Multi-brain region brain model construction method and device, storage medium and electronic equipment

Abstract

The disclosure provides a multi-brain region brain model construction method and device, a storage medium and an electronic device, and relates to the field of computational neuroscience. The method comprises the following steps: obtaining a positron emission tomography-magnetic resonance brain image, wherein the brain image represents the physiological function metabolism and anatomical structure of the brain; based on the brain image, a plurality of brain regions included in the brain image are divided into working memory brain regions and non-working memory brain regions, and the network topology of the plurality of brain regions included in the brain image is determined; based on the network topology and a chemical synapse plasticity model with time delay, the information transmission mechanism between the plurality of brain regions is determined respectively, wherein the information transmission mechanism represents the release and reception of chemicals between the brain regions; based on the plurality of brain regions and the information transmission mechanism between the plurality of brain regions, the brain model of the working memory brain regions and the non-working memory brain regions is constructed respectively, and a multi-brain region brain model corresponding to the brain image is obtained. In this way, the biological rationality and biological interpretability of the multi-brain region brain model can be improved.

Description

Technical Field

[0001] This disclosure relates to the field of computational neuroscience technology, specifically to a method and apparatus for constructing a multi-brain-region brain-like model, a storage medium, and an electronic device. Background Technology

[0002] For biological nervous systems, information transmission can be simplified to the unidirectional transmission of electrical signals by neurons. Inspired by biological nervous systems, spiking neural networks (SNNs) are a new type of neural network that more closely resembles the working mechanism of the human brain's neural network. SNN models use neurons as processing units of nonlinear state dynamics, supplemented by the regulation of synaptic connection weight dynamics, giving them powerful processing capabilities for complex spatiotemporal information. Therefore, in the field of computational neuroscience, SNNs are often used to simulate various activities within the brain. However, brain-like models in related technologies often suffer from the following problems.

[0003] First, the nodes of brain-like models are usually composed of single-type neuron models, without considering the functional division of brain regions, resulting in the problem of brain region homogenization in brain-like models. Second, the network topology of the nodes in brain-like models is usually generated by algorithms and lacks biological rationality. Finally, the synaptic plasticity model used to connect neurons often depends on the difference between the actual discharge and the target discharge of the neuron, which does not conform to the dynamic range of biological synaptic delay and lacks biological interpretability. Summary of the Invention

[0004] In view of this, the present disclosure provides a method and apparatus for constructing a multi-brain region brain-like model, a storage medium, and an electronic device.

[0005] In a first aspect, one embodiment of this disclosure provides a method for constructing a multi-brain-region brain-like model. The method includes: acquiring positron emission tomography-magnetic resonance (PET-MRI) brain images, wherein the brain images characterize the physiological functions, metabolism, and anatomical structure of the brain; based on the brain images, dividing the multiple brain regions included in the brain images into working memory brain regions and non-working memory brain regions, and determining the network topology of the multiple brain regions included in the brain images; based on the network topology and a time-delayed chemical synaptic plasticity model, determining the information transmission mechanism between the multiple brain regions, wherein the information transmission mechanism characterizes the release and reception of chemical substances between brain regions; and based on the multiple brain regions and the information transmission mechanism between the multiple brain regions, constructing brain-like models of the working memory brain regions and non-working memory brain regions respectively, thereby obtaining a multi-brain-region brain-like model corresponding to the brain images.

[0006] In conjunction with the first aspect, in some implementations of the first aspect, the brain regions included in the brain image are divided into working memory brain regions and non-working memory brain regions, including: determining the location of each of the multiple brain regions; dividing the multiple brain regions into working memory brain regions and non-working memory brain regions based on their respective locations; constructing brain-like models of the working memory brain regions and non-working memory brain regions based on the information transmission mechanism between the multiple brain regions, including: generating nodes of the brain-like model in the working memory brain regions and nodes of the brain-like model in the non-working memory brain regions based on multiple preset neuron models; and constructing brain-like models of the working memory brain regions and non-working memory brain regions respectively, using the information transmission mechanism between the multiple brain regions as the edges of the brain-like models.

[0007] In conjunction with the first aspect, in some implementations of the first aspect, the preset neuron model includes a prefrontal cortex neuron model, a hippocampal neuron model, and a non-working memory neuron model. Based on multiple preset neuron models, nodes of the brain-like model in the working memory brain region and nodes of the brain-like model in the non-working memory brain region are generated respectively, including: dividing the working memory brain region into the prefrontal cortex brain region and the hippocampus brain region based on the respective locations of multiple brain regions; generating nodes of the brain-like model corresponding to the prefrontal cortex brain region based on the prefrontal cortex neuron model; generating nodes of the brain-like model corresponding to the hippocampus brain region based on the hippocampal neuron model; and generating nodes of the brain-like model corresponding to the non-working memory brain region based on the non-working memory neuron model.

[0008] In conjunction with the first aspect, in some implementations of the first aspect, information transmission mechanisms between multiple brain regions are determined based on network topology and a time-delayed chemical synaptic plasticity model, including: determining the synapses included in each brain region based on network topology; and determining the weight of each synapse based on the time-delayed chemical synaptic plasticity model, which serves as the information transmission mechanism between multiple brain regions.

[0009] In conjunction with the first aspect, in certain implementations of the first aspect, time-delayed chemical synaptic plasticity models include:

[0010]

[0011] Among them, I syn Indicates synaptic current; g syn Indicates synaptic conductance; r represents receptor binding fraction; V syn V represents the reversible synaptic potential; pre and V post α and β represent the presynaptic neuron membrane potential and the postsynaptic neuron membrane potential, respectively; α and β represent the forward and reverse rate constants of neurotransmitter binding, respectively; T represents the concentration of neurotransmitter; and τ represents the synaptic transmission delay.

[0012] In conjunction with the first aspect, in some implementations of the first aspect, synapses include excitatory synapses and inhibitory synapses. Based on the network topology, the number of synapses included in each brain region is determined, including: based on the network topology, determining the number of edges corresponding to each brain region in the network topology; generating synapses corresponding to each brain region based on the number of edges corresponding to each brain region; and determining each synapse as an excitatory synapse or an inhibitory synapse based on a preset ratio.

[0013] In conjunction with the first aspect, in some implementations of the first aspect, determining the network topology of multiple brain regions includes: determining multiple brain regions included in the brain images based on brain images, and using the multiple brain regions as nodes of the network topology; determining the functional connectivity strength between the multiple brain regions respectively; filtering the functional connectivity strength based on the functional connectivity strength threshold, and generating the edges of the network topology; and constructing the network topology based on the nodes and edges of the network topology.

[0014] Secondly, one embodiment of this disclosure provides a multi-brain-region brain-like model construction device, the device comprising: an acquisition module for acquiring positron emission tomography-magnetic resonance brain images, the brain images representing the physiological functions, metabolism, and anatomical structure of the brain; a first determination module for dividing multiple brain regions included in the brain images into working memory brain regions and non-working memory brain regions, and determining the network topology of the multiple brain regions; a second determination module for determining the information transmission mechanism between the multiple brain regions based on the network topology and a time-delayed chemical synaptic plasticity model, the information transmission mechanism representing the release and reception of chemical substances between brain regions; and a construction module for constructing brain-like models of working memory brain regions and non-working memory brain regions based on the information transmission mechanism between the multiple brain regions, thereby obtaining a multi-brain-region brain-like model corresponding to the brain images.

[0015] Thirdly, one embodiment of this disclosure provides an electronic device including: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute the multi-brain region brain-like model construction method of the first aspect by executing the executable instructions.

[0016] Fourthly, one embodiment of this disclosure provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the multi-brain region brain-like model construction method of the first aspect.

[0017] In this embodiment, the network topology of the multi-brain-region brain-like model is constructed based on real physiological function metabolism and anatomical brain images, which can reflect the real brain network connections. At the same time, multiple brain regions are divided into working memory brain regions and non-working memory brain regions, and their respective brain-like models are constructed based on the firing characteristics of brain regions in different regions, which can solve the problem of the single neuron model. In addition, a chemical synaptic plasticity model with time delay is used as the edge of the multi-brain-region brain-like model to connect the nodes of each brain region, which can accurately simulate the regulatory effect of synapses on neurons in biological nervous systems. Therefore, the constructed multi-brain-region brain-like model can overcome the limitations of insufficient biological rationality of existing spiking neural networks and has biological interpretability. Attached Figure Description

[0018] The above and other objects, features, and advantages of this disclosure will become more apparent from the more detailed description of the embodiments thereof in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this disclosure and form part of the specification. They are used together with the embodiments of this disclosure to explain the disclosure and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0019] Figure 1 The diagram shown is a system architecture schematic of a multi-brain region brain-like model construction method provided in an exemplary embodiment of this disclosure.

[0020] Figure 2 The diagram shown is a flowchart illustrating a method for constructing a multi-brain region brain-like model according to an exemplary embodiment of this disclosure.

[0021] Figure 3a The diagram shown is a schematic representation of an exemplary embodiment of the present disclosure, illustrating the process of dividing multiple brain regions included in a brain image into working memory brain regions and non-working memory brain regions.

[0022] Figure 3b The diagram shows a flowchart illustrating the process of constructing brain-like models of working memory brain regions and non-working memory brain regions based on an information transmission mechanism between multiple brain regions, as provided in an exemplary embodiment of this disclosure.

[0023] Figure 4 The diagram shown is a schematic diagram of a multi-brain region brain functional network topology provided in an exemplary embodiment of this disclosure.

[0024] Figure 5 The diagram shows a flowchart illustrating how a brain-like model node in the working memory brain region and a brain-like model node in the non-working memory brain region are generated based on multiple preset neuron models, according to an exemplary embodiment of this disclosure.

[0025] Figure 6aThe diagram shown is a schematic representation of the firing patterns of a prefrontal neuron model provided in an exemplary embodiment of this disclosure.

[0026] Figure 6b The diagram shown is a schematic representation of the firing patterns of a hippocampal neuron model provided in an exemplary embodiment of this disclosure.

[0027] Figure 6c The diagram shown is a schematic representation of the firing patterns of a non-working memory neuron model provided in an exemplary embodiment of this disclosure.

[0028] Figure 7 The diagram shown is a schematic representation of an exemplary embodiment of this disclosure, illustrating how a network topology-based and time-delayed chemical synaptic plasticity model is used to determine the information transmission mechanism between multiple brain regions.

[0029] Figure 8 The diagram shown is a network topology diagram provided in an exemplary embodiment of this disclosure.

[0030] Figure 9 The diagram shown is a schematic representation of an exemplary embodiment of this disclosure, illustrating the process of determining the synapses included in each brain region based on network topology.

[0031] Figure 10 The diagram shown is a flowchart illustrating the process of determining the network topology of multiple brain regions according to an exemplary embodiment of this disclosure.

[0032] Figure 11 The diagram shown is a schematic diagram of the structure of a multi-brain region brain-like model construction device provided in an exemplary embodiment of this disclosure.

[0033] Figure 12 The diagram shown is a structural schematic of an electronic device provided in an exemplary embodiment of this disclosure. Detailed Implementation

[0034] The technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0035] Computational neuroscientists have conducted extensive research on constructing more biologically plausible brain-like models, starting from the three basic elements of spiking neural networks: neuron models, synaptic plasticity models, and network topology.

[0036] In related technologies, the nodes of brain-like models are usually composed of single-type neuron models. For example, the Izhikevich neuron model has high computational performance and is easy to construct large-scale networks, but it cannot describe the dynamic changes of ion channels; or, the Hodgkin-Huxley neuron model can accurately and in detail describe the changes in neuronal membrane potential caused by the dynamic changes of various ion channels, quantitatively describe the changes in neuronal membrane voltage and current, and is the most biologically representative. However, the realization of biological brain functions is accomplished by the collaboration of multiple types of neurons. Brain-like models constructed solely from single-type neuron models suffer from the disadvantage of a single neuron model and are a bottleneck in forming multi-brain-region brain-like models.

[0037] In the study of synaptic plasticity models, drawing on research findings on biological synapses, the co-regulation of excitatory and inhibitory synaptic plasticity can improve the performance of population rate encoding, and this has already been applied to the construction of brain-like models. Furthermore, in biological nervous systems, the diffusion of neurotransmitters along concentration gradients in chemical synapses creates a time delay in information transmission; chemical synapses with time delays are better suited for higher cognitive functions. Therefore, some scholars have attempted to introduce synaptic time delays into learning algorithms based on spiking neural networks. Studies have shown that this algorithm outperforms traditional synaptic weight learning methods in both accuracy and efficiency. However, the synaptic time delay in related technologies depends on the difference between the actual discharge and the target discharge of neurons, which does not conform to the time dynamic range of biological synaptic time delays and lacks biological interpretability.

[0038] In the field of network topology research, numerous neuroscientists have discovered that the topological structure of biological brain networks possesses small-world and scale-free properties. Based on these findings, some researchers have attempted to generate network topologies with small-world and scale-free properties using algorithms. For example, some researchers have generated small-world spiking neural networks using the Watts-Strogatz algorithm, investigated the impact of topology generation parameters on the firing dynamics of spiking neural networks, and established the relationship between network topology generation parameters and the stability of spiking neural network firing modes. Alternatively, some researchers have generated scale-free spiking neural networks using the Barbasi-Albert model algorithm, discovering that spike-timing-dependent plasticity (STDP) can simultaneously optimize multicoherent resonance and synchronization through self-synaptic delays. However, most current spiking neural network topologies are algorithmically generated. Although their overall characteristics have some consistency with real brain networks, exhibiting small-world or scale-free properties, the actual connections of their nodes differ from those in real brain networks and exhibit significant randomness, failing to reflect the true connectivity of real brain networks.

[0039] To address the aforementioned technical problems, this disclosure provides a method for constructing a multi-brain-region brain-like model. The method includes: acquiring positron emission tomography-magnetic resonance (PET-MRI) brain images, which characterize the physiological functions, metabolism, and anatomical structure of the brain; based on the brain images, dividing the multiple brain regions included in the images into working memory regions and non-working memory regions, and determining the network topology of the multiple brain regions included in the images; based on the network topology and a time-delayed chemical synaptic plasticity model, determining the information transmission mechanism between the multiple brain regions, which characterizes the release and reception of chemical substances between brain regions; and based on the multiple brain regions and the information transmission mechanism between them, constructing brain-like models of the working memory regions and non-working memory regions respectively, thereby obtaining a multi-brain-region brain-like model corresponding to the brain images. In the method disclosed herein, the network topology of the multi-brain-region brain-like model is constructed based on real physiological function metabolism and anatomical brain images, thus reflecting the real brain network connections. Simultaneously, multiple brain regions are divided into working memory and non-working memory regions, and separate brain-like models are constructed based on the firing characteristics of brain regions in each region, solving the problem of singular neuron models. Furthermore, a time-delayed chemical synaptic plasticity model is used as the edge of the brain-like model, connecting the nodes of each brain region, accurately simulating the regulatory effect of synapses on neurons in biological nervous systems. Therefore, the constructed multi-brain-region brain-like model can overcome the limitations of insufficient biological rationality in existing spiking neural networks and possesses biological interpretability.

[0040] This disclosure provides a method, apparatus, electronic device, and storage medium for constructing a multi-brain-region brain-like model. Specifically, the multi-brain-region brain-like model construction apparatus can be integrated into an electronic device, such as a terminal or server.

[0041] It is understood that the multi-brain region brain-like model construction method of this embodiment can be executed on a terminal, on a server, or jointly by a terminal and a server. The above examples should not be construed as limiting this disclosure.

[0042] Figure 1 An exemplary system architecture diagram is shown that can be applied to the multi-brain region brain-like model construction method or multi-brain region brain-like model construction apparatus in the embodiments of this disclosure.

[0043] like Figure 1 As shown, the system architecture 100 includes a terminal 110 and a server 120. The terminal 110 and the server 120 are connected via a network, such as a wired or wireless network. The multi-brain region brain-like model construction device can be integrated into the server 120.

[0044] Server 120 can be used to acquire positron emission tomography-magnetic resonance brain images, which characterize the physiological functions, metabolism, and anatomical structure of the brain. Based on the brain images, multiple brain regions included in the images are divided into working memory regions and non-working memory regions, and the network topology of the multiple brain regions included in the images is determined. Based on the network topology and a time-delayed chemical synaptic plasticity model, the information transmission mechanism between multiple brain regions is determined, which characterizes the release and reception of chemical substances between brain regions. Based on multiple brain regions and the information transmission mechanism between multiple brain regions, brain-like models of working memory regions and non-working memory regions are constructed respectively, resulting in a multi-brain region brain-like model corresponding to the brain images.

[0045] Server 120 can be a single server, a server cluster consisting of multiple servers, or a cloud server. For example, a server can be an interconnection server between multiple heterogeneous systems or a backend server. It can also be an independent physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, and big data and artificial intelligence platforms, etc.

[0046] Terminal 110 can send positron emission tomography-magnetic resonance (PET) brain images obtained from a database or captured in advance to server 120, or receive multi-brain region brain-like models generated by server 120. Terminal 110 may include mobile phones, personal computers, or medical imaging diagnostic equipment such as PET-MRI scanners.

[0047] Those skilled in the art will know that Figure 1 The number of terminals and servers shown is merely illustrative. Depending on actual needs, there may be any number of terminals and servers, and this disclosure does not impose any limitation on this.

[0048] The following will describe the exemplary implementation method in detail with reference to the accompanying drawings and embodiments.

[0049] Figure 2 The diagram shown is a flowchart illustrating a method for constructing a multi-brain region brain-like model according to an exemplary embodiment of this disclosure. Figure 2 As shown in the embodiments of this disclosure, the method for constructing a multi-brain region brain-like model includes the following steps.

[0050] S210, acquires positron emission tomography-magnetic resonance brain images.

[0051] Positron emission tomography-magnetic resonance (PET-MR) brain imaging can characterize the physiological functions, metabolism, and anatomical structure of the brain. Specifically, PET-MR brain imaging is an image captured using PET-MR imaging technology, which is a novel imaging technique that organically combines positron emission tomography (PET) and magnetic resonance imaging (MRI). PET can reflect the metabolic information of human physiological functions, and the biochemical signals it provides more accurately reflect the connectivity between brain regions, while MRI can reflect the anatomical structure of brain tissue.

[0052] In related technologies, functional magnetic resonance imaging (fMRI) is usually used to study brain functional connectivity. However, since brain activity is generated by the complex interaction of biochemical and electrical signals, fMRI, which measures neural activity by blood oxygen levels, cannot fully reflect the diversity of communication between brain regions.

[0053] The PET-MR images in this embodiment can use MRI images that reflect tissue structure to attenuate PET images, achieving information complementarity between functional and anatomical images, and can more effectively and accurately reflect the functional connectivity between multiple brain regions.

[0054] S220, divides the multiple brain regions included in the brain image into working memory brain regions and non-working memory brain regions, and determines the network topology of multiple brain regions.

[0055] Brain imaging reflects the anatomical structure of the brain. Therefore, by matching brain images with standard brain atlases, it is possible to identify multiple brain regions included in the images and further determine the individual brain region numbers. These brain region numbers allow for further identification of whether the brain region is located in working memory or non-working memory areas. It is understood that both working memory and non-working memory areas each contain multiple brain regions.

[0056] Specifically, different brain regions perform different tasks when humans work or think. Studies have shown that brain regions located in the hippocampus are responsible for forming and storing memories of new things and events; brain regions located in the prefrontal cortex play an important role in the encoding and retrieval of working memory, short-term memory, and long-term memory. In this disclosure, brain regions involved in working memory tasks are referred to as working memory brain regions. When processing these working memory tasks, the firing characteristics of working memory brain regions also differ from those of other non-working memory brain regions; in related technologies, brain-like models constructed using single-neuron models cannot reflect the differences between working memory brain regions and non-working memory brain regions.

[0057] Next, by analyzing the physiological functions and metabolism between multiple brain regions as reflected in brain imaging, the functional connectivity strength between these regions is determined. Furthermore, based on this connectivity strength, the connection relationships between the multiple brain regions are determined, generating a network topology corresponding to the brain imaging. This generated network topology allows for the differentiation of brain regions with different functions, facilitating the subsequent construction of multi-region brain-like models and addressing the issue of brain region homogenization in related technologies.

[0058] S230, based on network topology and a time-delayed chemical synaptic plasticity model, identifies the information transmission mechanisms between multiple brain regions.

[0059] Edges in a network topology can reflect the existence of connections between multiple brain regions, but they cannot reflect the strength, excitability, or inhibitory nature of information transmission between connected brain regions. Therefore, after determining the connections, it is necessary to determine the connection weights of each connection using a time-delayed chemical synaptic plasticity model to identify the information transmission mechanism between multiple brain regions.

[0060] In biological nervous systems, synapses are key structures that allow two neurons to contact each other and transmit information. During information transmission, when the presynaptic neuron is stimulated by signals from the environment or other neurons, neurotransmitters stored in the presynaptic membrane are released into the synaptic cleft, acting on the corresponding receptors on the postsynaptic membrane and transmitting the neurotransmitter signal to the postsynaptic neuron.

[0061] In this embodiment, the time-delayed chemical synaptic plasticity model can simulate the release and reception processes of neurotransmitters in the aforementioned biological nervous system, and the time delay is introduced to simulate the time interval for neurotransmitter diffusion in the synaptic cleft. Therefore, the information transmission mechanism established by the time-delayed chemical synaptic plasticity model can better simulate the temporal dynamic range of biological synaptic time delays, has biological interpretability, and can also improve the model's information processing capabilities.

[0062] S240, based on the information transmission mechanism between multiple brain regions, constructs brain-like models of working memory brain regions and non-working memory brain regions respectively, and obtains multi-brain-region brain-like models corresponding to brain images.

[0063] The above steps identified multiple brain regions included in the working memory and non-working memory brain regions, as well as the information transmission mechanisms between these regions. Therefore, nodes for a brain-like model can be generated based on the individual firing characteristics of each brain region within the working memory and non-working memory brain regions. The information transmission mechanisms between these regions can then be used as edges in the multi-brain-region brain-like model, resulting in a multi-brain-region brain-like model corresponding to the brain image.

[0064] In this embodiment, the network topology of the multi-brain-region brain-like model is constructed based on real physiological function metabolism and anatomical brain images, which can reflect the real brain network connections. At the same time, multiple brain regions are divided into working memory brain regions and non-working memory brain regions, and their respective brain-like models are constructed based on the firing characteristics of brain regions in different regions, which can solve the problem of the single neuron model. In addition, a chemical synaptic plasticity model with time delay is used as the edge of the multi-brain-region brain-like model to connect the nodes of each brain region, which can accurately simulate the regulatory effect of synapses on neurons in biological nervous systems. Therefore, the constructed multi-brain-region brain-like model can overcome the limitations of insufficient biological rationality of existing spiking neural networks and has biological interpretability.

[0065] The following is combined with Figure 3a , Figure 3b This paper further introduces a specific implementation method for constructing a multi-brain-region brain-like model.

[0066] Figure 3a The diagram illustrates a process for dividing a brain image into working memory regions and non-working memory regions, according to an exemplary embodiment of this disclosure. Figure 3a As shown, the brain regions included in the brain images are divided into working memory regions and non-working memory regions, specifically including the following steps.

[0067] S310, which determines the location of each of the multiple brain regions.

[0068] As described above, by matching brain images with standard brain atlases, the brain region numbers of multiple brain regions included in the brain images can be determined. Then, by comparing brain images with standard brain atlases, the locations of each brain region can be determined based on the brain region numbers, thus obtaining the brain region coordinates.

[0069] Standard brain atlases are digital brain structure atlases developed with reference to different brain anatomical structures. For example, a standard brain atlas could be the SPM standard brain atlas.

[0070] S320 divides multiple brain regions into working memory regions and non-working memory regions based on their respective locations.

[0071] The location of a brain region can reflect its function. For example, as listed above, the prefrontal cortex and hippocampus play important roles in spatial learning and memory-guided decision-making. Therefore, the prefrontal cortex and hippocampus are both working memory regions.

[0072] Therefore, it is possible to determine whether a brain region participates in working memory tasks based on its individual location; or, based on the individual locations of multiple brain regions, multiple brain regions can be divided into multiple working memory brain regions, and brain regions other than working memory brain regions can be identified as non-working memory brain regions. It is worth noting that the hippocampus and prefrontal cortex regions listed above are exemplary, and the working memory brain regions in this embodiment are not limited to the hippocampus and prefrontal cortex regions mentioned above; the specific classification of brain regions should be determined by their location.

[0073] Figure 4 The diagram shown is a schematic representation of a multi-brain region brain functional network topology provided in an exemplary embodiment of this disclosure. Figure 4 The network topology shown clearly distinguishes multiple brain regions into working memory regions or non-working memory regions, and also reflects the connectivity between multiple brain regions.

[0074] Correspondingly, such as Figure 3b As shown, based on the information transmission mechanism between multiple brain regions, brain-like models of working memory brain regions and non-working memory brain regions are constructed respectively, specifically including the following steps.

[0075] S330 generates nodes of the brain-like model in the working memory brain region and nodes of the brain-like model in the non-working memory brain region, based on multiple preset neuron models.

[0076] As described above, when processing the working memory task, the firing characteristics of the working memory brain region differ from those of other non-working memory brain regions. Therefore, embodiments of this disclosure simulate the firing characteristics of multiple brain regions in different areas using multiple preset neuron models to generate nodes of brain-like models in both the working memory and non-working memory brain regions.

[0077] Specifically, the preset neuron model is divided into working memory neuron model and non-working memory neuron model. When generating nodes of the brain-like model, nodes of the brain-like model corresponding to the working memory brain region are generated based on the working memory neuron model, and nodes of the brain-like model corresponding to the non-working memory brain region are generated based on the non-working memory neuron model.

[0078] S340 uses the information transmission mechanism between multiple brain regions as the edges of the brain-like model to construct brain-like models of working memory brain regions and non-working memory brain regions respectively.

[0079] As mentioned above, the information transmission mechanisms between multiple brain regions can reflect the intensity, excitability, or inhibitory nature of information transmission between connected brain regions. Therefore, the information transmission mechanisms between multiple brain regions can be used as edges in a brain-like model, connecting the nodes of the brain-like model corresponding to each brain region.

[0080] In the embodiments disclosed herein, a multi-brain-region brain-like model can be constructed by using the locations of multiple brain regions, and the firing characteristics of different brain regions can be simulated by using different neuron models, thus solving the problem of brain region homogeneity in related technologies and improving the biological rationality of multi-brain-region brain-like models.

[0081] The following is combined with Figure 5 Furthermore, it introduces several pre-defined neuron models used to simulate the firing characteristics of different brain regions.

[0082] Figure 5 The diagram illustrates a process for generating nodes of a brain-like model in working memory brain regions and nodes of a brain-like model in non-working memory brain regions, based on multiple preset neuron models, according to an exemplary embodiment of this disclosure. Figure 5 As shown, based on multiple preset neuron models, nodes of the brain-like model in the working memory brain region and nodes of the brain-like model in the non-working memory brain region are generated respectively, specifically including the following steps.

[0083] S2421 divides the working memory brain region into the prefrontal cortex and hippocampus based on the location of multiple brain regions.

[0084] As mentioned above, the prefrontal cortex and hippocampus both belong to the working memory brain region, but they play different roles in working memory tasks, and their firing characteristics are not entirely the same. Therefore, the working memory brain region can be further divided into the prefrontal cortex and hippocampus based on the location of each of the multiple brain regions within the working memory brain region.

[0085] S2422 generates nodes of a brain-like model corresponding to the prefrontal cortex brain region based on the prefrontal cortex neuron model.

[0086] Specifically, the mathematical description of the prefrontal neuron model is as follows:

[0087]

[0088] Among them, C p V is the membrane capacitance of a prefrontal lobe neuron; p The membrane voltage of the prefrontal cortex neuron; g LP For leakage conductance; E L The effective resting potential; ΔT is the threshold slope coefficient; V T V is the effective threshold potential. r Reset voltage; I e ω is the sum of the external input current and the current conducted through multiple synapses; τ is the adaptive current; ω For time scale; I a For adaptive current sensitivity to voltage; I b It is a pulse-triggered adaptive current.

[0089] In accordance with the application scenarios of this disclosure embodiment, the parameters of the prefrontal neuron model are set as follows: C p =123.71pF; g LP =7.16nS; E L =-71.48mV; ΔT = 4.51mV; τ ω =120.98ms; V r =―84.23mV; V T =―55.38mV; I a =―7.16pA;I b =19.82pA. Its discharge mode is as follows: Figure 6a As shown.

[0090] The prefrontal cortex neuron model provided in this disclosure can accurately predict the peak firing time of neurons in the biological prefrontal cortex within the range of physiological reliability, and reproduce most of the dynamic characteristics of the prefrontal cortex. Therefore, the brain-like model nodes generated based on the prefrontal cortex neuron model can accurately simulate the firing characteristics of the prefrontal brain region.

[0091] S2423 generates nodes of a brain-like model corresponding to the hippocampal brain region based on the hippocampal neuron model.

[0092] Specifically, the mathematical description of the hippocampal neuron model is as follows:

[0093]

[0094]

[0095] Among them, C H V is the membrane capacitance of a hippocampal neuron. H g is the membrane voltage of a hippocampal neuron. LH Leakage conductance; V L Leakage voltage; I APP is the sum of the external input current and the current conducted through multiple synapses; n is the activation variable of the delayed rectified potassium ion current. g Na and V Na These are sodium ion conductivity and voltage, respectively; g Kdr and V K These are potassium ion conductivity and voltage, respectively.

[0096] m ∞ (V H ), n ∞ (V H ) and τ n (V H These are all parameters that change with the membrane potential of hippocampal neurons, and their mathematical descriptions are as follows:

[0097]

[0098]

[0099]

[0100] In accordance with the application scenarios of this disclosure embodiment, the parameters of the hippocampal neuron model are set as follows: C H =1uF / cm 2 G LH =0.05mS / cm 2 V L =―70mV; g Na =35mS / cm 2 V Na =55mV; g Kdr =6mS / cm 2 V K = -90mV. Its discharge mode is as follows: Figure 6b As shown.

[0101] The hippocampal neuron model provided in this disclosure can quantitatively reproduce various electrophysiological behaviors of biological neurons, exhibits consistent dynamic behavior with the electrophysiological data of the hippocampus, and is easy to simulate on a large scale. Therefore, the brain-like model nodes generated based on the hippocampal neuron model can accurately simulate the firing characteristics of the hippocampus.

[0102] S2424 generates nodes of a brain-like model corresponding to the non-working memory brain region based on the non-working memory neuron model.

[0103] Specifically, the Hodgkin-Huxley neuron model can be used as a model of non-working memory neurons. The mathematical description of the Hodgkin-Huxley neuron model is as follows:

[0104]

[0105] Where V is the neuronal membrane potential, and the unit is mV; C m This refers to film capacitance, measured in μF / cm. 2 ;I app The stimulation current is expressed in μA / cm. 2 g Na g K and g L These are the conductances of sodium ion channels, potassium ion channels, and leakage channels, respectively, in mS / cm. 2 E Na E K and E LThese are the inversion potentials of the sodium ion channel, potassium ion channel, and drain channel, respectively, in mV; m and h are the gate variables of the sodium ion channel; n is the gate variable of the potassium ion channel.

[0106] The α and β functions are rate functions related to membrane potential:

[0107]

[0108] In accordance with the application scenarios of this disclosure embodiment, the parameters of the Hodgkin-Huxley neuron model are set as follows: C m =1μF / cm 2 g Na =120mS / cm 2 g k =36mS / cm 2 g L =3mS / cm 2 E Na =50mV; E k = -77mV; E L = -54.4mV. Its discharge mode is as follows: Figure 6c As shown.

[0109] The Hodgkin-Huxley neuron model can accurately and comprehensively describe the changes in neuronal membrane potential caused by the dynamic changes of various ion channels, and quantitatively describe the changes in neuronal membrane voltage and current, making it the most biologically representative model.

[0110] In the embodiments disclosed herein, multi-region brain-like models can be constructed using prefrontal cortex neuron models, hippocampal neuron models, and Hodgkin-Huxley neuron models, respectively. This allows different brain regions to exhibit different firing characteristics, enabling the multi-region brain-like models to represent different brain functions, further improving the biological rationality and biological interpretability of the multi-region brain-like models.

[0111] The above details the node generation method of multi-brain region brain-inspired models. Below, we will combine... Figure 7 This section further introduces the edge generation method of the multi-brain region brain-like model.

[0112] Figure 7 The diagram illustrates a flowchart of an exemplary embodiment of this disclosure, illustrating how a network topology-based and time-delayed chemical synaptic plasticity model determines the information transmission mechanism between multiple brain regions. Figure 7 As shown, based on network topology and a chemical synaptic plasticity model with time delay, the information transmission mechanism between multiple brain regions is determined, specifically including the following steps.

[0113] S231, based on network topology, determines the synapses included in each brain region.

[0114] As mentioned above, synapses in biological nervous systems are key structures that connect two neurons, through which information is transmitted between them.

[0115] In this embodiment of the application, brain regions are regarded as neurons in a biological nervous system. Then, based on the network topology of multiple brain regions, the synapses included in each brain region can be determined.

[0116] For example, in Figure 8 In the network topology shown, brain region 1 is connected to brain region 2 and brain region 3 respectively. Therefore, brain region 1 includes two synapses, which connect brain region 1 to brain region 2 and brain region 3 respectively. Correspondingly, brain region 2 also includes a synapse corresponding to brain region 1, and brain region 3 also includes a synapse corresponding to brain region 1. There is no connection between brain region 2 and brain region 3, so there is no synapse between brain region 2 and brain region 3.

[0117] S232, based on a time-delayed chemical synaptic plasticity model, determines the weight of each synapse and serves as an information transmission mechanism between multiple brain regions.

[0118] After identifying the synapses included in each brain region, the weights of each synapse can be determined individually based on a time-delayed chemical synaptic plasticity model.

[0119] In biological nervous systems, the connection strength between neurons changes over time. That is, different stimuli, even with the same type and intensity, will produce different electrical effects on the postsynaptic membrane. Therefore, the dynamic behavior of synaptic weights is a crucial mechanism for the regulation of spiking neural networks. The synaptic plasticity provided in this disclosure accurately simulates the dynamic behavior of synaptic weights described above.

[0120] Furthermore, the model provided in this disclosure is based on the neurotransmitter transmission and reception mechanism in biological neural systems, which is more consistent with the connectivity relationships between brain regions constructed based on biochemical signals. Moreover, the synaptic time delay can simulate the time delay in information transmission caused by the diffusion of neurotransmitters with concentration gradients in synapses, further improving the biological rationality of the multi-brain-region brain-like model and enhancing the model's information processing capabilities.

[0121] The following section further describes the specific implementation methods for determining the information transmission mechanisms between multiple brain regions. For example... Figure 9 As shown, based on the network topology, the synapses included in each brain region are determined, specifically including the following steps.

[0122] S2311, Based on the network topology, determine the number of edges corresponding to each brain region in the network topology.

[0123] By analyzing the network topology, we can determine the connections between each brain region and other brain regions, and thus determine the number of edges corresponding to each brain region.

[0124] For example, continue to refer to Figure 8 The number of edges corresponding to brain region 1 is two, while the number of edges corresponding to brain regions 2 and 3 is one each.

[0125] S2312 generates synapses corresponding to each brain region based on the number of edges corresponding to each brain region.

[0126] The number of synapses corresponding to each brain region is the same as the number of edges corresponding to each brain region.

[0127] For example, continue to refer to Figure 8 The number of synapses corresponding to brain region 1 is two, while the number of synapses corresponding to brain regions 2 and 3 is one each.

[0128] S2313, based on a preset ratio, each synapse is identified as either an excitatory synapse or an inhibitory synapse.

[0129] Synapses include excitatory synapses and inhibitory synapses. In biological nervous systems, excitatory synapses are those where, when the presynaptic neuron is excited, synaptic vesicles release neurotransmitters with excitatory effects (e.g., acetylcholine, norepinephrine). These neurotransmitters depolarize the postsynaptic neuron, exhibiting an excitatory effect. Inhibitory synapses, on the other hand, are those where, when the presynaptic neuron is excited, synaptic vesicles release neurotransmitters with inhibitory effects (e.g., dopamine, glycine, etc.). These neurotransmitters hyperpolarize the postsynaptic neuron, exhibiting an inhibitory effect.

[0130] To simulate the excitatory and inhibitory effects described above, embodiments of this disclosure randomly divide the synapses generated in step S2312 into excitatory synapses or inhibitory synapses based on a preset ratio. The preset ratio is determined according to the actual ratio of excitatory to inhibitory synapses in a biological nervous system. For example, the preset ratio could be excitatory synapse : inhibitory synapse = 4:1.

[0131] In this embodiment, the excitatory and / or inhibitory synapses included in each brain region can be obtained based on network topology, and the ratio of excitatory to inhibitory synapses conforms to the ratio of excitatory to inhibitory synapses in biological neural systems. Since excitatory and inhibitory synapses have different regulatory rules on brain regions, this embodiment can distinguish the excitatory or inhibitory effects of different types of synapses on brain regions, further improving the biological rationality of multi-brain-region brain-like models.

[0132] In some embodiments, the mathematical description of the chemical synaptic plasticity model with time delay is as follows:

[0133]

[0134] Among them, I syn Indicates synaptic current; g syn Indicates synaptic conductance; r represents receptor binding fraction; V syn V represents the reversible synaptic potential; pre and V post α and β represent the presynaptic neuron membrane potential and the postsynaptic neuron membrane potential, respectively; α and β represent the forward and reverse rate constants of neurotransmitter binding, respectively; T represents the concentration of neurotransmitter; and τ represents the synaptic transmission delay.

[0135] In the application scenario of this disclosure embodiment, τ is set to 20ms. The parameters of the excitatory synapse are set as follows: V syn =0mV, α=2, β=1; the parameters of the inhibitory synapse are set as follows: V syn =-70mV, α=0.9, β=0.1.

[0136] The regulatory rules of excitatory and inhibitory synapses are described below. Both excitatory and inhibitory synapses regulate the spiking neural network by changing synaptic conductance. The regulatory rules for each type of synapse can be divided into two cases: the postsynaptic neuron does not receive the action potential of the presynaptic neuron, and the postsynaptic neuron receives the action potential of the presynaptic neuron.

[0137] (1) If the postsynaptic neuron j does not receive the action potential of the presynaptic neuron i, the changes in the conductance of the excitatory and inhibitory synapses are as follows:

[0138] Excitatory synapse:

[0139] Inhibitory synapses:

[0140] Where, τ ex and τ in τ represents the attenuation constant of the excitatory and inhibitory synaptic conductances, respectively. ex =τ in =5ms; g ex and g in These represent excitatory and inhibitory synaptic conductances, respectively.

[0141] (2) If the postsynaptic neuron j receives the action potential of the presynaptic neuron i, the changes in the conductance of the excitatory and inhibitory synapses are as follows:

[0142] Excitatory synapse:

[0143] Inhibitory synapses:

[0144] in, and These represent the excitatory and inhibitory conductance increments caused by the action potential, respectively. and Each is determined by the excitability correction function w ij and the inhibition correction function m ij Adjustments are made. The excitability correction function w ij and the inhibition correction function m ij The mathematical descriptions are as follows:

[0145]

[0146]

[0147] Among them, A + and A ― These are the maximum and minimum correction values ​​for excitatory synaptic conductance, respectively; B + and B ― These represent the maximum and minimum correction values ​​for inhibitory synaptic conductance, respectively. Δt is the firing interval between the presynaptic and postsynaptic neurons. τ + and τ ― These represent the time intervals between neuronal firing during synaptic enhancement and synaptic depletion, respectively. In the context of the application scenarios of this disclosure, τ + =τ ― =20ms, g max =0.015, A + =0.1, A ― =0.105, B + =0.02, B ― =0.03.

[0148] In some embodiments, maximum and minimum values ​​can be set for excitatory conductance and inhibitory synaptic conductance to further improve the biological rationality of the multi-brain-region brain-like model. When the excitatory conductance or inhibitory synaptic conductance calculated by the above method is greater than a preset maximum value g... max At that time, the preset maximum value g will be set. max As the final calculation result; when the excitatory conductance or inhibitory synaptic conductance calculated by the above method is less than the preset minimum value g min At that time, the preset minimum value g will be set. min The final calculation result is determined based on the actual synaptic conductance in the biological nervous system.

[0149] Therefore, the time-delayed chemical synaptic plasticity model provided in this application provides a reasonable and accurate mathematical expression of the excitation and inhibition relationships between brain regions, as well as the dynamic behavior of synaptic weights between brain regions, further improving the biological rationality of the multi-brain region brain-like model.

[0150] In the above embodiments, there is a one-to-one correspondence between multiple brain regions in the network topology and nodes of the multi-brain region brain-like model. Furthermore, the information transmission mechanism between multiple brain regions is constructed based on edges in the network topology. Therefore, a reasonable and reliable network topology is the foundation for constructing a biologically interpretable multi-brain region brain-like model. The following section combines... Figure 10 This section further introduces the implementation methods for constructing a reasonable and reliable network topology.

[0151] Figure 10 The diagram illustrates a flowchart of a method for determining the network topology of multiple brain regions, provided in an exemplary embodiment of this disclosure. Figure 10 As shown, the network topology of multiple brain regions is determined, specifically including the following steps.

[0152] S221, based on brain images, identify multiple brain regions included in the brain images and use these multiple brain regions as nodes in the network topology.

[0153] By matching brain images with standard brain maps, multiple brain regions included in the brain images can be identified, and these identified brain regions can be used as multiple nodes in the network topology.

[0154] S222 determines the strength of functional connectivity between multiple brain regions.

[0155] First, the strength of functional connectivity between multiple brain regions is determined by examining the physiological metabolic functions of the brain as reflected in brain imaging.

[0156] Specifically, the strength of functional connectivity between multiple brain regions can be determined by the Pearson correlation coefficient between the average time series of physiological and functional metabolic measurement signals from different brain regions, and its mathematical expression is as follows:

[0157]

[0158] Where, x i (t) and x j (t) represents the time series of physiological and functional metabolic measurement signals of nodes i and j at time t, respectively; and The average time series of physiological and functional metabolic measurement signals for nodes i and j are respectively; r ijLet be the correlation coefficient between node i and node j; T is the number of time points. The larger the absolute value of the correlation coefficient, the stronger the functional connection between the two nodes; the smaller the absolute value of the correlation coefficient, the weaker the functional connection between the two nodes.

[0159] S223, filter functional connection strength based on functional connection strength threshold, and generate edges of network topology.

[0160] Understandably, when the correlation coefficient is less than a certain value, two brain regions can be considered unrelated. Therefore, the correlation coefficients calculated in the above steps can be filtered using a preset functional connectivity strength threshold, retaining only partial connections between multiple brain regions with correlation coefficients greater than the functional connectivity strength threshold as edges in the network topology.

[0161] Specifically, the functional connectivity strength threshold X suitable for this embodiment can be determined based on the characteristics of biological brain network topology: the average degree of the network is greater than 2ln(N); the network density is in the range of 3.6%-39.3%; the network has a small-world characteristic greater than 1 and a power-law exponent of around 2. th .

[0162] Among them, network density is used to describe the density of interconnections between nodes in the entire network; in an unweighted network, node degree represents the number of edges directly connected to that node, and the larger the node degree, the more important the node is; small-world property is used to describe the global information transmission efficiency and the degree of clustering of the network; the degree distribution of scale-free networks follows a power-law distribution, and the power-law exponent represents the degree of difference in node degree in the scale-free network, and the larger the power-law exponent, the more obvious the difference in node degree.

[0163] S224: Construct the network topology based on the nodes and edges of the network topology.

[0164] Using the aforementioned brain regions as nodes in the network topology, the correlation coefficient between nodes is greater than the functional connectivity strength threshold X. th If the correlation coefficient between two nodes is less than or equal to the functional connection strength threshold X, then a connection is considered to exist between them, and the edge weight is set to 1; otherwise, the correlation coefficient between the nodes is less than or equal to the functional connection strength threshold X. th If no connection exists between two nodes, the edge weight is set to 0. The binary matrix constructed through the above steps represents the final network topology of multiple brain regions.

[0165] In this embodiment of the disclosure, a reasonable and reliable network topology is established by the functional connectivity strength between multiple brain regions, which lays the foundation for the subsequent construction of a multi-brain region brain-like model with biological interpretability.

[0166] In some embodiments, brain images may be preprocessed before matching them with standard brain atlases to address issues such as low signal-to-noise ratio and instability. For example, the DPABI (Data Processing & Analysis for Brain Imaging) brain imaging data processing and analysis toolkit can be used for preprocessing, including steps such as: temporal correction; head motion correction; spatial normalization; smoothing; and filtering.

[0167] Figure 11 The diagram shown is a structural schematic of a multi-brain region brain-like model construction device provided in an exemplary embodiment of this disclosure. Figure 11 As shown, the multi-brain-region brain-like model construction device 1100 provided in this embodiment includes: an acquisition module 1110, used to acquire positron emission tomography-magnetic resonance brain images, the brain images representing the physiological functions, metabolism, and anatomical structure of the brain; a first determination module 1120, used to divide the multiple brain regions included in the brain images into working memory brain regions and non-working memory brain regions, and determine the network topology of the multiple brain regions; a second determination module 1130, used to determine the information transmission mechanism between the multiple brain regions based on the network topology and a time-delayed chemical synaptic plasticity model, the information transmission mechanism representing the release and reception of chemical substances between brain regions; and a construction module 1140, used to construct brain-like models of working memory brain regions and non-working memory brain regions based on the information transmission mechanism between the multiple brain regions, to obtain a multi-brain-region brain-like model corresponding to the brain images.

[0168] In some embodiments, the first determining module 1120 is further configured to determine the location of each of the plurality of brain regions; and divide the plurality of brain regions into working memory brain regions and non-working memory brain regions based on the location of each of the plurality of brain regions. The constructing module 1140 is further configured to generate nodes of the brain-like model in the working memory brain region and nodes of the brain-like model in the non-working memory brain region based on a plurality of preset neuron models; and construct brain-like models of the working memory brain region and non-working memory brain region respectively, using the information transmission mechanism between the plurality of brain regions as the edges of the brain-like model.

[0169] In some embodiments, the preset neuron model includes a prefrontal cortex neuron model, a hippocampal neuron model, and a non-working memory neuron model. The construction module 1140 is further configured to: divide the working memory brain region into the prefrontal cortex brain region and the hippocampal brain region based on the respective locations of multiple brain regions; generate nodes of the brain-like model corresponding to the prefrontal cortex brain region based on the prefrontal cortex neuron model; generate nodes of the brain-like model corresponding to the hippocampal brain region based on the hippocampal neuron model; and generate nodes of the brain-like model corresponding to the non-working memory brain region based on the non-working memory neuron model.

[0170] In some embodiments, the second determining module 1130 is further configured to: determine the synapses included in each brain region based on the network topology; and determine the weight of each synapse based on a time-delayed chemical synaptic plasticity model, and use it as an information transmission mechanism between multiple brain regions.

[0171] In some embodiments, a time-delayed chemical synaptic plasticity model includes:

[0172]

[0173] Among them, I syn Indicates synaptic current; g syn Indicates synaptic conductance; r represents receptor binding fraction; V syn V represents the reversible synaptic potential; pre and V post α and β represent the presynaptic neuron membrane potential and the postsynaptic neuron membrane potential, respectively; α and β represent the forward and reverse rate constants of neurotransmitter binding, respectively; T represents the concentration of neurotransmitter; and τ represents the synaptic transmission delay.

[0174] In some embodiments, the synapse includes excitatory synapses and inhibitory synapses. The second determining module 1130 is further configured to: determine the number of edges corresponding to each brain region in the network topology based on the network topology; generate synapses corresponding to each brain region based on the number of edges corresponding to each brain region; and determine each synapse as an excitatory synapse or an inhibitory synapse based on a preset ratio.

[0175] In some embodiments, the first determining module 1120 is further configured to: determine multiple brain regions included in the brain images based on the brain images, and use the multiple brain regions as nodes of the network topology; determine the functional connectivity strength between the multiple brain regions respectively; filter the functional connectivity strength based on the functional connectivity strength threshold, and generate the edges of the network topology; and construct the network topology based on the nodes and edges of the network topology.

[0176] Below, for reference Figure 12 To describe an electronic device according to embodiments of the present disclosure. Figure 12 The diagram shown is a structural schematic of an electronic device provided in an exemplary embodiment of this disclosure.

[0177] like Figure 12 As shown, the electronic device 1200 includes one or more processors 1210 and memory 1220.

[0178] The processor 1210 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 1200 to perform desired functions.

[0179] The memory 1220 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 1210 may execute the program instructions to implement the multi-brain region brain-like model construction methods of the various embodiments of this disclosure described above, and / or other desired functions.

[0180] In some embodiments, the electronic device 1200 may further include an input device 1230 and an output device 1240, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).

[0181] The input device 1230 may include, for example, a keyboard, a mouse, etc.

[0182] The output device 1240 can output various information to the outside, including a brain-like model constructed based on the method of the embodiments of this disclosure. The output device 1240 may include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, etc.

[0183] Of course, for the sake of simplicity, Figure 12 Only some of the components of the electronic device 1200 relevant to this disclosure are shown, omitting components such as buses, input / output interfaces, etc. In addition, the electronic device 1200 may include any other suitable components depending on the specific application.

[0184] In addition to the methods and devices described above, embodiments of this disclosure may also be computer program products comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the methods for constructing multi-brain region brain-like models according to various embodiments of this disclosure as described above.

[0185] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this disclosure. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on a user's computing device, partially on a user's computing device, as a standalone software package, partially on a user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0186] Furthermore, embodiments of this disclosure may also be computer-readable storage media storing computer program instructions that, when executed by a processor, cause the processor to perform the steps in the multi-brain region brain-like model construction method according to various embodiments of this disclosure described above.

[0187] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0188] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0189] The block diagrams of devices, apparatuses, devices, and systems disclosed herein are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0190] It should also be noted that in the apparatus, devices, and methods of this disclosure, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions to this disclosure.

[0191] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0192] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.

Claims

1. A method for constructing a multi-brain-region brain-like model, characterized in that, include: Acquire positron emission tomography-magnetic resonance brain images, which characterize the physiological functions, metabolism, and anatomical structure of the brain; The brain images are divided into working memory brain regions and non-working memory brain regions, and the network topology of the brain regions is determined. Based on the network topology and the time-delayed chemical synaptic plasticity model, the information transmission mechanisms between the multiple brain regions are determined, and the information transmission mechanisms characterize the release and reception of chemical substances between brain regions. Based on the information transmission mechanism between the multiple brain regions, brain-like models of the working memory brain region and the non-working memory brain region are constructed respectively to obtain the multi-brain region brain-like model corresponding to the brain image. The determination of information transmission mechanisms between the multiple brain regions based on the network topology and the time-delayed chemical synaptic plasticity model includes: Based on the network topology, the synapses included in each of the brain regions are determined respectively; Based on the time-delayed chemical synaptic plasticity model, the weight of each synapse is determined and used as an information transmission mechanism between the multiple brain regions. The time-delayed chemical synaptic plasticity model includes: in, Indicates synaptic current; Indicates synaptic conductance; Indicates the receptor binding fraction; This indicates a reversible synaptic potential; and These represent the presynaptic neuron membrane potential and the postsynaptic neuron membrane potential, respectively. and These represent the forward and reverse rate constants for neurotransmitter binding, respectively. Indicates the concentration of neurotransmitters; This indicates the time delay in synaptic transmission.

2. The method according to claim 1, characterized in that, The division of the brain regions included in the brain image into working memory regions and non-working memory regions includes: The locations of each of the plurality of brain regions were determined respectively; Based on the location of each of the multiple brain regions, the multiple brain regions are divided into the working memory brain region and the non-working memory brain region; The method of constructing brain-like models of the working memory brain region and the non-working memory brain region based on the information transmission mechanism between the multiple brain regions includes: Based on multiple preset neuron models, nodes of the brain-like model in the working memory brain region and nodes of the brain-like model in the non-working memory brain region are generated respectively. Using the information transmission mechanism between the multiple brain regions as the edges of the brain-like model, brain-like models of the working memory brain region and the non-working memory brain region are constructed respectively.

3. The method according to claim 2, characterized in that, The preset neuron model includes a prefrontal cortex neuron model, a hippocampal neuron model, and a non-working memory neuron model. The step of generating nodes of the brain-like model in the working memory brain region and nodes of the brain-like model in the non-working memory brain region based on multiple preset neuron models includes: Based on the location of each of the multiple brain regions, the working memory brain region is divided into the prefrontal cortex and the hippocampus. Based on the prefrontal neuron model, nodes of the brain-like model corresponding to the prefrontal brain region are generated; Based on the hippocampal neuron model, nodes of the brain-like model corresponding to the hippocampal brain region are generated; Based on the non-working memory neuron model, nodes of the brain-like model corresponding to the non-working memory brain region are generated.

4. The method according to claim 1, characterized in that, The synapses include excitatory synapses and inhibitory synapses. The determination of the synapses included in each brain region based on the network topology includes: Based on the network topology, determine the number of edges corresponding to each brain region in the network topology; Based on the number of edges corresponding to each brain region, generate synapses corresponding to each brain region; Based on a preset ratio, each synapse is identified as either an excitatory synapse or an inhibitory synapse.

5. The method according to claim 1, characterized in that, Determining the network topology of the plurality of brain regions includes: Based on the brain images, multiple brain regions included in the brain images are identified, and these multiple brain regions are used as nodes in the network topology. Determine the functional connectivity strength between the plurality of brain regions respectively; The functional connection strength is filtered based on the functional connection strength threshold to generate the edges of the network topology; The network topology is constructed based on the nodes and edges of the network topology.

6. A device for constructing a multi-brain-region brain-like model, characterized in that, include: The acquisition module is used to acquire positron emission tomography-magnetic resonance brain images, which characterize the physiological functions, metabolism, and anatomical structure of the brain. The first determining module is used to divide the multiple brain regions included in the brain image into working memory brain regions and non-working memory brain regions, and to determine the network topology of the multiple brain regions; The second determining module is used to determine the information transmission mechanism between the plurality of brain regions based on the network topology and the chemical synaptic plasticity model with time delay, wherein the information transmission mechanism characterizes the release and reception of chemical substances between brain regions; A construction module is used to construct brain-like models of the working memory brain region and the non-working memory brain region respectively based on the information transmission mechanism between the multiple brain regions, so as to obtain a multi-brain-region brain-like model corresponding to the brain image. The second determining module is further configured to: determine the synapses included in each brain region based on the network topology; and determine the weight of each synapse based on the time-delayed chemical synaptic plasticity model, and use it as an information transmission mechanism between the multiple brain regions. The time-delayed chemical synaptic plasticity model includes: in, Indicates synaptic current; Indicates synaptic conductance; Indicates the receptor binding fraction; This indicates a reversible synaptic potential; and These represent the presynaptic neuron membrane potential and the postsynaptic neuron membrane potential, respectively. and These represent the forward and reverse rate constants for neurotransmitter binding, respectively. Indicates the concentration of neurotransmitters; This indicates the time delay in synaptic transmission.

7. An electronic device, characterized in that, include: processor; as well as Memory for storing the executable instructions of the processor; The processor is configured to execute the multi-brain region brain-like model construction method according to any one of claims 1 to 5 by executing the executable instructions.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for constructing a multi-brain region brain-like model as described in any one of claims 1 to 5.

Images