A personalized target prediction method based on twin brain functional dynamics model

Abstract

The present application provides an individualized target prediction method based on a twin brain function dynamics model, the method comprising: constructing a sensitivity evaluation model based on sensitivity evaluation indicators, interaction relationships, and interaction networks; extracting action mechanism characteristic elements based on the sensitivity evaluation model, constructing a brain function network dynamics model using the action mechanism characteristic elements, and incrementally updating the brain function network dynamics model based on the action mechanism characteristic elements to obtain an incrementally updated brain function network dynamics model; constructing a twin brain function dynamics model based on the incrementally updated brain function network dynamics model, and generating a semantic description of the action mechanism formation process based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model. Through the method, the accuracy of individualized target calculation can be improved.

Description

Technical Field

[0001] The present application relates to the field of target prediction technology, and in particular to an individualized target prediction method based on a twin brain functional dynamics model. Background Art

[0002] With the development of medical technology, brain imaging data can be used to study and understand brain structure and function, as well as the occurrence and development mechanism of brain diseases. Transcranial magnetic stimulation, which relies on brain imaging data for target positioning, has gained increasing recognition in the fields of clinical psychiatry, neurological diseases and rehabilitation. However, the differences in brain structure and functional connections among patients make individual therapeutic effects highly heterogeneous, and there is a lag in the analysis of brain state changes, which makes it impossible to correct the stimulation target in a timely manner when the individual cause of the disease is unclear.

[0003] The effects of transcranial magnetic stimulation are not limited to the brain areas that are directly intervened, but may also be transmitted to other brain areas that have structural or functional connections with the intervened brain areas. Traditional manual diagnosis can only observe brain structure through low-level features such as brightness, size, and shape of the examination area in neuroimaging, making it impossible to correctly observe the high-level brain functions representing abnormal brain areas hidden behind the brain structure; at the same time, due to the high heterogeneity of patients' conditions, there is a lag in the analysis of individual differences in brain functional connectivity and brain state changes, which may lead to heterogeneity in the efficacy of transcranial magnetic stimulation and make the calculation of dynamic targets of transcranial magnetic stimulation more difficult. Summary of the invention

[0004] In view of this, the purpose of this application is to provide an individualized target prediction method based on a twin brain function dynamics model. The user's individualized target is determined by their respective functional neuroimaging data. The action mechanism and characteristic parameter identification method of brain functional connections based on complex brain information are used to capture the time-varying characteristics of the brain network. The mechanism-feature-guided brain function network dynamics incremental update method is adopted to achieve compensation and correction of the dynamics model, and a brain function twin model is constructed to accurately calculate the individualized target, thereby improving the accuracy of individualized target calculation.

[0005] In a first aspect, the present application embodiment provides an individualized target prediction method based on a twin brain functional dynamics model, the individualized target prediction method comprising:

[0006] Analyze the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generate an interaction network, extract sensitive feature parameters that affect the action mechanism based on the interaction relationship and the interaction network, and construct a sensitivity evaluation index based on the sensitive feature parameters;

[0007] Based on the sensitivity evaluation index, the interaction relationship and the interaction network, constructing a sensitivity evaluation model;

[0008] Extracting characteristic elements of the action mechanism based on the sensitivity evaluation model, constructing a brain function network dynamics model using the characteristic elements of the action mechanism, and incrementally updating the brain function network dynamics model based on the characteristic elements of the action mechanism to obtain an incrementally updated brain function network dynamics model;

[0009] A twin brain function dynamics model is constructed based on the incrementally updated brain function network dynamics model, and a semantic description of the action mechanism formation process is generated based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model.

[0010] Furthermore, the method of analyzing the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generating an interaction network includes:

[0011] Modeling the functional connectivity relationship between different brain regions in the functional magnetic resonance imaging data to determine the interaction relationship between different brain regions;

[0012] The functional magnetic resonance imaging data are preprocessed and feature extracted to obtain characteristic parameters of each functional area of ​​the brain, and a brain network science method is used to construct an interaction network of multiple brain areas.

[0013] Furthermore, constructing a sensitivity evaluation model based on the sensitivity evaluation index, the interaction relationship and the interaction network includes:

[0014] Extracting characteristic parameters affecting the action mechanism according to the interaction relationship and the interaction network;

[0015] Based on the characteristic parameters and the sensitivity evaluation index, a sensitivity evaluation model integrating the variation characteristics of brain functional connectivity and the action characteristic parameters is constructed.

[0016] Furthermore, extracting characteristic elements of the action mechanism based on the sensitivity evaluation model and constructing a brain function network dynamics model using the characteristic elements of the action mechanism includes:

[0017] Extracting characteristic elements of the action mechanism under the complex brain network based on the sensitivity evaluation model;

[0018] Iteratively simulating and modeling the clinical characteristics and regions associated with each element to generate simulation data related to the characteristic elements of the mechanism of action;

[0019] The simulation data is used to calculate the correlation coefficients between the nodes of the brain function network, so as to construct the brain function network dynamics model using the correlation coefficients.

[0020] Furthermore, the incrementally updating the brain function network dynamics model based on the action mechanism characteristic elements to obtain the incrementally updated brain function network dynamics model includes:

[0021] According to the characteristic elements of the mechanism of action, the twin data selection result is determined by using characteristic signal enhancement;

[0022] Based on the twin data selection results, the characteristic elements of the action mechanism are analyzed from the perspective of multi-level spatiotemporal scales, and the dynamic changes of brain functional connections are processed in real time and fed back to the brain functional network dynamics model to achieve incremental updating of the brain functional network dynamics model and obtain the incrementally updated brain functional network dynamics model.

[0023] Furthermore, the twin brain function dynamics model is constructed based on the incrementally updated brain function network dynamics model, and a semantic description of the mechanism formation process is generated based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model, including:

[0024] Based on the incrementally updated brain function network dynamics model, constructing the twin brain function dynamics model by mathematical deduction;

[0025] Determining a semantic description of the process of forming the mechanism of action based on the characteristic elements of the mechanism of action and the incrementally updated brain functional network dynamics model;

[0026] Based on the semantic description, the progression pattern of neuronal activity under the twin brain functional dynamics model is determined to perform individualized target prediction based on the twin brain functional dynamics model.

[0027] In a second aspect, the present application also provides an individualized target prediction device based on the twin brain functional dynamics model, the individualized target prediction device comprising:

[0028] A sensitivity evaluation index construction module, used to analyze the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generate an interaction network, extract sensitive feature parameters that affect the action mechanism based on the interaction relationship and the interaction network, and construct a sensitivity evaluation index based on the sensitive feature parameters;

[0029] A sensitivity evaluation model construction module, used to construct a sensitivity evaluation model based on the sensitivity evaluation index, the interaction relationship and the interaction network;

[0030] An incremental update module, used for extracting characteristic elements of the action mechanism based on the sensitivity evaluation model, constructing a brain function network dynamics model using the characteristic elements of the action mechanism, and incrementally updating the brain function network dynamics model based on the characteristic elements of the action mechanism to obtain an incrementally updated brain function network dynamics model;

[0031] A prediction module is used to construct a twin brain function dynamics model based on the incrementally updated brain function network dynamics model, and to generate a semantic description of the action mechanism formation process based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model.

[0032] Furthermore, when the sensitivity evaluation index construction module is used to analyze the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generate an interaction network, the sensitivity evaluation index construction module is also used to:

[0033] Modeling the functional connectivity relationship between different brain regions in the functional magnetic resonance imaging data to determine the interaction relationship between different brain regions;

[0034] The functional magnetic resonance imaging data are preprocessed and feature extracted to obtain characteristic parameters of each functional area of ​​the brain, and a brain network science method is used to construct an interaction network of multiple brain areas.

[0035] In a third aspect, an embodiment of the present application further provides an electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, the steps of the individualized target prediction method based on the twin brain functional dynamics model are performed as described above.

[0036] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, executes the steps of the above-mentioned individualized target prediction method based on the twin brain functional dynamics model.

[0037] In this application, the user's individualized target is determined by their respective functional neuroimaging data. The action mechanism and characteristic parameter identification method of brain functional connections based on complex brain information are used to capture the time-varying characteristics of the brain network. The mechanism-feature-guided brain functional network dynamics incremental update method is adopted to achieve compensation and correction of the dynamic model, and a brain function twin model is constructed to accurately calculate the individualized target, thereby improving the accuracy of individualized target calculation.

[0038] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 A flowchart of an individualized target prediction method based on a twin brain functional dynamics model provided in an embodiment of the present application;

[0041] Figure 2 A schematic diagram of the structure of an individualized target prediction device based on a twin brain functional dynamics model provided in an embodiment of the present application;

[0042] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0043] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work belongs to the scope of protection of the present application.

[0044] First, the application scenarios to which the present application is applicable are introduced. The present application can be applied to the field of target prediction technology.

[0045] With the development of medical technology, brain imaging data can be used to study and understand brain structure and function, as well as the occurrence and development mechanism of brain diseases. Transcranial magnetic stimulation, which relies on brain imaging data for target positioning, has gained increasing recognition in the fields of clinical psychiatry, neurological diseases and rehabilitation. However, the differences in brain structure and functional connections among patients make individual therapeutic effects highly heterogeneous, and there is a lag in the analysis of brain state changes, which makes it impossible to correct the stimulation target in a timely manner when the individual cause of the disease is unclear.

[0046] Studies have found that the effects of transcranial magnetic stimulation are not limited to the directly intervened brain areas, but may also be transmitted to other brain areas that have structural or functional connections with the intervened brain areas. Traditional manual diagnosis can only observe brain structure through low-level features such as brightness, size, and shape of the examination area in neuroimaging, making it impossible to correctly observe the high-level brain functions representing abnormal brain areas hidden behind the brain structure; at the same time, due to the high heterogeneity of patients' conditions, there is a lag in the analysis of individual differences in brain functional connectivity and brain state changes, which may lead to heterogeneity in the efficacy of transcranial magnetic stimulation and make the calculation of dynamic targets of transcranial magnetic stimulation more difficult.

[0047] Based on this, an embodiment of the present application provides an individualized target prediction method based on the twin brain functional dynamics model to improve the accuracy of individualized target calculation.

[0048] See also Figure 1 , Figure 1 This is a flowchart of a personalized target prediction method based on a twin brain functional dynamics model provided in an embodiment of the present application. Figure 1 As shown in , the individualized target prediction method provided in the embodiment of the present application includes:

[0049] S101, analyzing the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generating an interaction network, extracting sensitive feature parameters that affect the action mechanism based on the interaction relationship and the interaction network, and constructing a sensitivity evaluation index based on the sensitive feature parameters.

[0050] First, it is necessary to analyze the dynamic characteristics of the brain functional network under complex brain information, the correlation relationship of time-varying characteristics, the coupling relationship and the synergistic relationship according to the user's brain functional magnetic resonance medical imaging data to construct the sensitivity evaluation index of the characteristic parameters. For the above step S101, in the specific implementation, firstly, based on the brain functional magnetic resonance medical imaging data, the interaction relationship of multiple information (disease principle, disease cause, disease state, etc.) and multiple brain region information (frontal lobe, parietal lobe, temporal lobe and occipital lobe and other brain regions) is analyzed to construct a multi-information, multi-region interaction network, and then based on the interaction relationship and the interaction network, the sensitive characteristic parameters affecting the mechanism of action are extracted, and the sensitivity evaluation index is constructed based on the sensitive characteristic parameters. Based on the above analysis, the formation elements in the formation process of the multi-information, multi-region mechanism of action and the correlation elements between the formation elements are extracted, and by analyzing the interaction network, the key elements affecting the formation of the multi-information, multi-region mechanism of action are extracted, including disease-related factors, characteristics of brain regions, etc. Analyze the correlation between the extracted formative factors, use statistical methods such as correlation analysis to identify the correlated factors, and construct an index system for evaluating the sensitivity of the elements of the brain functional connection mechanism under complex brain information, which covers indicators such as dynamic correlation and temporal variability.

[0051] As an optional implementation, with respect to the above step S101, analyzing the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generating an interaction network includes:

[0052] Step 1011 , modeling the functional connectivity relationship between different brain regions in the functional magnetic resonance imaging data to determine the interaction relationship between different brain regions.

[0053] Step 1012, preprocessing and feature extraction are performed on the functional magnetic resonance imaging data to obtain feature parameters of each functional area of ​​the brain, and using brain network science methods to construct an interaction network of multiple brain areas.

[0054] For the above steps 1011-1012, in the specific implementation, the functional connection relationship between different brain regions in the brain functional magnetic resonance imaging data is first modeled to determine the interaction relationship between different brain regions. Then the brain functional magnetic resonance imaging data is preprocessed and feature extracted to obtain the characteristic parameters of each functional region of the brain, and the brain network scientific method is used to construct an interaction network of multiple brain regions. Here, specifically, the brain imaging data is preprocessed and feature extracted to obtain the characteristic parameters of each functional region, and the brain network scientific method is used to construct a multi-region interaction network, and the functional connection relationship between different brain regions is modeled to reveal the synergistic relationship between different brain regions. Analyze the interaction between multiple information (disease principles, disease causes, disease status, etc.) and multiple regions (frontal lobe, parietal lobe, temporal lobe, occipital lobe and other brain regions). Based on the disease principles, disease causes, disease status and other multi-information and the characteristic parameters of neuroimaging data extracted from multiple brain regions, construct a multi-information, multi-region interaction network, associate disease information with brain region information, analyze the topological structure characteristics of the network, such as node degree, betweenness centrality, etc., to identify key nodes and key paths, and reveal the interaction between different information and regions.

[0055] S102: constructing a sensitivity evaluation model based on the sensitivity evaluation index, the interaction relationship and the interaction network.

[0056] Regarding the above step S102, in a specific implementation, after the sensitivity evaluation index, the interaction relationship and the interaction network are determined in step S101, a sensitivity evaluation model is constructed based on the sensitivity evaluation index, the interaction relationship and the interaction network.

[0057] As an optional embodiment, for the above step S102, constructing a sensitivity evaluation model based on the sensitivity evaluation index, the interaction relationship and the interaction network includes:

[0058] Step 1021: extract characteristic parameters affecting the action mechanism according to the interaction relationship and the interaction network.

[0059] For the above step 1021, in the specific implementation, according to the interaction relationship between the clinical features and the interaction network between the brain functional areas analyzed in the above steps, the characteristic parameters affecting the mechanism of action are extracted. Here, as an example, the characteristic parameters may include the importance of the clinical features, the connection strength of the brain functional areas, etc., which are not specifically limited in this application.

[0060] Step 1022: Based on the characteristic parameters and the sensitivity evaluation index, a sensitivity evaluation model integrating the variation characteristics of brain functional connectivity and the action characteristic parameters is constructed.

[0061] For the above step 1022, in the specific implementation, based on the extracted characteristic parameters and sensitivity evaluation indicators, the sensitivity of the action mechanism characteristic parameters to the individual brain functional connection variation is reflected. The clinical characteristics and brain functional connection characteristic parameters are integrated to construct a sensitivity evaluation model that integrates the brain functional connection variation characteristics and the action characteristic parameters. Here, specifically, a sensitivity evaluation model that integrates the brain functional connection variation characteristics and the action characteristic parameters is constructed in the following manner: extracting variation characteristics and action characteristics from brain functional magnetic resonance medical imaging data, and performing statistical difference analysis on the brain functional connection data of the patient group and the healthy control group. The variation characteristics include the strength, direction, stability and other adverse change characteristics of the brain functional connection caused by the patient's illness, and the action characteristics include the degree of effect of the brain functional connection on important functions such as cognition, perception, and movement. Based on the selected variation characteristics and action characteristics, the machine learning model support vector machine is used to construct a sensitivity evaluation model. The constructed model is verified and optimized using cross-validation and other methods. The evaluation models of variation characteristics and action characteristics are integrated to construct a sensitivity evaluation model. The influence of variation characteristics and action characteristics on the sensitivity of patients' brain functional connections is analyzed to assist doctors in making decisions such as disease diagnosis and treatment plan selection, thereby improving the effect of personalized medicine.

[0062] S103, extracting action mechanism characteristic elements based on the sensitivity evaluation model, constructing a brain function network dynamics model using the action mechanism characteristic elements, and incrementally updating the brain function network dynamics model based on the action mechanism characteristic elements to obtain an incrementally updated brain function network dynamics model.

[0063] In order to realize the compensation correction of the dynamic model, based on the identification results of the sensitive characteristic parameters of the above-mentioned mechanism model, the following steps are followed: ① Based on the local element characteristics constrained by the association rules, the connection weights between neurons are defined to construct the brain function network dynamic model; ② According to the guidance of the mechanism sensitive characteristic parameters, combined with the characteristics of the dynamic changes of brain function connections, the update and self-learning of the brain function network dynamic model based on the chaotic neural network is realized; ③ The accuracy of the brain function network dynamic model is incrementally updated to realize the compensation correction of the dynamic model. With respect to the above-mentioned step S103, in the specific implementation, the action mechanism characteristic elements are extracted based on the sensitivity evaluation model, the brain function network dynamic model is constructed using the action mechanism characteristic elements, and the brain function network dynamic model is incrementally updated based on the action mechanism characteristic elements to obtain the incrementally updated brain function network dynamic model, so as to realize the incremental update of brain function network dynamics guided by the mechanism characteristics.

[0064] As an optional embodiment, for the above step S103, extracting the action mechanism characteristic elements based on the sensitivity evaluation model and constructing the brain function network dynamics model using the action mechanism characteristic elements include:

[0065] A: Based on the sensitivity evaluation model, characteristic elements of the mechanism of action in complex brain networks are extracted.

[0066] For the above step A, in the specific implementation, the characteristic elements of the action mechanism under the complex brain network are extracted based on the sensitivity evaluation model. Specifically, it is necessary to analyze the associated features, associated information, and associated regions based on incomplete data, and extract the action mechanism and mechanism formation elements between multiple information and multiple regions under the complex brain network under incomplete data based on the sensitivity evaluation model.

[0067] B: Iterative simulation and simulation are performed on the clinical characteristics and regions associated with each factor to generate simulation data related to the characteristic elements of the mechanism of action.

[0068] C: Calculate the correlation coefficients between the nodes of the brain function network using the simulation data, and construct the brain function network dynamics model using the correlation coefficients.

[0069] For the above steps B-C, in the specific implementation, the clinical characteristics and regions associated with each element are iteratively simulated and simulated to generate simulation data related to the characteristic elements of the mechanism of action. The simulation data is used to calculate the correlation coefficient between the nodes of the brain function network, so as to construct a brain function network dynamics model using the correlation coefficient. Specifically, based on the complex brain information digital twin brain function model, the clinical characteristics and regions associated with each element are iteratively simulated and simulated to generate simulation data related to the elements forming the mechanism of action, and the data is used to calculate the Pearson correlation coefficient between the nodes of the brain function network to realize the construction of the brain function network dynamics model.

[0070] Collect existing clinical feature data and brain region data, and establish a digital twin model, that is, a statistical model based on existing data, to simulate individual clinical features and brain region status. Use the digital twin model to iteratively predict the clinical features and brain region status of each individual, and generate a simulated data set for simulation and simulation based on the clinical features and regional status predicted by the model. For clinical feature data, information including patient disease progression and changes in physiological indicators can be generated; for brain region status data, information including functional connection strength and brain region activity can be generated. Use simulated data to calculate the correlation coefficient between brain function network nodes (brain regions) to reflect the functional connection relationship between nodes. Based on the correlation coefficient between brain function network nodes, a dynamic mathematical model describing the dynamic evolution of brain function network is established to simulate the changes and evolution of brain function network at different time points.

[0071] As an optional embodiment, for the above step S103, the incremental updating of the brain function network dynamics model based on the action mechanism characteristic elements to obtain the incrementally updated brain function network dynamics model includes:

[0072] a: Based on the characteristic elements of the mechanism of action, the twin data selection results are determined by using characteristic signal enhancement.

[0073] b: Based on the twin data selection results, the characteristic elements of the action mechanism are analyzed from the perspective of multiple levels of spatiotemporal scales, and the dynamic changes of brain functional connections are processed in real time and fed back to the brain functional network dynamics model to achieve incremental update of the brain functional network dynamics model, thereby obtaining the incrementally updated brain functional network dynamics model.

[0074] For the above steps a-b, in the specific implementation, firstly, according to the characteristic elements of the action mechanism, the twin data selection results are determined by characteristic signal enhancement. Then, based on the twin data selection results, the characteristic elements of the action mechanism are analyzed from the perspective of multi-level spatiotemporal scales, and the dynamic changes of brain function connections are processed in real time and fed back to the brain function network dynamics model to realize the incremental update of the brain function network dynamics model, and obtain the brain function network dynamics model after incremental update. Specifically, according to the extraction results of the characteristic elements of the action mechanism, the characteristic signal enhancement method is used to realize the deep identification and selection of the elements forming the action mechanism of the twin data, and based on the twin data selection results, the elements of the action mechanism of brain function connections under complex brain information are analyzed from the perspective of multi-level spatiotemporal scales, and the dynamic changes of brain function connections are processed in real time and fed back to the brain function network dynamics model, so as to realize the update and self-learning of the brain function network dynamics model based on the chaotic neural network. Feature signal enhancement refers to enhancing feature signals with important information to improve the recognition and selection of features. For the feature signals in the twin data, feature selection is first performed to screen out features related to the formation of the action mechanism. The feature signal is processed by the signal processing technology filtering enhancement algorithm to highlight its information related to the formation of the action mechanism. Based on the enhanced feature signal, deep identification and selection are performed to further determine the features related to the formation of the action mechanism. The elements of the action mechanism of brain functional connections are analyzed from different spatiotemporal scales, from micro to macro, from short time scale to long time scale, from small area to large area, etc. Considering the dynamic change characteristics of brain functional networks at different scales, the functional connection relationship between different brain regions is analyzed. The dynamic changes of brain functional connections are monitored in real time, and the dynamic changes of brain functional connections are processed and fed back in real time by real-time data stream processing. The monitoring results are fed back to the brain functional network dynamics model to realize dynamic updating and self-learning of the model.

[0075] As an optional embodiment, in order to evaluate and verify the accuracy of the brain function network dynamics model and perform compensation correction, the dynamic characteristics of the brain region are analyzed mainly from the following aspects: 1. Multi-scale entropy is used to directly extract the pattern information contained in the original signal and analyze the dynamic characteristics of the brain region; 2. The output signal of the brain network is analyzed using the power spectrum analysis method; 3. The brain wave type of each brain region is thresholded and statistically classified. If the number of brain waves of the same type in a certain brain region is greater than or equal to the threshold, it is considered that this brain wave type is the brain wave type of the brain region in a certain state.

[0076] S104, constructing a twin brain function dynamics model based on the incrementally updated brain function network dynamics model, and generating a semantic description of the action mechanism formation process based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model.

[0077] For the above step S104, in the specific implementation, a twin brain function dynamics model is constructed according to the incrementally updated brain function network dynamics model, and a semantic description of the action mechanism formation process is generated based on the characteristic elements of the action mechanism and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model.

[0078] As an optional embodiment, for the above step S104, the twin brain function dynamics model is constructed based on the incrementally updated brain function network dynamics model, and a semantic description of the mechanism formation process is generated based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model:

[0079] Step 1041, based on the incrementally updated brain function network dynamics model, construct the twin brain function dynamics model by mathematical deduction.

[0080] For the above step 1041, in the specific implementation, according to the incrementally updated brain function network dynamics model, a twin brain function dynamics model is constructed by mathematical deduction. Specifically, according to the results of updating and self-learning of the brain function network dynamics model based on chaotic neural network under the above multi-temporal and spatial scales, the formation process of the action mechanism under multi-feature, multi-information, and multi-regional coordination, as well as the brain function connection mechanism under complex brain information are semantically described, and the model is constructed by mathematical deduction.

[0081] Step 1042, based on the characteristic elements of the action mechanism and the incrementally updated brain functional network dynamics model, determine the semantic description of the action mechanism formation process.

[0082] It should be noted that semantic description is to describe the interaction, influence relationship and mechanism formation process between different features, information and regions through professional knowledge and domain understanding, to describe the qualitative understanding of the brain function connection mechanism and mechanism formation process, and to describe the relationship and mechanism between different elements. Using the results of the above-mentioned brain function network dynamics model update and self-learning, a mathematical equation model is established to describe the relationship and mechanism between different elements. Using the constructed mechanism model, the known features, information and regional data are input to predict the interaction factors between multiple targets, that is, the degree of association and influence between different elements.

[0083] For the above step 1042, in the specific implementation, the semantic description of the process of the formation of the mechanism of action is determined according to the characteristic elements of the mechanism of action and the brain function network dynamics model after incremental update. Specifically, based on the semantic description of the process of the formation of the mechanism of action, the brain function twin model is used to study the marching pattern of neuronal activity, so as to understand how the neuronal activity induced by transcranial magnetic stimulation and / or spontaneous is propagated in the brain and manifested as functional connections; then, the concept of effective symmetry in complex networks is used to establish a multi-level tree of the structural network and a multi-level tree of the functional network, and it is found that there is a very close connection between the two trees, which can help explain the multi-rhythm of the brain in the experiment.

[0084] Step 1043, based on the semantic description, determine the progression pattern of neuronal activity under the twin brain functional dynamics model to perform individualized target prediction based on the twin brain functional dynamics model.

[0085] For the above step 1043, in the specific implementation, based on the brain function network dynamics model and the existing neuroscience knowledge, a brain function twin model is constructed to simulate the individual's brain function state and its change process, and the model can accurately reflect the pattern and change of neuronal activity. By observing the changes in neuronal activity in the simulation of the brain function twin model, the traveling mode of neuronal activity is determined, and the spatiotemporal characteristics of neuronal activity are analyzed, including the time series and spatial distribution of activity, to reveal the propagation law and pattern of neuronal activity. Based on the simulation results of the brain function twin model, the propagation process of transcranial magnetic stimulation or spontaneous neuronal activity in the brain is studied, and how it manifests as functional connection is explored, the connection strength and pattern between different regions in the brain function network are analyzed, and the propagation path and influence degree of neuronal activity are revealed, so as to understand the formation process of functional connection. According to the structural connection relationship of the brain, a multi-level tree of the structural network is constructed, and the structural organization and connection relationship of different levels are hierarchically represented. Based on anatomical knowledge or brain imaging data, a multi-level structural network is constructed, such as a multi-level structure from the brain region to the neuron level. According to the connection pattern and activity characteristics of the brain functional network, a multi-level tree of the functional network is constructed to represent the functional connection relationship at different levels. Based on the simulation results of the brain functional twin model and the simulation experimental data, a multi-level functional network is constructed, such as a multi-level functional network from large-scale brain regions to small-scale neuronal activities. Using the close connection between the trees, a universal principle about the brain network is found: the structural network of the brain is naturally inlaid with many internal groups. The structural network acts as a basis vector. In the normal or abnormal state of the brain, by activating and reorganizing the inherent groups in the structural network, a rich combination of synergistic and non-synergistic states is formed, and the identification of functional antagonism and synergistic areas of brain regions is realized, so as to accurately calculate individualized targets. The characteristics of the individual brain structural network and functional network are analyzed to understand the characteristics of their brain structure and function, provide a basis for the selection of individualized targets, and use the characteristics of the internal groups in the brain network to identify functional antagonism and synergistic areas. Combined with the individual's clinical condition and treatment goals, appropriate targets are selected from the functional antagonism and synergistic areas. Target selection is based on existing neuroimaging data and clinical feature data, combined with the prediction model or guiding algorithm of the treatment effect, for accurate calculation and evaluation.

[0086] In this application, the user's individualized target is determined by their respective functional neuroimaging data. The action mechanism and characteristic parameter identification method of brain functional connections based on complex brain information are used to capture the time-varying characteristics of the brain network. The mechanism-feature-guided brain functional network dynamics incremental update method is adopted to achieve compensation and correction of the dynamic model, and a brain function twin model is constructed to accurately calculate the individualized target, thereby improving the accuracy of individualized target calculation.

[0087] See also Figure 2 , Figure 2This is a schematic diagram of the structure of an individualized target prediction device based on a twin brain functional dynamics model provided in an embodiment of the present application. Figure 2 As shown in , the individualized target prediction device 200 includes:

[0088] A sensitivity evaluation index construction module 201 is used to analyze the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generate an interaction network, extract sensitive feature parameters that affect the action mechanism based on the interaction relationship and the interaction network, and construct a sensitivity evaluation index based on the sensitive feature parameters;

[0089] A sensitivity evaluation model construction module 202, used to construct a sensitivity evaluation model based on the sensitivity evaluation index, the interaction relationship and the interaction network;

[0090] An incremental update module 203, configured to extract characteristic elements of the action mechanism based on the sensitivity evaluation model, construct a brain function network dynamics model using the characteristic elements of the action mechanism, and incrementally update the brain function network dynamics model based on the characteristic elements of the action mechanism to obtain an incrementally updated brain function network dynamics model;

[0091] The prediction module 204 is used to construct a twin brain function dynamics model based on the incrementally updated brain function network dynamics model, and to generate a semantic description of the action mechanism formation process based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model.

[0092] Furthermore, when the sensitivity evaluation index construction module 201 is used to analyze the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generate an interaction network, the sensitivity evaluation index construction module 201 is also used to:

[0093] Modeling the functional connectivity relationship between different brain regions in the functional magnetic resonance imaging data to determine the interaction relationship between different brain regions;

[0094] The functional magnetic resonance imaging data are preprocessed and feature extracted to obtain characteristic parameters of each functional area of ​​the brain, and a brain network science method is used to construct an interaction network of multiple brain areas.

[0095] Further, when the sensitivity evaluation model construction module 202 is used to construct a sensitivity evaluation model based on the sensitivity evaluation index, the interaction relationship and the interaction network, the sensitivity evaluation model construction module 202 is also used to:

[0096] Extracting characteristic parameters affecting the action mechanism according to the interaction relationship and the interaction network;

[0097] Based on the characteristic parameters and the sensitivity evaluation index, a sensitivity evaluation model integrating the variation characteristics of brain functional connectivity and the action characteristic parameters is constructed.

[0098] Furthermore, when the incremental update module 203 is used to extract the characteristic elements of the action mechanism based on the sensitivity evaluation model and construct the brain function network dynamics model using the characteristic elements of the action mechanism, the incremental update module 203 is also used to:

[0099] Extracting characteristic elements of the action mechanism under the complex brain network based on the sensitivity evaluation model;

[0100] Iteratively simulating and modeling the clinical characteristics and regions associated with each element to generate simulation data related to the characteristic elements of the mechanism of action;

[0101] The simulation data is used to calculate the correlation coefficients between the nodes of the brain function network, so as to construct the brain function network dynamics model using the correlation coefficients.

[0102] Furthermore, when the incremental update module 203 is used to incrementally update the brain function network dynamics model based on the action mechanism characteristic elements to obtain the incrementally updated brain function network dynamics model, the incremental update module 203 is also used to:

[0103] According to the characteristic elements of the mechanism of action, the twin data selection result is determined by using characteristic signal enhancement;

[0104] Based on the twin data selection results, the characteristic elements of the action mechanism are analyzed from the perspective of multi-level spatiotemporal scales, and the dynamic changes of brain functional connections are processed in real time and fed back to the brain functional network dynamics model to achieve incremental updating of the brain functional network dynamics model and obtain the incrementally updated brain functional network dynamics model.

[0105] Further, when the prediction module 204 is used to construct a twin brain function dynamics model based on the incrementally updated brain function network dynamics model, and to generate a semantic description of the mechanism formation process based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model, the prediction module 204 is also used to:

[0106] Based on the incrementally updated brain function network dynamics model, constructing the twin brain function dynamics model by mathematical deduction;

[0107] Determining a semantic description of the process of forming the mechanism of action based on the characteristic elements of the mechanism of action and the incrementally updated brain functional network dynamics model;

[0108] Based on the semantic description, the progression pattern of neuronal activity under the twin brain functional dynamics model is determined to perform individualized target prediction based on the twin brain functional dynamics model.

[0109] See also Figure 3 , Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 3 As shown in , the electronic device 300 includes a processor 310 , a memory 320 and a bus 330 .

[0110] The memory 320 stores machine-readable instructions executable by the processor 310. When the electronic device 300 is running, the processor 310 communicates with the memory 320 via the bus 330. When the machine-readable instructions are executed by the processor 310, the above-mentioned Figure 1 The steps of the individualized target prediction method based on the twin brain functional dynamics model in the method embodiment shown in the method embodiment, the specific implementation method can be found in the method embodiment, and will not be repeated here.

[0111] The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the computer program can execute the above-mentioned Figure 1 The steps of the individualized target prediction method based on the twin brain functional dynamics model in the method embodiment shown in the method embodiment, the specific implementation method can be found in the method embodiment, and will not be repeated here.

[0112] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0113] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0114] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0115] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0116] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application can essentially be embodied in the form of a software product, or in other words, the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0117] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A personalized target prediction method based on the twin brain functional dynamics model, characterized in that: The individualized target prediction method comprises: Analyze the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generate an interaction network, extract sensitive feature parameters that affect the action mechanism based on the interaction relationship and the interaction network, and construct a sensitivity evaluation index based on the sensitive feature parameters; Based on the sensitivity evaluation index, the interaction relationship and the interaction network, constructing a sensitivity evaluation model; Extracting characteristic elements of the action mechanism based on the sensitivity evaluation model, constructing a brain function network dynamics model using the characteristic elements of the action mechanism, and incrementally updating the brain function network dynamics model based on the characteristic elements of the action mechanism to obtain an incrementally updated brain function network dynamics model; Constructing a twin brain function dynamics model based on the incrementally updated brain function network dynamics model, and generating a semantic description of the mechanism formation process based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model; The constructing of a sensitivity evaluation model based on the sensitivity evaluation index, the interaction relationship and the interaction network includes: Extracting characteristic parameters affecting the action mechanism according to the interaction relationship and the interaction network; Based on the characteristic parameters and the sensitivity evaluation index, a sensitivity evaluation model integrating the variation characteristics of brain functional connectivity and the action characteristic parameters is constructed; The incrementally updating the brain function network dynamics model based on the action mechanism characteristic elements to obtain the incrementally updated brain function network dynamics model includes: According to the characteristic elements of the mechanism of action, the twin data selection result is determined by using characteristic signal enhancement; Based on the twin data selection results, the characteristic elements of the action mechanism are analyzed from the perspective of multi-level spatiotemporal scales, and the dynamic changes of brain functional connections are processed in real time and fed back to the brain functional network dynamics model to achieve incremental updating of the brain functional network dynamics model and obtain the incrementally updated brain functional network dynamics model.

2. The personalized target prediction method according to claim 1, characterized in that: The method of analyzing the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generating an interaction network includes: Modeling the functional connectivity relationship between different brain regions in the functional magnetic resonance imaging data to determine the interaction relationship between different brain regions; The functional magnetic resonance imaging data are preprocessed and feature extracted to obtain characteristic parameters of each functional area of ​​the brain, and a brain network science method is used to construct an interaction network of multiple brain areas.

3. The personalized target prediction method according to claim 1, characterized in that: The method of extracting characteristic elements of the action mechanism based on the sensitivity evaluation model and constructing a brain function network dynamics model using the characteristic elements of the action mechanism includes: Extracting characteristic elements of the action mechanism under the complex brain network based on the sensitivity evaluation model; Iteratively simulating and modeling the clinical characteristics and regions associated with each element to generate simulation data related to the characteristic elements of the mechanism of action; The simulation data is used to calculate the correlation coefficients between the nodes of the brain function network, so as to construct the brain function network dynamics model using the correlation coefficients.

4. The personalized target prediction method according to claim 1, characterized in that: The step of constructing a twin brain function dynamics model based on the incrementally updated brain function network dynamics model, and generating a semantic description of the mechanism formation process based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model, includes: Based on the incrementally updated brain function network dynamics model, constructing the twin brain function dynamics model by mathematical deduction; Determining a semantic description of the process of forming the mechanism of action based on the characteristic elements of the mechanism of action and the incrementally updated brain functional network dynamics model; Based on the semantic description, the progression pattern of neuronal activity under the twin brain functional dynamics model is determined to perform individualized target prediction based on the twin brain functional dynamics model.

5. An individualized target prediction device based on a twin brain functional dynamics model, characterized in that: The individualized target prediction device comprises: A sensitivity evaluation index construction module, used to analyze the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generate an interaction network, extract sensitive feature parameters that affect the action mechanism based on the interaction relationship and the interaction network, and construct a sensitivity evaluation index based on the sensitive feature parameters; A sensitivity evaluation model construction module, used to construct a sensitivity evaluation model based on the sensitivity evaluation index, the interaction relationship and the interaction network; An incremental update module, used for extracting characteristic elements of the action mechanism based on the sensitivity evaluation model, constructing a brain function network dynamics model using the characteristic elements of the action mechanism, and incrementally updating the brain function network dynamics model based on the characteristic elements of the action mechanism to obtain an incrementally updated brain function network dynamics model; A prediction module, for constructing a twin brain function dynamics model based on the incrementally updated brain function network dynamics model, and generating a semantic description of the mechanism formation process based on the action mechanism characteristic elements and the incrementally updated brain function network dynamics model, so as to predict individualized targets based on the semantic description and the twin brain function dynamics model; When the sensitivity evaluation model construction module is used to construct a sensitivity evaluation model based on the sensitivity evaluation index, the interaction relationship and the interaction network, the sensitivity evaluation model construction module is also used to: Extracting characteristic parameters affecting the action mechanism according to the interaction relationship and the interaction network; Based on the characteristic parameters and the sensitivity evaluation index, a sensitivity evaluation model integrating the variation characteristics of brain functional connectivity and the action characteristic parameters is constructed; When the incremental update module is used to incrementally update the brain function network dynamics model based on the action mechanism characteristic elements to obtain the incrementally updated brain function network dynamics model, the incremental update module is also used to: According to the characteristic elements of the mechanism of action, the twin data selection result is determined by using characteristic signal enhancement; Based on the twin data selection results, the characteristic elements of the action mechanism are analyzed from the perspective of multi-level spatiotemporal scales, and the dynamic changes of brain functional connections are processed in real time and fed back to the brain functional network dynamics model to achieve incremental updating of the brain functional network dynamics model and obtain the incrementally updated brain functional network dynamics model.

6. The individualized target prediction device according to claim 5, characterized in that: When the sensitivity evaluation index construction module is used to analyze the interaction relationship of brain region information based on the user's brain functional magnetic resonance medical imaging data and generate an interaction network, the sensitivity evaluation index construction module is also used to: Modeling the functional connectivity relationship between different brain regions in the functional magnetic resonance imaging data to determine the interaction relationship between different brain regions; The functional magnetic resonance imaging data are preprocessed and feature extracted to obtain characteristic parameters of each functional area of ​​the brain, and a brain network science method is used to construct an interaction network of multiple brain areas.

7. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate through the bus. When the machine-readable instructions are run by the processor, the steps of the individualized target prediction method based on the twin brain functional dynamics model as described in any one of claims 1 to 4 are executed.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the individualized target prediction method based on the twin brain functional dynamics model as described in any one of claims 1 to 4 are executed.

Images