Effective connection information derivation method, digital twin brain model training method, device and equipment

Abstract

The invention discloses a brain effective connection information derivation method, a digital twin brain model training method, a device and equipment. The method comprises the following steps: determining disturbance data of a first brain node in at least two brain nodes and brain neural data of the at least two brain nodes in a time sequence; obtaining first prediction neural data of the at least two brain nodes at the next moment based on the disturbance data of the first brain node and the brain neural data of the at least two brain nodes in the time sequence through a digital twin brain model; and determining effective brain connection information from the first brain node to the second brain node based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment. And training the time sequence prediction network based on the training data set to obtain a digital twin brain model.

Description

Technical Field

[0001] The present invention relates to the field of brain neuroscience technology, and in particular to a method for deriving effective brain connection information, and a training method, device, and equipment for a digital twin brain model. Background Art

[0002] Effective connectivity (EC) can characterize the causal interactions between brain regions and is the basis for understanding brain information processing. Among them, effective connectivity can be obtained by experimental methods and data-driven methods.

[0003] However, common experimental methods are not applicable to stimulation and observation across the entire human brain; common data-driven effective connection inference methods and model-based methods have high computational complexity, and model-free methods can only distinguish whether directed connections exist.

[0004] It can be seen that common experimental methods and data-driven methods are unable to obtain accurate information about effective brain connections. Summary of the invention

[0005] The embodiments of the present application provide a method for deriving effective brain connection information, a training method, device, and equipment for a digital twin brain model, which can obtain accurate effective brain connection information.

[0006] The technical solution of the embodiment of the present application is implemented as follows:

[0007] In a first aspect, an embodiment of the present application provides a method for deriving effective brain connection information, the method comprising:

[0008] determining disturbance data of a first brain node among at least two brain nodes and brain neural data of the at least two brain nodes in a time series;

[0009] Obtaining, by means of a digital twin brain model, first predicted neural data of the at least two brain nodes at the next moment based on the disturbance data of the first brain node and the brain neural data of the at least two brain nodes in a time series; wherein the digital twin brain model is obtained by training based on a time series prediction network;

[0010] Based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment, the effective brain connection information from the first brain node to the second brain node is determined; wherein the second brain node is the other brain node among the at least two brain nodes except the first brain node; the second predicted neural data represents the predicted data obtained when the disturbance data is not added.

[0011] In an embodiment of the present application, based on the digital twin brain model obtained through training, the brain's response to the disturbance stimulus is predicted and the effective brain connection information is inferred by adding disturbance data to the brain nodes. Among them, the digital twin brain model is obtained through training based on the time series prediction network. Therefore, the digital twin brain model can effectively and accurately predict the dynamic changes of brain neural data, thereby obtaining more accurate effective brain connection information.

[0012] In a second aspect, an embodiment of the present application provides a method for training a digital twin brain model, the method comprising:

[0013] The time series prediction network is trained based on the training data set to obtain a digital twin brain model; wherein the digital twin brain model is used to predict the neural data of the next moment based on the brain neural data of the brain node in the time series;

[0014] The training data set includes neural training data of at least two brain nodes corresponding to a preset time length p+1; wherein the neural training data corresponding to the preset time length p+1 includes neural training data from the mpth moment to the mth moment, and the training data set also includes neural training data of the at least two brain nodes at the m+1th moment; p is an integer greater than or equal to 0, and m is an integer greater than p;

[0015] The time series prediction network is trained based on the training data set to obtain a digital twin brain model, including:

[0016] Obtaining, by the time series prediction network, prediction data of the at least two brain nodes at the m+1th moment based on the neural training data of the at least two brain nodes from the mpth moment to the mth moment;

[0017] Based on the prediction data of the at least two brain nodes at the m+1th moment and the neural training data of the at least two brain nodes at the m+1th moment, the time series prediction network is corrected to obtain the digital twin brain model.

[0018] In an embodiment of the present application, a time series prediction network can be used to learn and predict the dynamic changes of brain neural signals to obtain a digital twin brain model with time series prediction function. Among them, the digital twin brain model can effectively and accurately predict the dynamic changes of brain neural data.

[0019] In a third aspect, an embodiment of the present application provides a device for deriving effective brain connection information, the device for deriving effective brain connection information comprising:

[0020] a determining unit, configured to determine disturbance data of a first brain node among at least two brain nodes and brain neural data of the at least two brain nodes in a time series;

[0021] An acquisition unit, configured to obtain, through a digital twin brain model, first predicted neural data of the at least two brain nodes at a next moment based on the disturbance data of the first brain node and the brain neural data of the at least two brain nodes in a time series; wherein the digital twin brain model is obtained by training based on a time series prediction network;

[0022] The determination unit is further used to determine the effective brain connection information between the first brain node and the second brain node based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment; wherein the second brain node is the other brain node among the at least two brain nodes except the first brain node; and the second predicted neural data represents the predicted data obtained when the disturbance data is not added.

[0023] In a fourth aspect, an embodiment of the present application provides a training device for a digital twin brain model, wherein the training device for the digital twin brain model comprises:

[0024] A training unit, used to train a time series prediction network based on a training data set to obtain a digital twin brain model; wherein the digital twin brain model is used to predict the neural data of the next moment based on the brain neural data of the brain node in the time series;

[0025] The training data set includes neural training data of at least two brain nodes corresponding to a preset time length p+1; wherein the neural training data corresponding to the preset time length p+1 includes neural training data from the mpth moment to the mth moment, and the training data set also includes neural training data of the at least two brain nodes at the m+1th moment; p is an integer greater than or equal to 0, and m is an integer greater than p;

[0026] The time series prediction network is trained based on the training data set to obtain a digital twin brain model, including:

[0027] Obtaining, by the time series prediction network, prediction data of the at least two brain nodes at the m+1th moment based on the neural training data of the at least two brain nodes from the mpth moment to the mth moment;

[0028] Based on the prediction data of the at least two brain nodes at the m+1th moment and the neural training data of the at least two brain nodes at the m+1th moment, the time series prediction network is corrected to obtain the digital twin brain model.

[0029] In a fifth aspect, an embodiment of the present application provides a computer device, the computer device comprising: a processor and a memory; wherein:

[0030] The memory is used to store a computer program that can be run on the processor;

[0031] The processor is used to execute the method described in the first aspect or the second aspect when running the computer program.

[0032] In a sixth aspect, an embodiment of the present application provides a computer-readable storage medium having a program stored thereon, and when the program is executed by a processor, the method described in the first aspect or the second aspect above is implemented.

[0033] In a seventh aspect, an embodiment of the present application provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements the method described in the first or second aspect above.

[0034] It can be seen that in the embodiments of the present application, the digital twin brain model is obtained based on the time series prediction network training. Therefore, the digital twin brain model can effectively and accurately predict the dynamic changes of brain neural data, thereby obtaining more accurate brain effective connection information. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 Schematic diagram of the implementation process of the method for deriving effective brain connection information proposed in the embodiment of the present application Figure 1 ;

[0036] Figure 2 Schematic diagram of the brain node proposed in the embodiment of the present application Figure 1 ;

[0037] Figure 3 Schematic diagram of the implementation process of the method for deriving effective brain connection information proposed in the embodiment of the present application Figure 2 ;

[0038] Figure 4 Schematic diagram of the derivation of effective brain connection information proposed in the embodiment of the present application Figure 1 ;

[0039] Figure 5 Schematic diagram of the derivation of effective brain connection information proposed in the embodiment of the present application Figure 2 ;

[0040] Figure 6 Schematic diagram of the derivation of effective brain connection information proposed in the embodiment of the present application Figure 3 ;

[0041] Figure 7 Schematic diagram of the derivation of effective brain connection information proposed in the embodiment of the present application Figure 4 ;

[0042] Figure 8 Schematic diagram of the derivation of effective brain connection information proposed in the embodiment of the present application Figure 5 ;

[0043] Fig. 9 A schematic diagram of the implementation flow of the training of the digital twin brain model proposed in the embodiment of the present application;

[0044] Fig.10 Schematic diagram of the brain node proposed in the embodiment of the present application Figure 2 ;

[0045] Fig.11 Schematic diagram of neural data analysis under real stimulation;

[0046] Fig.12 A schematic diagram of the implementation of the present application embodiment for deriving effective brain connection information based on a digital twin brain model obtained through training;

[0047] Fig.13 A schematic diagram of the structure of the device for deriving effective brain connection information proposed in an embodiment of the present application;

[0048] Fig.14 A schematic diagram of the composition structure of a training device for a digital twin brain model proposed in an embodiment of the present application;

[0049] Fig.15 A schematic diagram of the composition structure of the computer device proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. It is understood that the specific embodiments described herein are only used to explain the related applications, rather than to limit the applications. It should also be noted that, for ease of description, only the parts related to the related applications are shown in the drawings.

[0051] Brain network models are often used to characterize the connections between different brain regions and the transitions between brain states. They are an important method for understanding the working mechanisms of the brain at different scales and in different states. Exploring the causal relationship between different regions in the brain network is crucial for a deeper understanding of the brain's advanced cognitive functions and for locating the lesions of functional brain diseases.

[0052] The causal relationship in the brain network describes how one brain region directly affects other brain regions. Constructing a causal brain network requires determining the magnitude and direction of this effect and distinguishing between excitatory and inhibitory effects. This causal relationship is also called effective connectivity.

[0053] However, the dynamic changes of brain networks are highly nonlinear. It is difficult to accurately describe the complexity of the whole brain network dynamics by constructing a linear dynamic model with finite state variables. The direction and magnitude of information flow between brain regions and the causality of their interactions are still unclear. Therefore, how to accurately establish a brain network dynamics model to describe the complex dynamic processes in the brain is a huge challenge.

[0054] Currently, effective connectivity can be obtained by experimental and data-driven methods.

[0055] The experimental method is to perturb the brain signal of one area through electrical or magnetic stimulation and observe how this perturbation propagates to other brain areas to obtain effective connections. However, it is technically not feasible to apply stimulation and observation to the whole brain at the same time. Therefore, many studies have tried to infer effective connections from neural data, such as functional magnetic resonance imaging data and electroencephalogram data.

[0056] Data-driven methods include model-based methods and model-free methods.

[0057] Model-based methods usually parameterize effective connectivity in a generative model and estimate effective connectivity by fitting neural signals based on model assumptions. For example, dynamic causal models use biophysical models with effective connectivity as parameters in neural dynamics, and convert the generated neural signals into blood oxygen level-dependent signals through hemodynamic functions for fitting functional magnetic resonance imaging data. However, although dynamic causal models have been widely used to estimate effective connectivity between a few regions, due to their high computational complexity in parameter estimation, this method is not suitable for studying whole-brain effective connectivity that requires considering signals from all brain regions. Another limitation of such model-based methods is that they rely on model assumptions, which may lead to large inference biases if the model and actual brain dynamics do not match.

[0058] Model-free methods do not rely on explicit assumptions about the underlying neural dynamics and use statistical methods to infer effective connections from observed neural signals. For example, Granger causality uses time series analysis to determine the extent to which past neural activity in one region can predict future neural activity in another region. This method can be used to infer the strength and direction of effective connections, but it cannot distinguish between excitatory and inhibitory connections. In addition, the validity of Granger causality analysis is also controversial due to conceptual and performance limitations.

[0059] In addition to statistical methods, some causal inference models based on deep learning have been proposed in recent years. Such neural network-based causal inference models generally represent brain networks with graph structures, use data-driven methods to take effective connections in neural signals as implicit tasks, and use explicit tasks such as neural signal prediction or classification to drive the learning of connection relationships between different brain regions in the brain network in implicit tasks. This type of data-driven causal inference neural network model has achieved excellent performance in many fields such as physical system trajectory prediction, and has great application potential for inferring effective connections in the brain. However, most of these methods can only distinguish whether directed connections exist, and cannot characterize the weights and signs of connections. The individual differences in the obtained connections are small, and their application is limited in the selection of targets for precise neural regulation.

[0060] It can be seen from this that common experimental methods, such as perturbing brain signals in one area through electrical or magnetic stimulation and observing how this perturbation propagates to other brain areas to obtain effective connections, are not suitable for stimulation and observation of the entire human brain. Common data-driven effective connection inference methods, model-based methods have high computational complexity, and model-free methods can only distinguish whether there is a directed connection.

[0061] In other words, neither common experimental methods nor data-driven methods can obtain accurate information about the brain's effective connections.

[0062] In order to solve the above problems, in the embodiments of the present application, on the one hand, a time series prediction network can be used to learn and predict the dynamic changes of brain neural signals to obtain a digital twin brain model with a time series prediction function; on the other hand, based on the digital twin brain model obtained through training, the brain's response under the disturbance stimulus is predicted by adding disturbance data to the brain nodes and the effective connection information of the brain is inferred. It can be seen that in the embodiments of the present application, the digital twin brain model is obtained based on the training of the time series prediction network. Therefore, the digital twin brain model can effectively and accurately predict the dynamic changes of brain neural data, thereby obtaining more accurate effective connection information of the brain.

[0063] The technical solutions 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.

[0064] An embodiment of the present application provides a method for deriving effective brain connectivity information, wherein the method for deriving effective brain connectivity information can be applied to a device or computer device for deriving effective brain connectivity information, and the present application does not specifically limit it. Below, taking a device for deriving effective brain connectivity information as an example, the method for deriving effective brain connectivity information proposed in the embodiment of the present application is exemplarily described.

[0065] Furthermore, in the embodiments of the application, Figure 1 Schematic diagram of the implementation process of the method for deriving effective brain connection information proposed in the embodiment of the present application Figure 1 ,like Figure 1 As shown, the method for deriving effective brain connection information may include the following steps:

[0066] Step 101: Determine disturbance data of a first brain node among at least two brain nodes and brain neural data of at least two brain nodes in a time series.

[0067] In an embodiment of the present application, the device for deriving effective brain connection information can first determine the disturbance data of a first brain node among at least two brain nodes, and can also determine the brain neural data of at least two brain nodes in a time series.

[0068] It can be understood that the derivation method of brain effective connection information proposed in the embodiment of the present application can determine the brain effective connection information between different brain nodes, wherein the brain node can represent the brain region, that is, it can be considered that one brain node corresponds to one brain region. The brain can be divided into at least two brain regions, and accordingly, one brain can correspond to at least two brain nodes.

[0069] For example, in some embodiments, Figure 2 Schematic diagram of the brain node proposed in the embodiment of the present application Figure 1 ,like Figure 2 As shown, assuming that the biological brain is divided into multiple regions, such as region A, region B, region C, and region D, accordingly, it can be considered that the brain at least corresponds to brain node a, brain node b, brain node c, and brain node d.

[0070] It is understandable that in the embodiment of the present application, in the process of deriving the effective brain connection information, the brain neural data corresponding to the time series of each of the at least two brain nodes can be first obtained. In other words, the derivation device of the effective brain connection information needs to first obtain the brain neural data corresponding to each brain node at multiple moments.

[0071] In the embodiments of the present application, the time series may include continuous time sampling intervals or discrete time sampling points, which is not specifically limited in the present application.

[0072] Further, in an embodiment of the present application, the brain neural data of at least two brain nodes in the time series may include brain neural data of at least two brain nodes corresponding to a preset time length p+1. When predicting the brain neural data at the next moment, the preset time length may determine the time length of the brain neural data used in the prediction process. Wherein, p is an integer greater than or equal to 0.

[0073] It is understood that in the embodiment of the present application, the time length of the brain neural data used in the prediction process, that is, the preset time length p+1, can be preset. For example, if p=2 is preset, then the value of the preset time length p+1 is 3.

[0074] It can be understood that in an embodiment of the present application, the brain neural data of at least two brain nodes in a time series may include brain neural data of at least two brain nodes corresponding to multiple moments, and the brain neural data at multiple moments include but are not limited to brain neural data of at least two brain nodes corresponding to a preset time length p+1, that is, the time series length corresponding to the time series is greater than or equal to the time length of the brain neural data used in the prediction process.

[0075] Exemplarily, in some embodiments, the time series length corresponding to the time series is 4800, while the time length of the brain neural data used in the prediction process is 3.

[0076] It can be understood that in the embodiments of the present application, for a determined preset time length, the time length of the data used for brain neural data prediction is fixed. For example, for the preset time length p+1, the brain neural data of the previous p+1 moments can be used to predict the brain neural data of the next moment.

[0077] That is to say, in the embodiments of the present application, the brain nerve data at one or more moments can be used to predict the brain nerve data at the next moment. For example, if p=2 is preset, the brain nerve data at three consecutive moments, namely, moment 1, moment 2, and moment 3, can be used to predict the brain nerve data at moment 4, the brain nerve data at three consecutive moments, namely, moment 2, moment 3, and moment 4, can be used to predict the brain nerve data at moment 5, and the brain nerve data at three consecutive moments, namely, moment 3, moment 4, and moment 5, can be used to predict the brain nerve data at moment 6.

[0078] Exemplarily, in some embodiments, assuming that the preset time length is p+1, the time length indicated by the preset time length includes p+1 moments from moment tp to moment t. Accordingly, the brain nerve data corresponding to the preset time length p+1 includes the brain nerve data from moment tp to moment t, where t is an integer greater than p.

[0079] That is, in some embodiments, the time series may include p+1 consecutive moments from moment tp to moment t. Accordingly, the brain neural data of each brain node in the time series may include p+1 brain neural data of the brain node corresponding to the p+1 moment.

[0080] Furthermore, in the embodiments of the present application, a variety of methods may be used to determine the brain neural data of any brain node within a time series, and the present application does not make any specific limitations.

[0081] Exemplarily, in some embodiments, brain neural data can be acquired through existing bioinformatics databases, wherein the bioinformatics databases include but are not limited to one or more public data sets such as human fMRI data sets, EEG data sets, and MEG data sets.

[0082] Exemplarily, in some embodiments, brain neural data may also be generated through computational models.

[0083] Further, in an embodiment of the present application, the first brain node may be any one of the at least two brain nodes. For example, the first brain node may be brain node a, or may also be brain node b.

[0084] In an embodiment of the present application, in the process of deriving effective brain connection information, it is also necessary to apply certain disturbance data to the first brain node, so as to utilize the changes in the activities of other brain nodes other than the first brain node after adding the disturbance data to characterize the effective connection of the first brain node to other brain nodes.

[0085] It can be understood that in the embodiment of the present application, the disturbance data of the first brain node can be understood as a virtual disturbance corresponding to the first brain node. Among them, the present application does not specifically limit the form of expression of the disturbance data of the first brain node. For example, an N-dimensional virtual disturbance vector corresponding to N brain nodes can be pre-constructed, wherein N is an integer greater than 1, and the disturbance data of the first brain node can be an element corresponding to the N-dimensional virtual disturbance vector, and the value of the element is a non-zero value, which represents the intensity of the virtual disturbance increased on the first brain node, and the value of other elements is 0, that is, other brain nodes other than the first brain node do not increase virtual disturbances.

[0086] Further, in an embodiment of the present application, when determining disturbance data of a first brain node among the at least two brain nodes, the disturbance data of the first brain node may be set based on a preset disturbance intensity.

[0087] For example, in some embodiments, for an N-dimensional virtual perturbation vector P corresponding to N brain nodes, assuming that each element p in the N-dimensional virtual perturbation vector P corresponds to a brain node, if the i-th brain node is the first brain node to which the virtual perturbation is applied, then when determining the perturbation data of the first brain node, it is possible to select the i-th element p in the N-dimensional virtual perturbation vector P corresponding to the first brain node as iSet to Δ, where Δ can be the preset disturbance intensity; at the same time, the i-th element p in the N-dimensional virtual disturbance vector P can be i All other elements except are set to 0. For example, the following formula (1):

[0088] p i =Δ, p k，k≠i =0 (1);

[0089] Step 102: Using a digital twin brain model, based on the disturbance data of a first brain node and brain neural data of at least two brain nodes in a time series, obtain first predicted neural data of at least two brain nodes at the next moment; wherein the digital twin brain model is obtained by training based on a time series prediction network.

[0090] In an embodiment of the present application, after determining the disturbance data of a first brain node among at least two brain nodes and the brain neural data of at least two brain nodes in a time series, the digital twin brain model can be used to further determine the first predicted neural data of at least two brain nodes at the next moment based on the disturbance data of the first brain node and the brain neural data of at least two brain nodes in a time series, that is, when there is a virtual disturbance in the first brain node, the brain neural data of other brain nodes are predicted by the digital twin brain model.

[0091] Furthermore, in an embodiment of the present application, the digital twin brain model can be obtained by training based on a time series prediction network. Accordingly, the digital twin brain model can be used to determine the effective brain connection information based on the disturbance data of the brain nodes and the brain neural data of the brain nodes in the time series.

[0092] That is to say, in the embodiments of the present application, the digital twin brain model is obtained based on the time series prediction network training. Therefore, the digital twin brain model can effectively predict the dynamic changes of brain neural signals (brain neural data), and thus can obtain more accurate brain effective connection information.

[0093] Furthermore, in an embodiment of the present application, the time series prediction network used to train the digital twin brain model can be an artificial neural network of any type and structure for time series prediction, and the present application does not make any specific limitations.

[0094] Exemplarily, in some embodiments, the time series prediction network may use classic models such as multi-layer perceptron, convolutional neural network, recurrent neural network, etc., or other network architectures suitable for time series prediction may be selected, which is not specifically limited in this application.

[0095] Exemplarily, in some embodiments, for a time series of p+1 consecutive moments from the tpth moment to the tth moment, when obtaining the first predicted neural data of at least two brain nodes at the next moment based on the disturbance data of the first brain node and the brain neural data of at least two brain nodes in the time series through the digital twin brain model, the perturbed brain neural data of at least two brain nodes at the tth moment can be first determined based on the disturbance data of the first brain node and the brain neural data of at least two brain nodes at the tth moment; and then the first predicted neural data of at least two brain nodes at the t+1th moment can be further obtained through the digital twin brain model based on the perturbed brain neural data of at least two brain nodes at the tth moment and the brain neural data of at least two brain nodes from the tpth moment to the t-1th moment.

[0096] It can be understood that in the embodiments of the present application, in the process of dynamically predicting brain neural data through a digital twin brain model, the disturbed brain neural data with added virtual disturbance corresponding to the current moment can be generated based on the disturbance data of the first brain node corresponding to the current moment (such as the tth moment) and the brain neural data of all brain nodes corresponding to the current moment.

[0097] Further, in an embodiment of the present application, based on the disturbance data of the first brain node and the brain nerve data of at least two brain nodes at the tth moment, the disturbed brain nerve data at the tth moment can be generated in a variety of ways. For example, including but not limited to performing mathematical operations on the disturbance data of the first brain node and the brain nerve data of at least two brain nodes at the tth moment, so as to obtain the disturbed brain nerve data at the tth moment.

[0098] Exemplarily, in some embodiments, when determining the disturbed brain neural data of at least two brain nodes at moment t based on the disturbance data of the first brain node and the brain neural data of at least two brain nodes at moment t, it is possible to choose to superimpose the disturbance data of the first brain node and the brain neural data of the first brain node at moment t, and then combine them with the brain neural data of other brain nodes at moment t to obtain the disturbed brain neural data of at least two brain nodes at moment t.

[0099] It can be understood that in the embodiment of the present application, the perturbed brain neural data of at least two brain nodes at time t and the brain neural data of at least two brain nodes from time tp to time t-1 can be used as the perturbed brain neural data of the digital twin brain model to predict the brain neural data at the next moment, and finally the first predicted neural data of at least two brain nodes at time t+1 can be output. Among them, the perturbed brain neural data of at least two brain nodes at time t are generated based on the perturbed data of the first brain node, so it can be considered that the digital twin brain model predicts the brain neural data with the addition of virtual disturbance.

[0100] That is to say, in an embodiment of the present application, the first predicted neural data of at least two brain nodes at time t+1 are the brain neural data predicted after adding the virtual disturbance.

[0101] Exemplarily, in some embodiments, the first predicted neural data of at least two brain nodes at time t+1 are determined by the digital twin brain model. The process can be referred to the following formula (2):

[0102]

[0103] The digital twin brain model can be expressed as f(·), P is the disturbance data, which can be an N-dimensional virtual disturbance vector, x (t-p) , x (t-p+1) , ..., x t It can represent brain neural data from time point (tp) to time point t.

[0104] Exemplarily, in some embodiments, the first predicted neural data of at least two brain nodes at time t+1 are determined by the digital twin brain model. The process can also be referred to the following formula (3):

[0105]

[0106] Among them, θ represents all parameters in the digital twin brain model that can be optimized through training iterations.

[0107] Furthermore, in the embodiments of the present application, Figure 3 Schematic diagram of the implementation process of the method for deriving effective brain connection information proposed in the embodiment of the present application Figure 2 ,like Figure 3 As shown, the method for deriving brain effective connection information may also include the following steps:

[0108] Step 104: Using the digital twin brain model, based on the brain neural data of at least two brain nodes in a time series, obtain second predicted neural data of at least two brain nodes at the next moment.

[0109] In an embodiment of the present application, after determining the disturbance data of a first brain node among at least two brain nodes and the brain neural data of at least two brain nodes in a time series, the digital twin brain model can be used to further determine the second predicted neural data of at least two brain nodes at the next moment based on the brain neural data of at least two brain nodes in a time series, that is, when there is no virtual disturbance in the first brain node, the brain neural data of other brain nodes are predicted by the digital twin brain model.

[0110] Exemplarily, in some embodiments, for a time series of p+1 consecutive moments from the tpth moment to the tth moment, when the digital twin brain model is used to obtain the first predicted neural data of at least two brain nodes at the next moment based on the brain neural data of at least two brain nodes in the time series, the brain neural data of at least two brain nodes from the tpth moment to the tth moment can be used as input data of the digital twin brain model to predict the brain neural data at the next moment, and finally the second predicted neural data of at least two brain nodes at the t+1th moment can be output.

[0111] It can be understood that in the embodiments of the present application, compared with the first predicted neural data, the second predicted neural data of at least two brain nodes at time t+1 can be understood as predicted brain neural data without adding virtual disturbance.

[0112] That is, in an embodiment of the present application, the first predicted neural data may represent the predicted data obtained when the disturbance data is added, and the second predicted neural data may represent the predicted data obtained when the disturbance data is not added.

[0113] For example, in some embodiments, the second predicted neural data of at least two brain nodes at the time t+1 is determined by the digital twin brain model. The process can be referred to the following formula (4):

[0114]

[0115] The digital twin brain model can be expressed as f(·), x (t-p) , x (t-p+1) , ..., x t It can represent brain neural data from time point (tp) to time point t.

[0116] For example, in some embodiments, the second predicted neural data of at least two brain nodes at the time t+1 is determined by the digital twin brain model. The process can also be referred to the following formula (5):

[0117]

[0118] Among them, θ represents all parameters in the digital twin brain model that can be optimized through training iterations.

[0119] Step 103: Determine effective brain connection information from the first brain node to the second brain node based on first predicted neural data of at least two brain nodes at the next moment and second predicted neural data of at least two brain nodes at the next moment; wherein the second brain node is other brain node among the at least two brain nodes except the first brain node; and the second predicted neural data represents predicted data obtained when no disturbance data is added.

[0120] In an embodiment of the present application, after the digital twin brain model predicts the first predicted neural data and the second predicted neural data at the next moment with and without adding virtual disturbances, the effective brain connection information from the first brain node to the second brain node can be further determined based on the first predicted neural data of at least two brain nodes at the next moment and the second predicted neural data of at least two brain nodes at the next moment.

[0121] It can be understood that in the embodiment of the present application, the second brain node can be other brain nodes other than the first brain node among the at least two brain nodes. For example, the first brain node is brain node a, and the second brain node can be one or more of brain node b, brain node c, and brain node d.

[0122] In an embodiment of the present application, the effective brain connection information from the first brain node to the second brain node can be used to determine the corresponding changes generated by the second brain node after the virtual disturbance is applied to the first brain node.

[0123] It can be understood that in the embodiment of the present application, the first predicted neural data represents the predicted data obtained when the disturbance data is added, and the second predicted neural data represents the predicted data obtained when the disturbance data is not added. Therefore, based on the first predicted neural data and the second predicted neural data, the brain effective connection information between the first brain node and any other brain node can be determined.

[0124] Further, in an embodiment of the present application, the effective brain connection information from the first brain node to the second brain node determined based on the first predicted neural data and the second predicted neural data may include but is not limited to the effective connection type, the effective connection direction and the effective connection strength.

[0125] It is understood that in the embodiments of the present application, the effective connection type can determine and distinguish the influence property of the effective brain connection between two brain nodes, wherein the effective connection type can be but is not limited to excitatory type and inhibitory type.

[0126] It can be understood that, in the embodiments of the present application, the effective connection direction can determine the transmission direction of the effective brain connection between two brain nodes.

[0127] It can be understood that, in the embodiments of the present application, the effective connection strength can determine the signal size of the effective brain connection between two brain nodes.

[0128] It can be seen that in the embodiments of the present application, the effective brain connection information between two brain nodes obtained based on the digital twin brain model can not only clarify the causal interactions within the brain, but also provide effective connection data with directionality, intensity and excitation / inhibition characteristics.

[0129] Further, in an embodiment of the present application, when determining the effective brain connection information from the first brain node to the second brain node based on the first predicted neural data of at least two brain nodes at the next moment and the second predicted neural data of at least two brain nodes at the next moment, the effective connection type can be determined to be an excitatory type when the first predicted neural data of the second brain node at the next moment is greater than the second predicted neural data of the second brain node at the next moment; and the effective connection type can be determined to be an inhibitory type when the first predicted neural data of the second brain node at the next moment is less than the second predicted neural data of the second brain node at the next moment. That is, in an embodiment of the present application, the first predicted neural data and the second predicted neural data obtained when the virtual disturbance is added and when the virtual disturbance is not added are compared, and the corresponding effective connection type can be further determined according to the comparison result. Among them, if the first predicted neural data is greater than the second predicted neural data, it can be considered that the neural activity corresponding to the second brain node in the case of the presence of disturbance data corresponding to the first brain node has increased, and at this time, the corresponding effective connection type can be determined to be an excitatory type; if the first predicted neural data is less than the second predicted neural data, it can be considered that the neural activity corresponding to the second brain node in the case of the presence of disturbance data corresponding to the first brain node has decreased, and at this time, the corresponding effective connection type can be determined to be an inhibitory type.

[0130] It can be seen that in the embodiments of the present application, the effective connection type can reflect the increase or decrease in the corresponding neural activities of other brain nodes after adding a virtual disturbance to the first brain node, and the resulting effects correspond to excitatory or inhibitory effects respectively.

[0131] Exemplarily, in some embodiments, for a time series of p+1 consecutive moments from moment tp to moment t, when determining the effective brain connection information from the first brain node to the second brain node based on the first predicted neural data of at least two brain nodes at the next moment and the second predicted neural data of at least two brain nodes at the next moment, when the first predicted neural data of the second brain node at moment t+1 is greater than the second predicted neural data of the second brain node at moment t+1, the effective connection type can be determined to be an excitatory type; and when the first predicted neural data of the second brain node at moment t+1 is less than the second predicted neural data of the second brain node at moment t+1, the effective connection type can be determined to be an inhibitory type. That is to say, in an embodiment of the present application, disturbance data is added to the first brain node at time t, and the first predicted neural data of the second brain node at time t+1 is predicted based on the brain neural data of the first brain node and the second brain node from time tp to time t; at the same time, the second predicted neural data of the second brain node at time t+1 is directly predicted based on the brain neural data of the first brain node and the second brain node from time tp to time t; if the first predicted neural data of the second brain node at time t+1 is greater than the second predicted neural data, then it can be considered that the neural activity corresponding to the second brain node has increased, that is, the disturbance data has an excitatory effect on the second brain node, and therefore the effective connection type can be determined to be excitatory; conversely, if the first predicted neural data of the second brain node at time t+1 is less than the second predicted neural data, then it can be considered that the neural activity corresponding to the second brain node has decreased, that is, the disturbance data has an inhibitory effect on the second brain node, and therefore the effective connection type can be determined to be inhibitory.

[0132] Further, in an embodiment of the present application, when determining the effective brain connection information from the first brain node to the second brain node, the effective connection direction may be directly determined as from the first brain node to the second brain node.

[0133] It can be understood that in the embodiments of the present application, when the neural data of other brain nodes are monitored while applying a virtual disturbance to the first brain node, it can be considered that the effective connection direction corresponding to the effective brain connection information between the first brain node and the second brain node is from the first brain node to the second brain node.

[0134] Furthermore, in an embodiment of the present application, when determining the effective brain connection information from a first brain node to a second brain node based on the first predicted neural data of at least two brain nodes at the next moment and the second predicted neural data of at least two brain nodes at the next moment, it is also possible to determine the strength information of the first predicted neural data of the second brain node at moment t+1 and simultaneously determine the strength information of the second predicted neural data of the second brain node at moment t+1; then, based on the strength information of the first predicted neural data and the strength information of the second predicted neural data, determine the effective connection strength.

[0135] It can be understood that in the embodiment of the present application, at the tth moment, disturbance data is added to the first brain node, and the first predicted neural data of the second brain node at the t+1th moment is predicted by combining the brain neural data of the first brain node and the second brain node from the tpth moment to the tth moment; at the same time, the second predicted neural data of the second brain node at the t+1th moment is directly predicted based on the brain neural data of the first brain node and the second brain node from the tpth moment to the tth moment; the strength information of the first predicted neural data and the strength information of the second predicted neural data are determined respectively, and then the effective connection strength is further determined based on the strength information of the two predicted neural data. For example, the strength information of the first predicted neural data and the strength information of the second predicted neural data can be differenced, and the effective connection strength can be determined according to the difference result.

[0136] Further, in an embodiment of the present application, after respectively determining the strength information of the first predicted neural data and the strength information of the second predicted neural data and comparing the two, if the comparison result is that the strength information of the first predicted neural data and the strength information of the second predicted neural data are equal, it can be considered that regardless of whether disturbance data is added to the first brain node, the predicted neural data corresponding to the second brain node will not change, that is, the disturbance data added to the first brain node will not affect the second brain node. At this time, it can be considered that there is no effective brain connection information between the first brain node and the second brain node.

[0137] For example, in some embodiments, assuming that brain node i is the first brain node and brain node j is the second brain node, then, based on the above formulas (3) and (5), the brain effective connection information EC between the two brain nodes is determined by the digital twin brain model. ij The process can be expressed as follows:

[0138] EC ij =f(x (t-p) , x (t-p+1) , ..., x t +P,θ) j -f(x(t-p) , x (t-p+1) , ..., x t ,θ) j (6);

[0139] For example, in some embodiments, assuming that brain node i is the first brain node and brain node j is the second brain node, then, based on the above formulas (3) and (5), the brain effective connection information EC between the two brain nodes is determined by the digital twin brain model. ij The process can also be expressed as follows:

[0140] EC ij =E t [f(x (t-p) , x (t-p+1) , ..., x t +P,θ) j -f(x (t-p) , x (t-p+1) , ..., x t ,θ) j ] (7);

[0141] Among them, E t (·) indicates averaging over time t.

[0142] Furthermore, in an embodiment of the present application, each of the at least two brain nodes can be traversed, so that the effective brain connection information between any two brain nodes of the at least two brain nodes can be determined, and then the whole-brain effective connection of the at least two brain nodes can be obtained.

[0143] That is to say, in an embodiment of the present application, virtual disturbance can be applied to each brain node, that is, disturbance data is added as input information, and then the brain neural data of other brain nodes are predicted respectively through the digital twin brain model, and the brain neural data of other brain nodes predicted when no virtual disturbance is applied are combined to further determine the effective brain connection information between any two brain nodes.

[0144] It can be understood that in the embodiment of the present application, after determining the effective brain connection information between every two brain nodes, the determination of the effective connection of the whole brain can be further completed.

[0145] For example, in some embodiments, Figure 4 Schematic diagram of the derivation of effective brain connection information proposed in the embodiment of the present application Figure 1 ,like Figure 4As shown, assuming that brain node a is the first brain node, after applying virtual disturbance to brain node a, the brain neural data of brain node a at the tth moment is superimposed with the corresponding disturbance data, and the brain neural data of brain node b, brain node c and brain node d at the tth moment are combined to obtain the perturbed brain neural data of these four brain nodes at the tth moment as the input of the digital twin brain model; at the same time, the brain neural data of these four brain nodes from the tpth moment to the t-1th moment can also be used as the input of the digital twin brain model. Through the digital twin brain model, the brain neural data of the four brain nodes of brain node a, brain node b, brain node c and brain node d at the next moment can be predicted, that is, the first predicted neural data obtained by prediction can include the first predicted neural data of the four brain nodes of brain node a, brain node b, brain node c and brain node d at the t+1th moment. At the same time, the brain neural data of the four brain nodes from the tpth moment to the tth moment can also be used as the input of the digital twin brain model, and the second predicted neural data can be obtained through the prediction of the digital twin brain model, wherein the second predicted neural data can include the second predicted neural data of the four brain nodes, namely, brain node a, brain node b, brain node c and brain node d, at the t+1th moment. By comparing the first predicted neural data and the second predicted neural data of brain node b, brain node c and brain node d at the t+1th moment respectively, it can be determined that the brain neural data of brain node b and brain node c have not changed, while the brain neural data of brain node d has been enhanced, and then it can be determined that the effective brain connection information from brain node a to brain node d includes: the effective connection direction is from brain node a to brain node d, the effective connection type is the excitatory type, and the effective connection strength is the strength change value of the brain neural data of brain node d.

[0146] For example, in some embodiments, Figure 5 Schematic diagram of the derivation of effective brain connection information proposed in the embodiment of the present application Figure 2 ,like Figure 5As shown, assuming that brain node b is the first brain node, after applying virtual disturbance to brain node b, the brain neural data of brain node b at the tth moment is superimposed with the corresponding disturbance data, and the brain neural data of brain node a, brain node c and brain node d at the tth moment are combined to obtain the perturbed brain neural data of these four brain nodes at the tth moment as the input of the digital twin brain model; at the same time, the brain neural data of these four brain nodes from the tpth moment to the t-1th moment can also be used as the input of the digital twin brain model. Through the digital twin brain model, the brain neural data of the four brain nodes of brain node a, brain node b, brain node c and brain node d at the next moment can be predicted, that is, the first predicted neural data obtained by prediction can include the first predicted neural data of the four brain nodes of brain node a, brain node b, brain node c and brain node d at the t+1th moment. At the same time, the brain neural data of the four brain nodes from the tpth moment to the tth moment can also be used as the input of the digital twin brain model, and the second predicted neural data can be obtained through the prediction of the digital twin brain model, wherein the second predicted neural data can include the second predicted neural data of the four brain nodes of brain node a, brain node b, brain node c and brain node d at the t+1th moment. By comparing the first predicted neural data and the second predicted neural data of brain node a, brain node c and brain node d at the t+1th moment respectively, it can be determined that the brain neural data of brain node d has not changed, while the brain neural data of brain node a has weakened, and the brain neural data of brain node c has strengthened, and then it can be determined that the effective brain connection information from brain node b to brain node a includes: the effective connection direction is from brain node b to brain node a, the effective connection type is inhibition type, and the effective connection strength is the intensity change value of the brain neural data of brain node a, and at the same time, it is determined that the effective brain connection information from brain node b to brain node c includes: the effective connection direction is from brain node b to brain node c, the effective connection type is excitation type, and the effective connection strength is the intensity change value of the brain neural data of brain node c.

[0147] For example, in some embodiments, Figure 6 Schematic diagram of the derivation of effective brain connection information proposed in the embodiment of the present application Figure 3 ,like Figure 6As shown, assuming that brain node c is the first brain node, after applying virtual disturbance to brain node c, the brain neural data of brain node c at the tth moment is superimposed with the corresponding disturbance data, and the brain neural data of brain node b, brain node a and brain node d at the tth moment are combined to obtain the perturbed brain neural data of these four brain nodes at the tth moment as the input of the digital twin brain model; at the same time, the brain neural data of these four brain nodes from the tpth moment to the t-1th moment can also be used as the input of the digital twin brain model. Through the digital twin brain model, the brain neural data of the four brain nodes of brain node a, brain node b, brain node c and brain node d at the next moment can be predicted, that is, the first predicted neural data obtained by prediction can include the first predicted neural data of the four brain nodes of brain node a, brain node b, brain node c and brain node d at the t+1th moment. At the same time, the brain neural data of the four brain nodes from the tpth moment to the tth moment can also be used as the input of the digital twin brain model, and the second predicted neural data can be obtained through the prediction of the digital twin brain model, wherein the second predicted neural data can include the second predicted neural data of the four brain nodes, namely, brain node a, brain node b, brain node c and brain node d, at the t+1th moment. By comparing the first predicted neural data and the second predicted neural data of brain node b, brain node a and brain node d at the t+1th moment respectively, it can be determined that the brain neural data of brain node b and brain node d have not changed, while the brain neural data of brain node a has weakened, and then it can be determined that the effective brain connection information from brain node c to brain node a includes: the effective connection direction is from brain node c to brain node a, the effective connection type is inhibition type, and the effective connection strength is the strength change value of the brain neural data of brain node a.

[0148] For example, in some embodiments, Figure 7 Schematic diagram of the derivation of effective brain connection information proposed in the embodiment of the present application Figure 4 ,like Figure 7As shown, assuming that brain node d is the first brain node, after applying virtual disturbance to brain node d, the brain neural data of brain node d at the tth moment is superimposed with the corresponding disturbance data, and the brain neural data of brain node b, brain node a and brain node c at the tth moment are combined to obtain the perturbed brain neural data of these four brain nodes at the tth moment as the input of the digital twin brain model; at the same time, the brain neural data of these four brain nodes from the tpth moment to the t-1th moment can also be used as the input of the digital twin brain model. Through the digital twin brain model, the brain neural data of the four brain nodes of brain node a, brain node b, brain node c and brain node d at the next moment can be predicted, that is, the first predicted neural data obtained by prediction can include the first predicted neural data of the four brain nodes of brain node a, brain node b, brain node c and brain node d at the t+1th moment. At the same time, the brain neural data of the four brain nodes from the tpth moment to the tth moment can also be used as the input of the digital twin brain model, and the second predicted neural data can be obtained through the prediction of the digital twin brain model, wherein the second predicted neural data can include the second predicted neural data of the four brain nodes, namely, brain node a, brain node b, brain node c and brain node d, at the t+1th moment. By comparing the first predicted neural data and the second predicted neural data of brain node b, brain node a and brain node c at the t+1th moment respectively, it can be determined that the brain neural data of brain node b and brain node a have not changed, while the brain neural data of brain node c has weakened, and then it can be determined that the effective brain connection information from brain node d to brain node c includes: the effective connection direction is from brain node d to brain node c, the effective connection type is inhibition type, and the effective connection strength is the strength change value of the brain neural data of brain node c.

[0149] For example, in some embodiments, Figure 8 Schematic diagram of the derivation of effective brain connection information proposed in the embodiment of the present application Figure 5 ,like Figure 8 As shown, after traversing brain node a, brain node b, brain node c and brain node d according to the above method, the effective brain connection information between every two brain nodes when virtual disturbances are applied to different brain nodes can be determined, and then the determination of the whole-brain effective connection can be completed, wherein the whole-brain effective connection can include the effective connection type, effective connection direction and effective connection strength between different brain nodes.

[0150] Furthermore, in an embodiment of the present application, since the time series length corresponding to the time series is greater than or equal to the time length of the brain neural data used in the prediction process, it is possible to select a digital twin brain model to predict the effective brain connection information at multiple moments, and perform mean calculation based on the multiple brain effective connection information obtained by prediction to determine the effective connection in the average sense, which can more accurately and comprehensively reflect the activity between brain nodes.

[0151] Furthermore, in an embodiment of the present application, a digital twin brain model can be used to first obtain a plurality of first predicted neural data corresponding to a plurality of moments of the at least two brain nodes based on a plurality of disturbance data corresponding to a plurality of moments of the first brain node and brain neural data of at least two brain nodes in a time series; at the same time, a plurality of second predicted neural data corresponding to a plurality of moments of the at least two brain nodes can be obtained based on the brain neural data of at least two brain nodes in a time series through the digital twin brain model; then, based on the plurality of first predicted neural data and the plurality of second predicted neural data, a plurality of brain effective connection information corresponding to a plurality of moments from the first brain node to the second brain node can be determined; finally, based on the plurality of brain effective connection information, the target effective connection information from the first brain node to the second brain node can be determined.

[0152] It can be understood that, in the embodiment of the present application, the first predicted neural data is the brain neural data predicted after adding the virtual disturbance, and the second predicted neural data is the brain neural data predicted without adding the virtual disturbance.

[0153] It can be understood that in the embodiments of the present application, the average calculation is performed based on multiple brain effective connection information corresponding to multiple moments from the first brain node to the second brain node, and the average result finally obtained is the target effective connection information from the first brain node to the second brain node, and the target effective connection information can characterize the effective connection from the first brain node to the second brain node in an average sense.

[0154] That is to say, in the embodiments of the present application, for a sufficiently large time series length, a plurality of optional parameters t can be selected based on the brain neural data in the time series to predict the brain neural data, that is, a plurality of predicted neural data are predicted respectively, including a plurality of first predicted neural data obtained when the disturbance data is added corresponding to a plurality of moments and a plurality of second predicted neural data obtained when the disturbance data is not added, and finally the effective connection in the average sense is determined using the plurality of first predicted neural data and the plurality of second predicted neural data corresponding to a plurality of moments. Among them, for a moment t, it can be considered that the effective connection obtained at a given t depends on the current state, and for a plurality of moments t, the effective connection is calculated and averaged under all optional t, and the effective connection in the average sense (target effective connection information) can be obtained.

[0155] Exemplarily, in some embodiments, it is assumed that the digital twin brain predicts 1 step in 3 steps (using brain neural data at three consecutive moments to predict brain neural data at the next moment), that is, p=2, the preset time length is 3, and the brain neural data in the time series includes brain neural data at five moments from moment 1 to moment 5, that is, the time series length is 5. Then, the effective brain connection information at the three moments of moment 4, moment 5, and moment 6 can be obtained according to the brain neural data at moment 1, moment 2, and moment 3, the brain neural data at moment 2, moment 3, and moment 4, and the brain neural data at moment 3, moment 4, and moment 5, respectively, and the obtained results depend on the brain states represented by different moments, and the results obtained at different moments are not exactly the same. Finally, the effective brain connection information at the three moments of moment 4, moment 5, and moment 6 can be further used for mean calculation, and the average result is used as the effective connection in the average sense, that is, the target effective connection information.

[0156] In summary, through the derivation method of the brain's effective connection information proposed in the above steps 101 to 104, the digital twin brain model can be stimulated by different virtual stimuli, and the digital twin brain model can be virtually perturbed to predict the brain's response under stimulation and infer the effective connection of the whole brain. In this way, a large number of brain areas can be considered at the same time, which is suitable for inferring the effective connection of the whole brain, and can simultaneously characterize the connection's sign (effective connection type), weight (effective connection strength) and direction (effective connection direction).

[0157] An embodiment of the present application provides a method for deriving effective brain connection information. The derivation device for deriving effective brain connection information determines the disturbance data of a first brain node among at least two brain nodes and the brain neural data of at least two brain nodes in a time series; through a digital twin brain model, based on the disturbance data of the first brain node and the brain neural data of at least two brain nodes in a time series, obtains the first predicted neural data of at least two brain nodes at the next moment; wherein, the digital twin brain model is obtained by training based on a time series prediction network; based on the first predicted neural data of at least two brain nodes at the next moment, and the second predicted neural data of at least two brain nodes at the next moment, determines the effective brain connection information from the first brain node to the second brain node; wherein, the second brain node is the other brain node among the at least two brain nodes except the first brain node; the second predicted neural data represents the predicted data obtained when no disturbance data is added. That is to say, in an embodiment of the present application, based on the digital twin brain model obtained through training, disturbance data is added to the brain nodes to predict the brain's response under disturbance stimulation and infer the effective connection information of the brain. The digital twin brain model can effectively and accurately predict the dynamic changes of brain neural data, thereby obtaining more accurate effective connection information of the brain.

[0158] Another embodiment of the present application provides a training method for a digital twin brain model, wherein the training method for the digital twin brain model can be applied to a training device or computer device for the digital twin brain model, and the present application does not specifically limit it. Below, taking the training device for the digital twin brain model as an example, the training method for the digital twin brain model proposed in the embodiment of the present application is exemplarily described.

[0159] Furthermore, in the embodiments of the application, Fig. 9 This is a schematic diagram of the implementation process of the training of the digital twin brain model proposed in the embodiment of the present application, such as Fig. 9 As shown, the training method of the digital twin brain model may include the following steps:

[0160] Step 901: Train the time series prediction network based on the training data set to obtain a digital twin brain model; wherein the digital twin brain model is used to predict the neural data at the next moment based on the brain neural data of the brain nodes in the time series.

[0161] In an embodiment of the present application, the training device of the digital twin brain model can use neural training data including at least two brain nodes in a time series to train the time series prediction network, so as to obtain a digital twin brain model. Among them, the digital twin brain model obtained based on the time series prediction network training can predict the dynamic changes of neural data, that is, it can predict the neural data of the next moment based on the brain neural data of the brain node in the time series, so that the effective brain connection information between the brain nodes can be further determined based on the obtained predicted neural data, thereby obtaining more accurate brain effective connection information.

[0162] It can be understood that in the embodiments of the present application, a brain node can represent a brain region, that is, it can be considered that one brain node corresponds to one brain region. The brain can be divided into at least two brain regions, and accordingly, one brain can correspond to at least two brain nodes.

[0163] For example, in some embodiments, Fig.10 Schematic diagram of the brain node proposed in the embodiment of the present application Figure 2 ,like Fig.10 As shown, assuming that the biological brain is divided into multiple regions, such as region A, region B and region C, accordingly, it can be considered that the brain at least corresponds to brain node a, brain node b and brain node c.

[0164] Furthermore, in an embodiment of the present application, the neural training data of at least two brain nodes in a time series may include the neural training data of each brain node corresponding to the time series, that is, the training device of the digital twin brain model needs to first obtain the neural training data of multiple moments corresponding to each brain node.

[0165] In the embodiments of the present application, the time series may include continuous time sampling intervals or discrete time sampling points, which is not specifically limited in the present application.

[0166] Further, in an embodiment of the present application, the training data set may include brain neural data corresponding to a preset time length p+1 of at least two brain nodes; wherein, when predicting the brain neural data at the next moment, the preset time length may determine the time length of the brain neural data used in the prediction process. wherein p is an integer greater than or equal to 0.

[0167] It is understood that in the embodiment of the present application, the time length of the brain neural data used in the prediction process, that is, the preset time length p+1, can be preset. For example, if p=2 is preset, then the value of the preset time length p+1 is 3.

[0168] It can be understood that in an embodiment of the present application, the training data set may include brain neural data corresponding to at least two brain nodes at multiple moments, and the brain neural data at multiple moments include but are not limited to brain neural data corresponding to at least two brain nodes for a preset time length p+1, that is, the time series length corresponding to the training data set is greater than or equal to the time length of the brain neural data used in the prediction process.

[0169] Exemplarily, in some embodiments, the time series length corresponding to the training data set is 5000, while the time length of the brain neural data used in the prediction process is 3.

[0170] It can be understood that in the embodiments of the present application, for a determined preset time length, the time length of the data used for brain neural data prediction is fixed. For example, for the preset time length p+1, the brain neural data of the previous p+1 moments can be used to predict the brain neural data of the next moment.

[0171] That is to say, in the embodiments of the present application, the brain nerve data at one or more moments can be used to predict the brain nerve data at the next moment. For example, if p=2 is preset, the brain nerve data at three consecutive moments, namely, moment 1, moment 2, and moment 3, can be used to predict the brain nerve data at moment 4, the brain nerve data at three consecutive moments, namely, moment 2, moment 3, and moment 4, can be used to predict the brain nerve data at moment 5, and the brain nerve data at three consecutive moments, namely, moment 3, moment 4, and moment 5, can be used to predict the brain nerve data at moment 6.

[0172] Exemplarily, in some embodiments, assuming that the preset time length is p+1, the time length indicated by the preset time length includes p+1 moments from the mpth moment to the mth moment. Accordingly, the neural training data corresponding to the preset time length p+1 includes the neural training data from the mpth moment to the mth moment, where m is an integer greater than p.

[0173] Furthermore, in an embodiment of the present application, the training data set also includes neural training data of at least two brain nodes at the (m+1)th moment.

[0174] Furthermore, in the embodiments of the present application, a variety of methods can be used to determine the neural training data of any brain node in a time series, and the present application does not make specific limitations.

[0175] Exemplarily, in some embodiments, neural training data can be acquired through existing bioinformatics databases, wherein the bioinformatics database includes but is not limited to one or more public datasets such as human fMRI datasets, EEG datasets, and MEG datasets.

[0176] Exemplarily, in some embodiments, neural training data may also be generated through a computational model.

[0177] Furthermore, in an embodiment of the present application, the time series prediction network used to train the digital twin brain model can be an artificial neural network of any type and structure for time series prediction, and the present application does not make any specific limitations.

[0178] Exemplarily, in some embodiments, the time series prediction network may use classic models such as multi-layer perceptron, convolutional neural network, recurrent neural network, etc., or other network architectures suitable for time series prediction may be selected, which is not specifically limited in this application.

[0179] Furthermore, in an embodiment of the present application, when the time series prediction network is trained based on the training data set to obtain a digital twin brain model, the time series prediction network can be used to obtain prediction data of at least two brain nodes at the m+1th moment based on the neural training data of at least two brain nodes at the mpth moment to the mth moment; based on the prediction data of at least two brain nodes at the m+1th moment and the neural training data of at least two brain nodes at the m+1th moment, the time series prediction network is corrected to obtain the digital twin brain model.

[0180] For example, in some embodiments, when training a time series prediction network based on a training data set, the prediction data of at least two brain nodes at the m+1th time are determined by the digital twin brain model. The process can be referred to the following formula (8):

[0181]

[0182] Among them, the time series prediction network can be expressed as f(·), x (m-p) , x (m-p+1) , ..., x m The neural training data from time point (mp) to time point m can be represented.

[0183] For example, in some embodiments, when training a time series prediction network based on a training data set, the prediction data of at least two brain nodes at the m+1th time are determined by the digital twin brain model. The process can be referred to the following formula (9):

[0184]

[0185] Among them, θ represents all parameters in the time series prediction network that can be optimized through training iterations.

[0186] Furthermore, in an embodiment of the present application, when the time series prediction network is corrected based on the prediction data of at least two brain nodes at the m+1th moment and the neural training data of at least two brain nodes at the m+1th moment to obtain a digital twin brain model, model parameters are determined based on the prediction data of at least two brain nodes at the m+1th moment, the neural training data of at least two brain nodes at the m+1th moment and the preset time length corresponding to the training data set; and the digital twin brain model is determined based on the model parameters.

[0187] It can be understood that in the embodiments of the present application, after the neural data is predicted based on the input neural training data through the time series prediction network and the predicted data for the next moment is obtained, the neural training data for the next moment can be further combined to complete the training of the time series prediction network by minimizing the loss function to obtain the corresponding model parameters, and finally the model parameters can be used to obtain the corresponding digital twin brain model.

[0188] In an embodiment of the present application, the loss function may be defined as any type of objective function. For example, the loss function may be defined by minimizing the mean square error of time series prediction. This application does not make any specific limitation.

[0189] For example, in some embodiments, it is assumed that the time series prediction network is trained by minimizing the mean square error of the time series prediction, that is, the loss function is defined as follows:

[0190]

[0191] Wherein, T represents the number of prediction data corresponding to multiple moments of the at least two brain nodes obtained based on the training data set. By minimizing the loss function L(θ) using optimization methods such as stochastic gradient descent and adaptive learning rate algorithm, the network parameters of the time series prediction network suitable for describing the dynamics of large-scale brain networks can be obtained, that is, the model parameters are obtained, and then based on the model parameters, an artificial neural network that can be used as a digital twin brain model can be obtained, that is, a digital twin brain model is obtained.

[0192] It can be understood that in the embodiments of the present application, the digital twin brain model is obtained based on time series prediction network training. Therefore, the digital twin brain model can effectively predict the dynamic changes of brain neural signals (brain neural data), and thus obtain more accurate brain effective connection information.

[0193] Furthermore, in an embodiment of the present application, since the time series length corresponding to the training data set is greater than or equal to the time length of the brain neural data used in the prediction process, it is possible to select a time series prediction network to predict data at multiple moments, and perform difference calculations based on the multiple predicted data obtained and the multiple neural training data at the corresponding moments, and determine the model parameters by comprehensively considering the multiple errors obtained by calculation, thereby obtaining a digital twin brain model with better prediction performance.

[0194] Furthermore, in an embodiment of the present application, when a time series prediction network is trained based on a training data set to obtain a digital twin brain model, a plurality of prediction data corresponding to at least two brain nodes at multiple moments can be obtained through the time series prediction network based on the training data set; a plurality of data errors corresponding to multiple moments are determined based on the plurality of prediction data and neural training data of at least two brain nodes corresponding to multiple moments; model parameters are determined based on the plurality of data errors; and a digital twin brain model is determined based on the model parameters.

[0195] That is, in the embodiment of the present application, for a sufficiently large time series length, multiple optional parameters t can be selected based on the neural training data in the training data set to predict neural data, that is, multiple prediction data are predicted separately, and finally, multiple prediction data corresponding to multiple moments and multiple neural training data are used to perform difference calculation to determine multiple data errors (multiple errors) corresponding to multiple moments. These data errors are comprehensively considered to determine the model parameters.

[0196] Exemplarily, in some embodiments, it is assumed that the digital twin brain predicts 1 step in 3 steps (using the neural training data of three consecutive moments to predict the brain neural data of the next moment), that is, p=2, the preset time length is 3, and the neural training data in the training data set includes the neural training data of the five moments from moment 1 to moment 5, that is, the time series length is 5. Then, the prediction data of moments 4 and 5 can be obtained according to the neural training data of moments 1, 2, and 3, and the neural training data of moments 2, 3, and 4, respectively, and then the error between the prediction data and the neural training data at moment 4 and the error between the prediction data and the neural training data at moment 5 are calculated respectively, and the data error of moments 4 and 5 is obtained. Finally, the model parameters are determined based on these two data errors.

[0197] Furthermore, in the embodiments of the present application, in order to evaluate the effect of the digital twin brain model on the characterization of large-scale brain network dynamics, the performance of the trained digital twin brain model can be further measured and evaluated.

[0198] Furthermore, in an embodiment of the present application, when measuring and evaluating the performance of the digital twin brain model obtained through training, the determination coefficient corresponding to the digital twin brain model can be determined based on the training data set and the predicted data set corresponding to the training data set; through the digital twin brain model, based on the brain neural data and random noise, the first functional connection matrix predicted by the digital twin brain model is determined, and the matrix correlation coefficient between the first functional connection matrix and the second functional connection matrix corresponding to the training data set is determined; based on the determination coefficient and the matrix correlation coefficient, the model performance parameters of the digital twin brain model are determined.

[0199] It can be understood that in an embodiment of the present application, the training data set may include neural training data of at least two brain nodes corresponding to T moments, and corresponding to the training data set, the prediction data set may include prediction data of at least two brain nodes corresponding to T moments, that is, the prediction data set may include prediction data of T moments corresponding to the neural training data of T moments.

[0200] Furthermore, when determining the determination coefficient corresponding to the digital twin brain model based on the training data set and the predicted data set corresponding to the training data set, the data mean of at least two brain nodes corresponding to T moments is determined based on the neural training data of at least two brain nodes corresponding to T moments; the determination coefficient corresponding to the digital twin brain model is determined based on the data mean, the neural training data corresponding to T moments and the predicted data corresponding to T moments.

[0201] For example, in some embodiments, the determination coefficient R introduced for evaluating the trained digital twin brain model is 2 , can be calculated according to the following formula (11):

[0202]

[0203] Where N represents the number of nodes in the large-scale brain network (number of brain nodes), x i，t represents the value of node i at time point t in the brain neural data, represents the value of node i at time point t predicted by the digital twin brain model, represents the mean value of node i in the brain neural data at all time points, and T represents the length of the brain neural data used to train the time series prediction network.

[0204] It can be understood that in the embodiments of the present application, the determination coefficient R obtained by calculation is 2 , the performance of the digital twin brain model can be evaluated. The better the prediction effect of the digital twin brain model on the brain neural data, the better the digital twin brain model can characterize the state transition relationship of the brain network.

[0205] Further, in an embodiment of the present application, the digital twin brain model can be used to determine the first functional connection matrix predicted by the digital twin brain model based on brain neural data and random noise. At the same time, the second functional connection matrix can be predicted based on the training data set, that is, the second functional connection matrix corresponding to the training data set is determined, and then the matrix correlation coefficient can be further determined based on the first functional connection matrix and the second functional connection matrix. Among them, the matrix correlation coefficient can be used to determine the correlation between the first functional connection matrix and the second functional connection matrix.

[0206] It can be understood that in the embodiments of the present application, the functional connection matrix is ​​a correlation matrix between different brain node activities. Among them, the first functional connection matrix can be understood as a correlation matrix between different brain node activities predicted by the digital twin brain model obtained by training; and the second functional connection matrix can be understood as a correlation matrix between different brain node activities calculated based on the training data set.

[0207] Furthermore, in an embodiment of the present application, when determining the first functional connectivity matrix based on brain neural data and random noise, a digital twin brain model can be used to generate data driven by random noise to generate the corresponding first functional connectivity matrix.

[0208] For example, in some embodiments, all zero data may be used as the initial value (ie, x 0..p =0), through the formula x t+1 =f(x (t-p)..t +δ (t-p)..t ,θ) to generate the next step of simulation data in a loop until the length of the simulation data is consistent with the length of the brain neural data used to train the model. t is random noise with a mean of 0 and a normal distribution.

[0209] It can be understood that in the embodiment of the present application, the functional connection matrix is ​​reconstructed by the digital twin brain model, and then the matrix correlation coefficient between the first functional connection matrix reconstructed by the digital twin brain model and the second functional connection matrix corresponding to the training data set is further determined, and the matrix correlation coefficient can determine the reconstruction effect of the functional connection matrix. Among them, the better the reconstruction effect of the functional connection matrix, the better the digital twin brain model can portray the functional dependency relationship between brain nodes.

[0210] That is to say, in the embodiments of the present application, after the digital twin brain model is trained, in order to evaluate the effect of the digital twin brain model on the characterization of large-scale brain network dynamics, the determination coefficient and matrix correlation coefficient can be introduced to measure and evaluate the performance of the digital twin brain model. For example, the correlation coefficient (matrix correlation coefficient) between the determination coefficient predicted by brain neural data, the functional connection matrix of the data generated by the digital twin brain model under random noise drive, and the functional connection matrix of the brain neural data (training data set) used to train the model is used to measure the effect of the digital twin brain model.

[0211] Furthermore, in an embodiment of the present application, after the digital twin brain model is completed, the digital twin brain model can be used to further determine the effective brain connection information between different brain nodes according to the derivation method of the effective brain connection information proposed in the above embodiment while adding virtual disturbances, thereby obtaining the whole-brain effective connection of at least two brain nodes.

[0212] In summary, through the training method of the digital twin brain model proposed in the above step 901, an artificial neural network (such as a time series prediction network) is used to learn and predict the dynamic changes of brain neural signals, the artificial neural network is used as a digital twin brain model, and the digital twin brain model is trained by reconstructing or predicting brain neural data. The prediction error and the functional connection of the generated data can also be used as indicators to evaluate the digital twin brain model. This can avoid relying on specific model assumptions and can be flexibly applied to neural signals of different modalities without a lot of adjustments.

[0213] The embodiment of the present application provides a training method for a digital twin brain model, wherein the training device of the digital twin brain model trains a time series prediction network based on a training data set to obtain a digital twin brain model; wherein the digital twin brain model is used to predict the neural data of the next moment based on the brain neural data of the brain node in the time series; the training data set includes neural training data of at least two brain nodes in the time series. That is to say, in an embodiment of the present application, a time series prediction network can be used to learn and predict the dynamic changes of brain neural signals to obtain a digital twin brain model with a time series prediction function. wherein, the digital twin brain model can effectively and accurately predict the dynamic changes of brain neural data.

[0214] Based on the above embodiments, another embodiment of the present application proposes a method for deriving effective brain connection information, and at the same time proposes a method for training a digital twin brain model. On the one hand, the training method of the digital twin brain model can use an artificial neural network to fit large-scale neural dynamics, construct a digital twin brain model, and characterize the functional dependencies between different brain regions. On the other hand, the method for deriving effective brain connection information, based on the digital twin brain model obtained through training, applies virtual stimulation to the digital twin brain model, and observes the response of other regions to the stimulation to characterize the effective connection between brain nodes. It can predict the response pattern of the brain in different states, so that the symbol (effective connection type), weight (effective connection strength) and direction (effective connection direction) of the effective connection can be characterized throughout the brain.

[0215] Fig.11 This is a schematic diagram of neural data analysis under real stimulation, such as Fig.11 As shown, in general, effective connectivity can be assessed through neural stimulation experiments, which perturb specific brain regions through real stimulation, monitor neural responses in other regions, and simultaneously record neural data to directly prove causality. However, this approach limits the scalability of whole-brain analysis due to its invasiveness.

[0216] Furthermore, in an embodiment of the present application, an artificial neural network may be used to construct a digital twin brain model, wherein the artificial neural network may be a time series prediction network.

[0217] Exemplarily, in some embodiments, the large-scale brain network dynamics is expressed as the following formula (12) using an artificial neural network f(·):

[0218]

[0219] The artificial neural network f(·) can use classic models such as multi-layer perceptron, convolutional neural network, recurrent neural network, etc., or other network architectures suitable for time series prediction. θ represents all parameters in the artificial neural network that can be optimized through training iterations, x (t-p)..t represents the brain neural data from time point (tp) to time point t, Represents the brain neural data at the time point (t+1) predicted by the artificial neural network, and the value of the hyperparameter p is determined by the properties of the brain neural data.

[0220] Exemplarily, in some embodiments, the artificial neural network is trained by minimizing the mean square error of time series prediction, that is, the loss function is defined by referring to formula (10). Wherein, T represents the length of brain neural data used to train the artificial neural network. By minimizing the loss function L(θ) using optimization methods such as stochastic gradient descent and adaptive learning rate algorithm, the artificial neural network parameters suitable for describing the dynamics of large-scale brain networks can be obtained, and the artificial neural network that can be used as a digital twin brain model can be obtained.

[0221] Furthermore, in an embodiment of the present application, after the digital twin brain model is obtained through training, the digital twin brain model can be evaluated using the prediction error and the functional connectivity of the generated data.

[0222] It can be understood that in the embodiments of the present application, in order to evaluate the effect of the digital twin brain model on the characterization of large-scale brain network dynamics, the determination coefficient of the brain neural data prediction, the functional connection matrix of the data generated by the digital twin brain model under random noise drive, and the correlation coefficient between the functional connection matrix of the brain neural data (training data set) used to train the model can be used to measure the effect of the alternative model.

[0223] For example, in some embodiments, the coefficient of determination R of the brain neural data prediction is 2 It can be calculated by formula (12), where N represents the number of nodes in the large-scale brain network, x i,t represents the value of node i at time point t in the brain neural data, represents the value of node i at time point t predicted by the digital twin brain model, Represents the mean value of node i in the brain neural data at all time points. The better the prediction effect of the digital twin brain model on the brain neural data, the better the alternative model can describe the state transition relationship of the brain network.

[0224] Exemplarily, in some embodiments, the functional connectivity matrix is ​​a correlation matrix between different node activities. When using the digital twin brain model to generate data under random noise drive, all zero data can be used as the initial value (i.e., x 0..p =0), through the formula x t+1 =f(x (t-p)..t +δ (t-p)..t ,θ) to generate the next step of simulation data in a loop until the length of the simulation data is consistent with the length of the brain neural data used to train the model. t It is a random noise that follows a normal distribution with a mean of 0. The better the reconstruction effect of the digital twin brain model on the functional connection matrix is, the better the alternative model can characterize the functional dependency relationship between nodes.

[0225] Furthermore, in an embodiment of the present application, after training to obtain a digital twin brain model with time series prediction function, the effective connections of the whole brain can be inferred by virtually perturbing the digital twin brain model.

[0226] It is understandable that after obtaining and verifying the effectiveness of the digital twin brain model, the effective connection of the brain network can be obtained by applying virtual perturbations to the model. When characterizing the effective connection of a given node to other nodes, the virtual perturbation can be expressed as a slight increase in the signal of a given node at the last time point of the perturbed brain neural data. By observing the impact of applying virtual perturbations on the output of the alternative model, the effective connection of the node to other nodes can be characterized.

[0227] For example, in some embodiments, the effective connection EC from node i to node j is ij It can be calculated by formula (7), where E t (·) represents the mean over time t, f(·) j represents the component j of the surrogate model output, and the virtual perturbation P is consistent with p i =Δ, p k，k≠i =0 or other N-dimensional virtual perturbation vectors required by traditional experimental paradigms, where Δ is the intensity of the virtual perturbation applied to node i.

[0228] Further, in an embodiment of the present application, the effective connection of the whole brain can be obtained by traversing all nodes i and j. Among them, the effective connection obtained by the derivation method of the brain effective connection information proposed in the present application can be the brain effective connection information with a sign (effective connection type), a weight (effective connection strength) and a direction (effective connection direction), which can be used to guide precise neural regulation.

[0229] For example, in some embodiments, Fig.12 This is a schematic diagram of the implementation of the present application embodiment for deducing effective brain connection information based on the digital twin brain model obtained through training, as shown in FIG. Fig.12As shown, during the training process of the digital twin brain model, the number of model nodes can correspond to the number of brain nodes. For example, there can be four model nodes corresponding to the four brain nodes, namely, brain node a, brain node b, brain node c and brain node d; the training data set used for model training can include neural training data of brain nodes in a time series, for example, resting-state functional magnetic resonance imaging data from time t to time t+Δt; the neural training data x(t) at time t is used as input to predict the predicted data x(t+Δt) at time t+Δt, and combined with the neural training data at time t+Δt, the model can be iterated and corrected to finally obtain a digital twin brain (digital twin brain model). After the digital twin brain model is trained, the digital twin brain model is used to combine the virtual disturbances (disturbance data) applied to different brain nodes to further compare and analyze the predicted neural data obtained when the disturbance data is added and when the disturbance data is not added, so as to obtain the effective brain connection information between any two brain nodes, and finally obtain accurate whole-brain effective connections, which include effective connection type, effective connection direction and effective connection strength.

[0230] Further, the derivation method of brain effective connection information and the training method of digital twin brain model proposed in the embodiment of the present application are verified by simulation of three simulation models, which proves that the method is feasible, and the inference result of effective connection in the present application is better than the traditional data-driven method (Granger causality, dynamic causal model). The derivation method of brain effective connection information proposed in the embodiment of the present application has been successfully applied to four brain maps, and the effective connection of the whole brain under different brain maps is obtained based on resting state functional magnetic resonance imaging data, and the obtained connection is significantly different between autistic patients and normal people.

[0231] In summary, in the embodiments of the present application, on the one hand, an artificial neural network (such as a time series prediction network) is used to learn and predict the dynamic changes of brain neural signals, and the artificial neural network is used as a digital twin brain model, and the digital twin brain model is trained by reconstructing or predicting brain neural data. The prediction error and the functional connection of the generated data can also be used as indicators to evaluate the digital twin brain model, which can avoid relying on specific model assumptions and can be flexibly applied to neural signals of different modalities without a lot of adjustments. On the second hand, the digital twin brain model can be stimulated by different virtual stimuli, and the response of the brain under stimulation can be predicted by virtually perturbing the digital twin brain model and inferring the effective connection of the whole brain. In this way, a large number of brain areas can be considered at the same time, which is suitable for inferring the effective connection of the whole brain, and can simultaneously characterize the sign, weight and direction of the connection.

[0232] The embodiments of the present application provide a method for deriving effective brain connection information and a method for training a digital twin brain model. The digital twin brain model is obtained based on time series prediction network training. Therefore, the digital twin brain model can effectively and accurately predict the dynamic changes of brain neural data, thereby obtaining more accurate effective brain connection information.

[0233] Based on the above embodiment, in another embodiment of the present application, Fig.13 This is a schematic diagram of the composition structure of the device for deriving effective brain connection information proposed in the embodiment of the present application, as shown in Fig.13 As shown, the brain effective connection information derivation device 130 proposed in the embodiment of the present application may include a determination unit 1301, an acquisition unit 1302,

[0234] A determination unit 1301 is used to determine disturbance data of a first brain node among at least two brain nodes and brain nerve data of the at least two brain nodes in a time series;

[0235] An acquisition unit 1302 is used to obtain, through a digital twin brain model, first predicted neural data of the at least two brain nodes at the next moment based on the disturbance data of the first brain node and the brain neural data of the at least two brain nodes in a time series; wherein the digital twin brain model is obtained by training based on a time series prediction network;

[0236] The determination unit 1301 is further used to determine the effective brain connection information between the first brain node and the second brain node based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment; wherein the second brain node is the other brain node among the at least two brain nodes except the first brain node; and the second predicted neural data represents the predicted data obtained when the disturbance data is not added.

[0237] Furthermore, in the embodiments of the present application, Fig.14 This is a schematic diagram of the composition structure of the training device of the digital twin brain model proposed in the embodiment of the present application, as shown in Fig.14 As shown, the training device 140 of the digital twin brain model proposed in the embodiment of the present application may include a training unit 1401 and an evaluation unit 1402.

[0238] A training unit 1401 is used to train a time series prediction network based on a training data set to obtain a digital twin brain model; wherein the digital twin brain model is used to predict the neural data of the next moment based on the brain neural data of the brain node in the time series;

[0239] The training data set includes neural training data of at least two brain nodes corresponding to a preset time length p+1; wherein the neural training data corresponding to the preset time length p+1 includes neural training data from the mpth moment to the mth moment, and the training data set also includes neural training data of the at least two brain nodes at the m+1th moment; p is an integer greater than or equal to 0, and m is an integer greater than p;

[0240] The time series prediction network is trained based on the training data set to obtain a digital twin brain model, including:

[0241] Obtaining, by the time series prediction network, prediction data of the at least two brain nodes at the m+1th moment based on the neural training data of the at least two brain nodes from the mpth moment to the mth moment;

[0242] Based on the prediction data of the at least two brain nodes at the m+1th moment and the neural training data of the at least two brain nodes at the m+1th moment, the time series prediction network is corrected to obtain the digital twin brain model.

[0243] Furthermore, in an embodiment of the present application, the evaluation unit 1402 is used to determine the determination coefficient corresponding to the digital twin brain model based on the training data set and the predicted data set corresponding to the training data set; determine the first functional connection matrix predicted by the digital twin brain model based on the brain neural data and random noise through the digital twin brain model, and determine the matrix correlation coefficient between the first functional connection matrix and the second functional connection matrix corresponding to the training data set; based on the determination coefficient and the matrix correlation coefficient, determine the model performance parameters of the digital twin brain model.

[0244] In the embodiments of the present application, further, Fig.15 This is a schematic diagram of the structure of the computer device proposed in the embodiment of the present application, such as Fig.15 As shown, the computer device 150 proposed in the embodiment of the present application includes a processor 1501 and a memory 1502 storing executable instructions of the processor 1501. Furthermore, the computer device 1500 may also include a communication interface 1503 and a bus 1504 for connecting the processor 1501, the memory 1502 and the communication interface 1503.

[0245] In the embodiment of the present application, the above-mentioned processor 1501 can be at least one of an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a digital signal processor (Digital Signal Processor, DSP), a digital signal processing device (Digital Signal Processing Device, DSPD), a programmable logic device (ProgRAMmable Logic Device, PLD), a field programmable gate array (Field ProgRAMmable Gate Array, FPGA), a central processing unit (Central Processing Unit, CPU), a controller, a microcontroller, and a microprocessor. It can be understood that for different devices, the electronic device used to implement the above-mentioned processor function can also be other, and the embodiment of the present application is not specifically limited. The computer device 1500 can also include a memory 1502, which can be connected to the processor 1501, wherein the memory 1502 is used to store executable program code, the program code includes computer operation instructions, and the memory 1502 may include a high-speed RAM memory, and may also include a non-volatile memory, for example, at least two disk memories.

[0246] In the embodiment of the present application, the bus 1504 is used to connect the communication interface 1503, the processor 1501 and the memory 1502, as well as the mutual communication between these devices.

[0247] In the embodiment of the present application, the memory 1502 is used to store instructions and data.

[0248] Further, in an embodiment of the present application, the processor 1501 is used to: determine the disturbance data of a first brain node among at least two brain nodes and the brain neural data of the at least two brain nodes in a time series; obtain, through a digital twin brain model, first predicted neural data of the at least two brain nodes at the next moment based on the disturbance data of the first brain node and the brain neural data of the at least two brain nodes in a time series; wherein the digital twin brain model is obtained by training based on a time series prediction network; determine the effective brain connection information from the first brain node to the second brain node based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment; wherein the second brain node is the other brain node among the at least two brain nodes except the first brain node; the second predicted neural data represents the predicted data obtained when the disturbance data is not added.

[0249] Further, in an embodiment of the present application, the processor 1501 is further used to: train a time series prediction network based on a training data set to obtain a digital twin brain model; wherein the digital twin brain model is used to predict the neural data of the next moment based on the brain neural data of the brain node in the time series;

[0250] The training data set includes neural training data of at least two brain nodes corresponding to a preset time length p+1; wherein the neural training data corresponding to the preset time length p+1 includes neural training data from the mpth moment to the mth moment, and the training data set also includes neural training data of the at least two brain nodes at the m+1th moment; p is an integer greater than or equal to 0, and m is an integer greater than p;

[0251] The time series prediction network is trained based on the training data set to obtain a digital twin brain model, including:

[0252] Obtaining, by the time series prediction network, prediction data of the at least two brain nodes at the m+1th moment based on the neural training data of the at least two brain nodes from the mpth moment to the mth moment;

[0253] Based on the prediction data of the at least two brain nodes at the m+1th moment and the neural training data of the at least two brain nodes at the m+1th moment, the time series prediction network is corrected to obtain the digital twin brain model.

[0254] In practical applications, the memory 1502 may be a volatile memory, such as a random access memory (RAM); or a non-volatile memory, such as a read-only memory (ROM), a flash memory, a hard disk (HDD) or a solid-state drive (SSD); or a combination of the above types of memories, and provide instructions and data to the processor 1501.

[0255] In addition, each functional module in this embodiment can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or software functional modules.

[0256] If the integrated unit is implemented in the form of a software function module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) or a processor to perform all or part of the steps of the method of this embodiment. The aforementioned storage medium includes: U disk, mobile hard disk, read only memory (ROM), random access memory (RAM), disk or optical disk, etc., which can store program code.

[0257] The embodiments of the present application provide a device for deriving effective brain connection information, a training device for a digital twin brain model, and a computer device. The digital twin brain model is obtained based on time series prediction network training. Therefore, the digital twin brain model can effectively and accurately predict the dynamic changes of brain neural data, thereby obtaining more accurate effective brain connection information.

[0258] Specifically, the program instructions corresponding to the method for deriving effective brain connection information in this embodiment can be stored on a storage medium such as a CD, a hard disk, or a USB flash drive. When the program instructions corresponding to the method for deriving effective brain connection information in the storage medium are read or executed by an electronic device, the steps include:

[0259] determining disturbance data of a first brain node among at least two brain nodes and brain neural data of the at least two brain nodes in a time series;

[0260] Obtaining, by means of a digital twin brain model, first predicted neural data of the at least two brain nodes at the next moment based on the disturbance data of the first brain node and the brain neural data of the at least two brain nodes in a time series; wherein the digital twin brain model is obtained by training based on a time series prediction network;

[0261] Based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment, the effective brain connection information from the first brain node to the second brain node is determined; wherein the second brain node is the other brain node among the at least two brain nodes except the first brain node; the second predicted neural data represents the predicted data obtained when the disturbance data is not added.

[0262] Specifically, the program instructions corresponding to the training method of a digital twin brain model in this embodiment can be stored on a storage medium such as a CD, a hard disk, a USB flash drive, etc. When the program instructions corresponding to the training method of a digital twin brain model in the storage medium are read or executed by an electronic device, the following steps are included:

[0263] The time series prediction network is trained based on the training data set to obtain a digital twin brain model; wherein the digital twin brain model is used to predict the neural data of the next moment based on the brain neural data of the brain node in the time series;

[0264] The training data set includes neural training data of at least two brain nodes corresponding to a preset time length p+1; wherein the neural training data corresponding to the preset time length p+1 includes neural training data from the mpth moment to the mth moment, and the training data set also includes neural training data of the at least two brain nodes at the m+1th moment; p is an integer greater than or equal to 0, and m is an integer greater than p;

[0265] The time series prediction network is trained based on the training data set to obtain a digital twin brain model, including:

[0266] Obtaining, by the time series prediction network, prediction data of the at least two brain nodes at the m+1th moment based on the neural training data of the at least two brain nodes from the mpth moment to the mth moment;

[0267] Based on the prediction data of the at least two brain nodes at the m+1th moment and the neural training data of the at least two brain nodes at the m+1th moment, the time series prediction network is corrected to obtain the digital twin brain model.

[0268] The embodiment of the present application also provides a computer program product.

[0269] In some embodiments, the computer program product may include a computer program or instructions.

[0270] In some embodiments, the computer program product can be applied to the computer device in the embodiments of the present application, and the computer program instructions enable the computer to execute the corresponding processes implemented by the computer device in the various methods of the embodiments of the present application. For the sake of brevity, they will not be repeated here.

[0271] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of hardware embodiments, software embodiments, or embodiments in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) that contain computer-usable program code.

[0272] The present application is described with reference to implementation flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the process in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0273] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which is implemented in the implementation flow diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0274] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing the steps in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0275] The above description is only a preferred embodiment of the present application and is not intended to limit the protection scope of the present application.

Claims

1. A method for deriving effective brain connection information, characterized in that: The method comprises: determining disturbance data of a first brain node among at least two brain nodes and brain neural data of the at least two brain nodes in a time series; Obtaining, by means of a digital twin brain model, first predicted neural data of the at least two brain nodes at the next moment based on the disturbance data of the first brain node and the brain neural data of the at least two brain nodes in a time series; wherein the digital twin brain model is obtained by training based on a time series prediction network; Based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment, the effective brain connection information from the first brain node to the second brain node is determined; wherein the second brain node is the other brain node among the at least two brain nodes except the first brain node; the second predicted neural data represents the predicted data obtained when the disturbance data is not added.

2. The method according to claim 1, characterized in that The method further comprises: Through the digital twin brain model, based on the brain neural data of the at least two brain nodes in a time series, the second predicted neural data of the at least two brain nodes at the next moment are obtained.

3. The method according to claim 2, characterized in that The effective brain connection information includes an effective connection type, and determining the effective brain connection information from the first brain node to the second brain node based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment includes: When the first predicted neural data of the second brain node at the next moment is greater than the second predicted neural data of the second brain node at the next moment, determining that the effective connection type is an excitatory type; When the first predicted neural data of the second brain node at the next moment is less than the second predicted neural data of the second brain node at the next moment, it is determined that the effective connection type is an inhibition type.

4. The method according to claim 2, characterized in that: The brain nerve data of the at least two brain nodes in the time series include the brain nerve data of the at least two brain nodes corresponding to the preset time length p+1; wherein the brain nerve data corresponding to the preset time length p+1 includes the brain nerve data from the tpth moment to the tth moment, p is an integer greater than or equal to 0, and t is an integer greater than p; The obtaining, by means of the digital twin brain model, first predicted neural data of the at least two brain nodes at the next moment based on the disturbance data of the first brain node and the brain neural data of the at least two brain nodes in a time series, comprises: Determine the perturbed brain neural data of the at least two brain nodes at the tth moment based on the disturbance data of the first brain node and the brain neural data of the at least two brain nodes at the tth moment; Through the digital twin brain model, based on the brain neural data of the at least two brain nodes after disturbance at the tth moment, and the brain neural data of the at least two brain nodes from the tpth moment to the t-1th moment, the first predicted neural data of the at least two brain nodes at the t+1th moment are obtained.

5. The method according to claim 4, characterized in that The determining, based on the disturbance data of the first brain node and the brain nerve data of the at least two brain nodes at the tth moment, the brain nerve data of the at least two brain nodes after disturbance at the tth moment comprises: The disturbance data of the first brain node and the brain neural data of the at least two brain nodes at the t-th moment are superimposed to obtain the disturbed brain neural data of the at least two brain nodes at the t-th moment.

6. The method according to claim 4, characterized in that The method of obtaining second predicted neural data of the at least two brain nodes at a next moment based on the brain neural data of the at least two brain nodes in a time series through the digital twin brain model includes: Through the digital twin brain model, based on the brain neural data of the at least two brain nodes from the tp-th moment to the t-th moment, the second predicted neural data of the at least two brain nodes at the t+1-th moment are obtained.

7. The method according to claim 6, characterized in that The effective brain connection information includes an effective connection direction, and determining the effective brain connection information from the first brain node to the second brain node includes: The effective connection direction is determined to be from the first brain node to the second brain node.

8. The method according to claim 7, characterized in that The effective brain connection information includes an effective connection type, and determining the effective brain connection information from the first brain node to the second brain node based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment includes: When the first predicted neural data of the second brain node at the time t+1 is greater than the second predicted neural data of the second brain node at the time t+1, determining that the effective connection type is an excitatory type; When the first predicted neural data of the second brain node at the time t+1 is less than the second predicted neural data of the second brain node at the time t+1, it is determined that the effective connection type is an inhibition type.

9. The method according to claim 7 or 8, characterized in that: The effective brain connection information includes effective connection strength, and determining the effective brain connection information from the first brain node to the second brain node based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment includes: Determine the strength information of the first predicted neural data of the second brain node at the t+1th time, and determine the strength information of the second predicted neural data of the second brain node at the t+1th time; The effective connection strength is determined based on the strength information of the first predicted neural data and the strength information of the second predicted neural data.

10. The method according to any one of claims 1 to 8, characterized in that The determining of disturbance data of a first brain node of the at least two brain nodes comprises: Setting disturbance data of the first brain node based on a preset disturbance intensity; The method further comprises: The at least two brain nodes are traversed to determine brain effective connection information between any two brain nodes of the at least two brain nodes to obtain the whole-brain effective connection of the at least two brain nodes.

11. The method according to any one of claims 1 to 8, characterized in that: The method further comprises: Using the digital twin brain model, based on the plurality of disturbance data of the first brain node corresponding to the plurality of moments and the brain neural data of the at least two brain nodes in a time series, respectively obtain a plurality of first predicted neural data of the at least two brain nodes corresponding to the plurality of moments; Using the digital twin brain model, based on the brain neural data of the at least two brain nodes in a time series, respectively obtain a plurality of second predicted neural data of the at least two brain nodes corresponding to the plurality of moments; Determine, based on the plurality of first predicted neural data and the plurality of second predicted neural data, a plurality of brain effective connection information from the first brain node to the second brain node corresponding to the plurality of time instants; Based on the multiple brain effective connection information, target effective connection information from the first brain node to the second brain node is determined.

12. A training method for a digital twin brain model, characterized in that: The method comprises: The time series prediction network is trained based on the training data set to obtain a digital twin brain model; wherein the digital twin brain model is used to predict the neural data of the next moment based on the brain neural data of the brain node in the time series; The training data set includes neural training data of at least two brain nodes corresponding to a preset time length p+1; wherein the neural training data corresponding to the preset time length p+1 includes neural training data from the mpth moment to the mth moment, and the training data set also includes neural training data of the at least two brain nodes at the m+1th moment; p is an integer greater than or equal to 0, and m is an integer greater than p; The time series prediction network is trained based on the training data set to obtain a digital twin brain model, including: Obtaining, by the time series prediction network, prediction data of the at least two brain nodes at the m+1th moment based on the neural training data of the at least two brain nodes from the mpth moment to the mth moment; Based on the prediction data of the at least two brain nodes at the m+1th moment and the neural training data of the at least two brain nodes at the m+1th moment, the time series prediction network is corrected to obtain the digital twin brain model.

13. The method according to claim 12, characterized in that The method of correcting the time series prediction network based on the prediction data of the at least two brain nodes at the m+1th moment and the neural training data of the at least two brain nodes at the m+1th moment to obtain the digital twin brain model comprises: Determining model parameters based on the prediction data of the at least two brain nodes at the m+1th moment, the neural training data of the at least two brain nodes at the m+1th moment, and the preset time length corresponding to the training data set; The digital twin brain model is determined based on the model parameters.

14. The method according to claim 12, characterized in that The time series prediction network is trained based on the training data set to obtain a digital twin brain model, including: Obtaining, by means of the time series prediction network, a plurality of prediction data corresponding to a plurality of moments of the at least two brain nodes based on the training data set; Determining a plurality of data errors corresponding to the plurality of moments based on the plurality of prediction data and the neural training data of the at least two brain nodes corresponding to the plurality of moments; determining model parameters based on the plurality of data errors; The digital twin brain model is determined based on the model parameters.

15. The method according to claim 13, characterized in that The method further comprises: Determining a coefficient of determination corresponding to the digital twin brain model based on the training data set and a prediction data set corresponding to the training data set; Determine, by means of the digital twin brain model, a first functional connectivity matrix predicted by the digital twin brain model based on the brain neural data and random noise, and determine a matrix correlation coefficient between the first functional connectivity matrix and a second functional connectivity matrix corresponding to the training data set; Based on the determination coefficient and the matrix correlation coefficient, model performance parameters of the digital twin brain model are determined.

16. The method according to claim 15, characterized in that The training data set includes neural training data corresponding to T moments of at least two brain nodes, the prediction data set includes prediction data corresponding to T moments of the at least two brain nodes, T is an integer greater than 0, and determining the determination coefficient corresponding to the digital twin brain model based on the training data set and the prediction data set corresponding to the training data set includes: Based on the neural training data of the at least two brain nodes corresponding to T moments, determining the data means of the at least two brain nodes corresponding to T moments; Based on the data mean, the neural training data corresponding to T moments and the predicted data corresponding to T moments, the determination coefficient corresponding to the digital twin brain model is determined.

17. A device for deriving effective brain connection information, characterized in that: The device for deriving effective brain connection information comprises: a determining unit, configured to determine disturbance data of a first brain node among at least two brain nodes and brain neural data of the at least two brain nodes in a time series; An acquisition unit, configured to obtain, through a digital twin brain model, first predicted neural data of the at least two brain nodes at a next moment based on the disturbance data of the first brain node and the brain neural data of the at least two brain nodes in a time series; wherein the digital twin brain model is obtained by training based on a time series prediction network; The determination unit is used to determine the effective brain connection information between the first brain node and the second brain node based on the first predicted neural data of the at least two brain nodes at the next moment and the second predicted neural data of the at least two brain nodes at the next moment; wherein the second brain node is the other brain node among the at least two brain nodes except the first brain node; and the second predicted neural data represents the predicted data obtained when the disturbance data is not added.

18. A computer device, characterized in that: The computer device comprises: a processor and a memory; wherein, The memory is used to store a computer program that can be run on the processor; The processor is configured to execute the method according to any one of claims 1-11 or 12-16 when running the computer program.

Images