accordingly, the information processing method for auxiliary analysis of brain dysfunction also includes: Based on the target importance of each brain region, the target correlation between each brain region, and the classification results, corresponding visualization results are generated; the visualization results are used to show the relative importance of each brain region in the brain functional network, the correlation between each brain region, and the classification results.
[0013] Secondly, this application provides an information processing device for auxiliary analysis of brain dysfunction, applied to a computer device, comprising: The data acquisition module is used to acquire initial functional magnetic resonance (fMRI) data and preprocess the initial fMRI data to obtain target fMRI data; wherein the fMRI data is time series data. The correlation determination module is used to generate feature data corresponding to each brain region based on the target functional magnetic resonance data, and to determine the target importance corresponding to each brain region and the target correlation between each brain region based on the feature data. The graph structure encoding construction module is used to determine the target spatial encoding corresponding to each brain region based on the comparison results of the relevance of each target with a preset threshold, and to construct the corresponding target graph structure encoding based on the target importance, target relevance and target spatial encoding of each brain region; The classification result determination module is used to generate corresponding enhanced features based on the target map structure encoding and the feature data, and to determine the classification result corresponding to the target functional magnetic resonance data based on the enhanced features, so as to generate and output the corresponding brain dysfunction analysis result using the classification result.
[0014] Thirdly, this application provides an electronic device, comprising: Memory, used to store computer programs; A processor is used to execute the computer program to implement the aforementioned information processing method for auxiliary analysis of brain dysfunction.
[0015] Fourthly, this application provides a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned information processing method for auxiliary analysis of brain dysfunction.
[0016] In this application, initial functional magnetic resonance imaging (fMRI) data is first acquired and preprocessed to obtain target fMRI data; wherein, the fMRI data is time-series data; then, feature data corresponding to each brain region is generated based on the target fMRI data, and the target importance corresponding to each brain region and the target correlation between each brain region are determined based on each feature data; subsequently, based on the comparison results of each target correlation with a preset threshold, the target spatial code corresponding to each brain region is determined, and the corresponding target map structure code is constructed based on the target importance, target correlation and target spatial code of each brain region; finally, corresponding enhancement features are generated based on each target map structure code and each feature data, and the classification result corresponding to the target fMRI data is determined based on the enhancement features, so as to generate and output the corresponding brain dysfunction analysis result using the classification result. As can be seen from the above, this application first preprocesses the acquired time-series initial functional magnetic resonance imaging (fMRI) data to obtain standardized target fMRI data; then, based on the target fMRI data, it generates feature data corresponding to each brain region, determines the target importance of each brain region and the target correlation between brain regions based on the feature data, then determines the target spatial encoding of each brain region by combining the comparison results of each target correlation with the preset threshold, and integrates the target importance, target correlation and target spatial encoding of the brain region to construct the corresponding target map structure encoding; finally, it combines each target map structure encoding with each feature data to generate enhanced features, determines the classification result corresponding to the target fMRI data based on the enhanced features, and uses the classification result to generate and output the brain dysfunction analysis result. In this way, this application achieves a unified data foundation through standardized processing of functional magnetic resonance imaging (fMRI) data. By using brain regions as the core modeling unit to mine the importance of brain regions and the correlations between brain regions, and combining this with spatial coding introduced through threshold screening to construct a multi-source graph structure coding, it can fully integrate individual characteristics of brain regions, functional associations between brain regions, and spatial prior information. The enhanced features generated jointly by graph structure coding and brain region feature data not only enrich the feature dimensions and improve the discriminativeness and stability of the features, but also make the classification results based on the enhanced features more accurate. The classification results can provide a scientific and reliable basis for the auxiliary screening and assessment of brain function-related diseases, thereby generating analysis results of brain dysfunction. In this way, this application can achieve structured expression and feature enhancement of brain functional networks, improve the discriminative ability and stability of fMRI data classification, and provide medically meaningful and interpretable results for brain function analysis, meeting the dual requirements of accuracy and interpretability in medical image analysis. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0018] Figure 1 A flowchart of an information processing method for auxiliary analysis of brain dysfunction is provided in this application; Figure 2 This application provides a specific flowchart for generating graph structure encoding; Figure 3 This application provides a specific flowchart for generating visualization results; Figure 4 A specific functional magnetic resonance imaging (fMRI) data classification flowchart is provided for this application; Figure 5 This application provides a schematic diagram of an information processing device for auxiliary analysis of brain dysfunction. Figure 6 This application provides a structural diagram of an electronic device. Detailed Implementation
[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0020] Functional magnetic resonance imaging (fMRI) can acquire brain oxygenation level signals over time without introducing exogenous intervention, thereby revealing the functional activity characteristics and interactions of different brain regions. Analysis of rs-fMRI data can further construct brain functional networks, providing important evidence for the auxiliary research and assessment of brain functional abnormalities such as Alzheimer's disease and mild cognitive impairment. Therefore, how to effectively model and accurately classify rs-fMRI data has always been an important research direction in the field of intelligent medical image analysis. In existing technologies, automatic classification methods for rs-fMRI data typically use brain region time-series features or brain functional connectivity matrices as the analysis objects. One type of method focuses on modeling the time series of individual brain regions, which, while capable of depicting local functional changes, struggles to fully reflect the synergistic relationships between brain regions. Another type of method introduces functional connectivity networks to describe the inter-brain region association structures, but often treats connectivity information as a holistic feature, ignoring the relative importance differences of different brain regions within the network. Furthermore, while some deep learning-based models have achieved certain improvements in classification performance, their modeling process often focuses on global feature representation, lacking a clear characterization of the contributions of brain region-level features, making it difficult to provide interpretable results for medical analysis. Moreover, brain functional networks not only contain information on functional connectivity strength but also the topological attributes of brain regions within the network and the spatial structural relationships between brain regions. Existing methods, when modeling rs-fMRI data, often fail to simultaneously integrate brain region features, connectivity relationships, and spatial prior information within a unified framework, resulting in limited model representation of complex brain network structures and insufficient stability under conditions of sparse connectivity or noise interference. Therefore, this application provides an information processing scheme for auxiliary analysis of brain dysfunction, which can use brain regions as basic modeling units and output interpretable rs-fMRI data classification results at the brain region level while incorporating brain network structural information, thus meeting the dual requirements of accuracy and interpretability in medical image analysis.
[0021] See Figure 1 As shown, this embodiment of the invention discloses an information processing method for auxiliary analysis of brain dysfunction, applied to a computer device, and may include: Step S11: Obtain initial functional magnetic resonance (fMRI) data and preprocess the initial fMRI data to obtain target fMRI data; wherein, the fMRI data is time series data.
[0022] In this embodiment, initial functional magnetic resonance (fMRI) data is acquired and preprocessed to obtain target fMRI data. The specific process may include: firstly, acquiring the initial fMRI data in a resting state or under preset experimental conditions using a target fMRI imaging device; then, cropping, normalizing, or standardizing the initial fMRI data to obtain the target fMRI data.
[0023] Specifically, initial functional magnetic resonance (fMRI) data can be acquired by the target fMRI device in a resting state or under specific experimental conditions. The acquisition results typically represent a sequence of signal changes across multiple brain regions at consecutive time points, reflecting the functional activity characteristics of these brain regions over time. The acquired data serves as system input and enters the subsequent preprocessing process. This preprocessing reduces noise interference and standardizes the data scale. Specific operations may include cropping the initial fMRI data in brain region-level time-series format to obtain a uniform time length, and normalizing or standardizing the initial fMRI data to eliminate differences in acquisition duration and signal amplitude among different subjects. Through these preprocessing operations, a stable and consistent data foundation can be provided for subsequent feature construction and structural modeling.
[0024] Step S12: Generate feature data corresponding to each brain region based on the target functional magnetic resonance data, and determine the target importance corresponding to each brain region and the target correlation between each brain region based on the feature data.
[0025] In this embodiment, feature data corresponding to each brain region is generated based on the target functional magnetic resonance imaging (fMRI) data. The specific process may include: firstly, based on a preset brain region partitioning template, the target fMRI data is mapped to each brain region to obtain target fMRI data corresponding to each brain region; then, feature data corresponding to each brain region is generated based on the target fMRI data corresponding to each brain region; the feature data corresponds one-to-one with the brain regions in the preset brain region partitioning template.
[0026] Specifically, a standard brain region partitioning template with 116 brain regions is used to map the whole-brain target functional magnetic resonance imaging (fMRI) data to several brain regions with clear spatial locations and functional significance. Through mapping operations, the target fMRI data of each brain region is represented as an independent brain region-level feature unit, resulting in ROI-Token feature data (ROI, Region of Interest) corresponding to each brain region. This constructs a feature sequence composed of multiple ROI-Tokens, where each ROI-Token is used to characterize the functional activity features of each brain region in the time dimension.
[0027] It should be noted that determining the target importance corresponding to each brain region and the target correlation between each brain region based on each feature data includes: firstly, determining the relative importance of each brain region in the brain functional network based on each feature data, and determining the target importance corresponding to each brain region based on the relative importance of each brain region; then, determining the functional connectivity strength between each brain region based on each feature data, and determining the target correlation between each brain region based on the functional connectivity strength between each brain region.
[0028] Specifically, firstly, the importance-related representation of each brain region node is calculated based on the ROI-Token feature data of each brain region to obtain the target importance corresponding to each brain region, which is used to characterize the relative importance of different brain regions in the overall brain functional network; secondly, the functional connectivity strength between brain regions is calculated based on the statistical relationship between time series of different brain regions, thereby constructing the correlation encoding between brain regions to obtain the target correlation between brain regions, which is used to reflect the functional association between brain regions.
[0029] Step S13: Based on the comparison results of the relevance of each target with the preset threshold, determine the target spatial code corresponding to each brain region, and construct the corresponding target graph structure code based on the target importance, target relevance and target spatial code of each brain region.
[0030] In this embodiment, the process of determining the target spatial code corresponding to each brain region based on the comparison results of the relevance of each target with a preset threshold, and constructing the corresponding target graph structure code based on the target importance, target relevance, and target spatial code of each brain region, may include the following steps: For any two brain regions, if the target relevance between the two brain regions is greater than or equal to the preset threshold, then the target spatial code corresponding to the two brain regions is determined, and the target graph structure code corresponding to the two brain regions is constructed based on the target importance, target relevance, and target spatial code; if the target relevance between the two brain regions is less than the preset threshold, then the target graph structure code corresponding to the two brain regions is directly constructed based on the target importance and target relevance of the two brain regions. Specifically, see [link to documentation]. Figure 2 As shown, a preset correlation threshold condition is introduced based on the target correlation.
[0031] In one specific implementation, when the functional connectivity strength between two brain regions meets the threshold condition, that is, when the target relevance is greater than or equal to a preset threshold, prior information on the spatial distance between the brain regions is further introduced to spatially encode the two brain regions, thereby obtaining the corresponding target spatial code. Based on the target importance, target relevance, and target spatial code of the two brain regions, a target graph structure code corresponding to the two brain regions is constructed, ultimately forming a multi-source graph structure code that integrates brain region importance, functional connectivity, and spatial structure.
[0032] In another specific implementation, when the target relevance is less than a preset threshold, prior information about spatial distance between brain regions is not introduced. Instead, target graph structure encoding for the two brain regions is constructed solely based on their target importance and relevance, ultimately forming a multi-source graph structure encoding that integrates brain region importance and functional connectivity. Through these operations, the validity of structural information can be ensured while avoiding the introduction of spatial relationships into invalid or weakly correlated brain region pairs, thereby reducing noise interference and improving the stability of structural modeling.
[0033] Step S14: Generate corresponding enhanced features based on the target map structure encoding and the feature data, and determine the classification result corresponding to the target functional magnetic resonance data based on the enhanced features, so as to generate and output the corresponding brain dysfunction analysis result using the classification result.
[0034] In this embodiment, corresponding enhanced features are generated based on the structure encoding of each target graph and each feature data, and the classification result corresponding to the target functional magnetic resonance imaging data is determined based on the enhanced features. The specific process may include: using a preset deep learning model to model each of the target graph structure encodings and each of the feature data to generate the enhanced features, and generating the classification result based on the enhanced features; the classification result is a preset normal state, a preset cognitive impairment state, or a preset Alzheimer's disease state.
[0035] Specifically, after completing the multi-source structure encoding, the ROI-Token feature sequence and the encoding of each target graph structure are jointly input into the joint modeling module of the Graph-Explainable-ROI-Former classification model (GER-Former classification model), which is oriented towards brain region graph structures, for processing. The joint modeling module of the GER-Former classification model is used to fuse and model the ROI-Token feature sequence and the encoding of each target graph structure at the brain region level, thereby obtaining an enhanced feature representation that includes brain region functional features and brain region structural relationships. Through this joint modeling process, the expressive power of brain region features in the overall brain network structure can be effectively improved. Finally, classification results for the subjects are generated based on the enhanced features. The classification results are either a preset normal state, a preset cognitive impairment state, or a preset Alzheimer's disease state. The classification results can be used for auxiliary screening or assessment of Alzheimer's disease, mild cognitive impairment, or other brain functional disorders.
[0036] It should be noted that this embodiment can also generate corresponding visualization results based on the target importance of each brain region, the target relevance between brain regions, and the classification results. The visualization results are used to display the relative importance of each brain region in the brain functional network, the correlation between brain regions, and the classification results. Specifically, while outputting the classification results, further analysis of brain region-level feature response information is also performed: see [link to documentation]. Figure 3 As shown, by combining the classification results with the importance ranking information of brain region ROIs, ROI feature data corresponding to the categories of the classification results can be generated. Under the constraints of preset template rules, category feature visualization results can be constructed. This not only obtains the classification results, but also intuitively shows the degree of contribution of different brain regions in the classification decision-making process, thus providing an interpretable auxiliary basis for brain function analysis.
[0037] It should be noted that, see Figure 4 As shown, this embodiment can be used to process rs-fMRI data from different categories, such as NC (Normal Controls) data, MCI (Mild Cognitive Impairment) data, and AD (Alzheimer's Disease) data. The raw rs-fMRI data from different categories undergo uniform preprocessing under the constraints of predefined brain region templates to form a standardized functional magnetic resonance imaging dataset, which is then further input into the GER-Former classification model for analysis. The GER-Former classification model outputs corresponding classification results based on enhanced feature representations, used to characterize the subject's brain functional state category.
[0038] As can be seen from the above, in this embodiment, initial functional magnetic resonance imaging (fMRI) data is first acquired and preprocessed to obtain target fMRI data; wherein, the fMRI data is time series data; then, feature data corresponding to each brain region is generated based on the target fMRI data, and the target importance corresponding to each brain region and the target correlation between each brain region are determined based on each feature data; subsequently, based on the comparison results of each target correlation with a preset threshold, the target spatial code corresponding to each brain region is determined, and the corresponding target graph structure code is constructed based on the target importance, target correlation and target spatial code of each brain region; finally, corresponding enhancement features are generated based on each target graph structure code and each feature data, and the classification result corresponding to the target fMRI data is determined based on the enhancement features, so as to generate and output the corresponding brain dysfunction analysis result using the classification result. As can be seen from the above, in this embodiment, the initial functional magnetic resonance imaging (fMRI) data in the form of time series is first preprocessed to obtain standardized target fMRI data; then, feature data corresponding to each brain region is generated based on the target fMRI data, and the target importance and target correlation between brain regions are determined according to the feature data. Next, the target spatial encoding of each brain region is determined by combining the comparison results of each target correlation with the preset threshold, and the target importance, target correlation and target spatial encoding of the brain region are fused to construct the corresponding target map structure encoding; finally, each target map structure encoding is combined with each feature data to generate enhanced features, the classification result corresponding to the target fMRI data is determined according to the enhanced features, and the brain dysfunction analysis result is generated and output using the classification result. In this embodiment, the standardized processing of functional magnetic resonance imaging (fMRI) data achieves a unified data foundation. By using brain regions as the core modeling unit to mine the importance of brain regions and the correlations between brain regions, and combining this with spatial coding introduced through threshold screening to construct a multi-source graph structure coding, the individual characteristics of brain regions, functional associations between brain regions, and spatial prior information are fully integrated. The enhanced features generated jointly by the graph structure coding and brain region feature data not only enrich the feature dimensions and improve the discriminativeness and stability of the features, but also make the classification results based on the enhanced features more accurate. The classification results can provide a scientific and reliable basis for the auxiliary screening and assessment of brain function-related diseases, thereby generating analysis results of brain dysfunction. Thus, this application can achieve structured expression and feature enhancement of brain functional networks, improve the discriminative ability and stability of fMRI data classification, and provide medically meaningful interpretable results for brain function analysis, meeting the dual requirements of accuracy and interpretability in medical image analysis.
[0039] Accordingly, see Figure 5 As shown in the illustration, this application also provides an information processing device for auxiliary analysis of brain dysfunction, applied to a computer device, and may include: The data acquisition module 11 is used to acquire initial functional magnetic resonance (fMRI) data and preprocess the initial fMRI data to obtain target fMRI data; wherein the fMRI data is time series data. The correlation determination module 12 is used to generate feature data corresponding to each brain region based on the target functional magnetic resonance data, and to determine the target importance corresponding to each brain region and the target correlation between each brain region based on the feature data. The graph structure encoding construction module 13 is used to determine the target spatial encoding corresponding to each brain region based on the comparison results of the relevance of each target with a preset threshold, and to construct the corresponding target graph structure encoding based on the target importance, the target relevance and the target spatial encoding of each brain region; The classification result determination module 14 is used to generate corresponding enhanced features based on the target map structure encoding and the feature data, and to determine the classification result corresponding to the target functional magnetic resonance data based on the enhanced features, so as to generate and output the corresponding brain dysfunction analysis result using the classification result.
[0040] In some specific embodiments, the data acquisition module 11 may include: The data acquisition unit is used to acquire the initial functional magnetic resonance data using the target functional magnetic resonance imaging device in a resting state or under preset experimental conditions. The data trimming unit is used to trim, normalize, or standardize the initial functional magnetic resonance data to obtain the target functional magnetic resonance data.
[0041] In some specific embodiments, the correlation determination module 12 may include: The data mapping unit is used to map the target functional magnetic resonance imaging (fMRI) data to each brain region based on a preset brain region partitioning template, so as to obtain the target fMRI data corresponding to each brain region. The feature data generation unit is used to generate feature data corresponding to each brain region based on the target functional magnetic resonance imaging data corresponding to each brain region; the feature data corresponds one-to-one with the brain regions in the preset brain region partitioning template.
[0042] In some specific embodiments, the correlation determination module 12 may include: An importance determination unit is used to determine the relative importance of each brain region in the brain functional network based on each of the feature data, and to determine the target importance corresponding to each brain region based on the relative importance of each brain region; The correlation determination unit is used to determine the functional connectivity strength between the brain regions based on the feature data, and to determine the target correlation between the brain regions based on the functional connectivity strength between the brain regions.
[0043] In some specific embodiments, the graph structure encoding construction module 13 may include: The first graph structure encoding construction unit is used to determine the target spatial encoding corresponding to the two brain regions if the target correlation between the two brain regions is greater than or equal to the preset threshold, and construct the target graph structure encoding corresponding to the two brain regions based on the target importance, the target correlation and the target spatial encoding corresponding to the two brain regions.
[0044] In some specific embodiments, the graph structure encoding construction module 13 may further include: The second graph structure encoding construction unit is used to construct the target graph structure encoding corresponding to the two brain regions directly based on the target importance and target correlation corresponding to the two brain regions if the target correlation between the two brain regions is less than the preset threshold.
[0045] In some specific embodiments, the classification result determination module 14 may include: The classification result determination unit is used to model the structure encoding of each target graph and the feature data using a preset deep learning model to generate the enhanced features, and generate the classification result based on the enhanced features; the classification result is a preset normal state, a preset cognitive impairment state, or a preset Alzheimer's disease state. Accordingly, the information processing device for auxiliary analysis of brain dysfunction may further include: The results visualization module is used to generate corresponding visualization results based on the target importance of each brain region, the target correlation between each brain region, and the classification results; the visualization results are used to show the relative importance of each brain region in the brain functional network, the correlation between each brain region, and the classification results.
[0046] Furthermore, embodiments of this application also disclose an electronic device, Figure 6This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content of the diagram should not be construed as limiting the scope of this application. Specifically, the electronic device 20 may include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 stores a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the information processing method for assistive analysis of brain dysfunction disclosed in any of the foregoing embodiments. Furthermore, the electronic device 20 in this embodiment may specifically be an electronic computer.
[0047] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 25 is used to acquire external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.
[0048] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored thereon can include operating system 221, computer program 222, etc., and the storage method can be temporary storage or permanent storage.
[0049] The operating system 221 is used to manage and control the various hardware devices on the electronic device 20 and the computer program 222, which may be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of performing the information processing method for assistive analysis of brain dysfunction disclosed in any of the foregoing embodiments, the computer program 222 may further include computer programs capable of performing other specific tasks.
[0050] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned information processing method for auxiliary analysis of brain dysfunction. Specific steps of this method can be found in the corresponding content disclosed in the foregoing embodiments, and will not be repeated here.
[0051] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.
[0052] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0053] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.
[0054] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0055] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.
Claims
1. An information processing method for auxiliary analysis of brain dysfunction, characterized in that, Applied to computer devices, including: Initial functional magnetic resonance (fMRI) data is acquired and preprocessed to obtain target fMRI data; wherein, the fMRI data is time series data. Based on the target functional magnetic resonance imaging data, feature data corresponding to each brain region is generated, and the target importance corresponding to each brain region and the target correlation between each brain region are determined based on the feature data. Based on the comparison results of the relevance of each target with the preset threshold, the target spatial code corresponding to each brain region is determined, and the corresponding target graph structure code is constructed based on the target importance, target relevance and target spatial code of each brain region; Based on the target map structure encoding and the feature data, corresponding enhancement features are generated, and the classification results corresponding to the target functional magnetic resonance imaging data are determined based on the enhancement features, so as to generate and output the corresponding brain dysfunction analysis results using the classification results.
2. The information processing method for auxiliary analysis of brain dysfunction according to claim 1, characterized in that, The process of acquiring initial functional magnetic resonance imaging (fMRI) data and preprocessing the initial fMRI data to obtain target fMRI data includes: The initial functional magnetic resonance data were acquired using a target functional magnetic resonance imaging (fMRI) device under resting conditions or preset experimental conditions. The initial functional magnetic resonance imaging (fMRI) data is cropped, normalized, or standardized to obtain the target fMRI data.
3. The information processing method for auxiliary analysis of brain dysfunction according to claim 1, characterized in that, The generation of feature data corresponding to each brain region based on the target functional magnetic resonance imaging data includes: Based on a preset brain region partitioning template, the target functional magnetic resonance imaging (fMRI) data is mapped to each brain region to obtain the target fMRI data corresponding to each brain region. Based on the target functional magnetic resonance imaging data corresponding to each brain region, the feature data corresponding to each brain region is generated; the feature data corresponds one-to-one with the brain regions in the preset brain region partitioning template.
4. The information processing method for auxiliary analysis of brain dysfunction according to claim 1, characterized in that, The determination of the target importance corresponding to each brain region and the target correlation between each brain region based on the feature data includes: The relative importance of each brain region in the brain functional network is determined based on the feature data, and the target importance corresponding to each brain region is determined based on the relative importance of each brain region. The functional connectivity strength between the brain regions is determined based on the feature data, and the target correlation between the brain regions is determined based on the functional connectivity strength between the brain regions.
5. The information processing method for auxiliary analysis of brain dysfunction according to claim 1, characterized in that, The process of determining the target spatial code corresponding to each brain region based on the comparison results of the relevance of each target with a preset threshold, and constructing the corresponding target graph structure code based on the target importance, target relevance, and target spatial code of each brain region, includes: For any two brain regions, if the target correlation between the two brain regions is greater than or equal to the preset threshold, then the target spatial encoding corresponding to the two brain regions is determined, and the target graph structure encoding corresponding to the two brain regions is constructed based on the target importance, the target correlation and the target spatial encoding corresponding to the two brain regions.
6. The information processing method for auxiliary analysis of brain dysfunction according to claim 5, characterized in that, The process of determining the target spatial code corresponding to each brain region based on the comparison results of the relevance of each target with a preset threshold, and constructing the corresponding target graph structure code based on the target importance, target relevance, and target spatial code of each brain region, further includes: For any two brain regions, if the target correlation between the two brain regions is less than the preset threshold, then the target graph structure encoding corresponding to the two brain regions is directly constructed based on the target importance and target correlation of the two brain regions.
7. The information processing method for auxiliary analysis of brain dysfunction according to any one of claims 1 to 6, characterized in that, The step of generating corresponding enhanced features based on the target map structure encoding and the feature data, and determining the classification result corresponding to the target functional magnetic resonance data based on the enhanced features, includes: A preset deep learning model is used to model the structure encoding of each target graph and the feature data to generate the enhanced features, and the classification result is generated based on the enhanced features; the classification result is a preset normal state, a preset cognitive impairment state, or a preset Alzheimer's disease state. Accordingly, the information processing method for auxiliary analysis of brain dysfunction also includes: Based on the target importance of each brain region, the target correlation between each brain region, and the classification results, corresponding visualization results are generated; the visualization results are used to show the relative importance of each brain region in the brain functional network, the correlation between each brain region, and the classification results.
8. An information processing device for auxiliary analysis of brain dysfunction, characterized in that, Applied to computer devices, including: The data acquisition module is used to acquire initial functional magnetic resonance (fMRI) data and preprocess the initial fMRI data to obtain target fMRI data; wherein the fMRI data is time series data. The correlation determination module is used to generate feature data corresponding to each brain region based on the target functional magnetic resonance data, and to determine the target importance corresponding to each brain region and the target correlation between each brain region based on the feature data. The graph structure encoding construction module is used to determine the target spatial encoding corresponding to each brain region based on the comparison results of the relevance of each target with a preset threshold, and to construct the corresponding target graph structure encoding based on the target importance, target relevance and target spatial encoding of each brain region; The classification result determination module is used to generate corresponding enhanced features based on the target map structure encoding and the feature data, and to determine the classification result corresponding to the target functional magnetic resonance data based on the enhanced features, so as to generate and output the corresponding brain dysfunction analysis result using the classification result.
9. An electronic device, characterized in that, The electronic device includes a processor and a memory; wherein the memory is used to store a computer program, which is loaded and executed by the processor to implement the information processing method for auxiliary analysis of brain dysfunction as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, Used to store a computer program, which, when executed by a processor, implements the information processing method for auxiliary analysis of brain dysfunction as described in any one of claims 1 to 7.