Knowledge graph construction method based on BNCT, terminal and medium

Abstract

The invention provides a BNCT-based knowledge graph construction method, a BNCT treatment effect prediction method, a terminal and a computer storage medium, and the method comprises the steps: extracting each piece of sub-data in each treatment data set corresponding to a target tumor, and constructing an initial BNCT knowledge graph based on each piece of sub-data and an incidence relation between each piece of sub-data; and utilizing a sum-product network method to reinforce a causal association relationship in the initial BNCT knowledge graph network so as to obtain a final BNCT knowledge graph, according to the method provided by the invention, the BNCT knowledge graph based on the causal association relationship can be quickly constructed based on the treatment scheme data and the patient data, and the treatment effect corresponding to the target tumor treatment scheme can be conveniently obtained based on the BNCT knowledge graph, so that decision support for auxiliary judgment is provided for making the treatment scheme.

Description

Technical Field

[0001] This application belongs to the technical field of boron concentration acquisition, and relates to a method for constructing a knowledge graph based on BNCT, a method for predicting the treatment effect of BNCT, a terminal, and a computer storage medium. Background Art

[0002] Boron Neutron Capture Therapy (BNCT), as an emerging tumor-targeted therapy, realizes "cellular-level targeted blasting" of cancer cells through the precise capture reaction of boron-10 nuclide and thermal neutrons, while minimizing damage to surrounding healthy tissues; its clinical decision-making highly depends on the multi-modal knowledge integration of nuclear physics, radiobiology, and oncology; in the prior art, the formulation of BNCT treatment plans is often based on manual experience, that is, based on the knowledge and experience of BNCT experts; however, this formulation method often cannot quickly, efficiently, and accurately integrate and construct interdisciplinary knowledge systems (such as neutron dose calculation models, boron drug pharmacokinetic parameters, tumor microenvironment characteristics, etc.), and cannot quickly process the complex associations between a large number of clinical cases and dynamic medical evidence, resulting in the BNCT treatment knowledge architecture obtained by the traditional knowledge invocation mode based on manual experience not only having efficiency bottlenecks such as long decision-making cycles and lagging knowledge updates, but even the homogenization degree of treatment plans being limited due to the cognitive biases of experts, thus reducing the accuracy of decision-making and affecting the treatment effect.

[0003] Therefore, how to quickly and accurately obtain the treatment effect corresponding to the treatment plan of BNCT has become a technical problem to be solved in this field. Summary of the Invention

[0004] In view of the above-mentioned disadvantages of the prior art, the purpose of this application is to provide a method for constructing a knowledge graph based on BNCT, a method for predicting the treatment effect of BNCT, a terminal, and a medium, which are used to solve the problems such as long decision-making cycles and lagging knowledge updates existing in the prior art when obtaining the BNCT treatment knowledge architecture.

[0005] To achieve the above purpose and other related purposes, this application provides a method for constructing a knowledge graph based on BNCT in the first aspect, including:

[0006] Obtain a treatment dataset corresponding to the target tumor for BNCT; the treatment dataset contains treatment plan data and treatment effect data in each historical treatment; based on the causal association relationships of the data in the treatment dataset in BNCT treatment, construct an initial BNCT knowledge graph, and construct a sample dataset corresponding to the initial BNCT knowledge graph; wherein, the causal association relationships include weighted causal relationships and joint causal relationships; among the nodes corresponding to the treatment plan data, extract the nodes that satisfy the weighted causal relationship as summation nodes; and among the nodes corresponding to the treatment effect data, extract the nodes that satisfy the joint causal relationship as product nodes; according to the sample dataset, use the sum-product network method to obtain the causal relationship probability corresponding to each summation node and obtain the causal relationship probability corresponding to the product node, so as to obtain the target BNCT knowledge graph.

[0007] In some embodiments of the first aspect of the present invention, the construction of the sample dataset corresponding to the initial BNCT knowledge graph includes:

[0008] Based on the node information in the initial BNCT knowledge graph, extract the feature variables corresponding to the node information from the treatment dataset; construct each feature variable belonging to the same treatment case into a sample data entry; perform the construction of each sample data entry on the treatment dataset to obtain the sample dataset corresponding to the initial BNCT knowledge graph.

[0009] In some embodiments of the first aspect of the present invention, the implementation manner of using the sum-product network method to obtain the causal relationship probability corresponding to each summation node and obtain the causal relationship probability corresponding to the product node includes:

[0010] Adopt the calculation method of frequency weight to obtain the causal relationship probability of the summation root node corresponding to the summation node; adopt the calculation method of association effect probability to obtain the causal relationship probability of the product root node corresponding to the product node.

[0011] In some embodiments of the first aspect of the present invention, the obtaining manner of adopting the calculation method of frequency weight to obtain the causal relationship probability of the summation root node corresponding to the summation node includes:

[0012] In the sample dataset, extract the sample subset corresponding to the current summation node, and obtain the number of sample data entries in the sample subset as the total number of samples of the current summation node; in the sample subset, extract the current sample number corresponding to the current summation root node; based on the current sample number and the total number of samples, calculate the ratio between the current sample data and the total number of samples as the causal relationship probability corresponding to the current summation root node.

[0013] In some embodiments of the first aspect of the present invention, the method for calculating the probability of the association effect to obtain the causal relationship probability of the product root node corresponding to the product node includes:

[0014] In the initial BNCT knowledge graph, determine the node group corresponding to the current product node; according to the causal association relationship, determine the treatment variable, confounding variable and result variable in the node group; based on the causal association relationship, construct an association probability function between the result variable and the treatment variable and the confounding variable; based on the sample data in the sample data subset, perform relationship probability calculation according to the association probability function to obtain the causal probability value corresponding to the product node group.

[0015] In some embodiments of the first aspect of the present invention, the extraction method of the summation node includes:

[0016] For a single node in each node corresponding to the treatment plan data, determine whether it is a child node; when the node is a child node, determine whether there is a weighted causal relationship between each root node corresponding to the current child node and the current child node. If so, label the child node as a summation node; and, the extraction method of the product node includes: for a single node in each node corresponding to the treatment effect data, determine whether the node is a child node; when the node is a child node, determine whether there is a joint causal relationship between each root node corresponding to the current child node and the current child node. If so, label the child node as a product node.

[0017] In some embodiments of the first aspect of the present invention, when the product node is the tumor volume change rate and the product root node is the boron drug dosage, the node group includes: neutron beam irradiation time, neutron beam energy, boron drug dosage, tumor boron concentration, tumor volume and tumor volume change rate; set the boron drug dosage as the treatment variable, set the neutron beam irradiation time and neutron beam energy as confounding variables respectively, and set the tumor volume concentration, tumor volume and tumor volume change rate as result variables respectively; construct an association probability function between the result variable and the treatment variable and the confounding variable, which is:

[0018] In the formula, P is the association effect probability; is the association effect probability under the influence of the interference confounding factor on the treatment variable, is the association effect probability of the treatment variable on the result variable under the influence of multiple interference confounding factors; U1 is the neutron beam irradiation time, U2 is the neutron beam energy; S2 is the tumor volume change rate; the tumor volume change rate is represented by the weighted combination or functional relationship of the tumor volume concentration and the tumor volume, and satisfies:

[0019] Among them, is the tumor boron concentration influence coefficient, is the irradiation time influence coefficient; S0 is the tumor boron concentration; S1 is the tumor volume.

[0020] To achieve the above and other related purposes, in a second aspect, the present application further provides a method for predicting the treatment effect of BNCT based on a knowledge graph, including:

[0021] Obtain a pre-constructed target BNCT knowledge graph; use sample data to train the target BNCT knowledge graph to obtain a trained knowledge graph; based on the trained knowledge graph, obtain the tumor treatment effect of a patient under a given treatment plan; wherein, the target BNCT knowledge graph is obtained by using the knowledge graph construction method based on BNCT described above in any way.

[0022] In a third aspect, the present application provides a terminal, including: a processor and a memory, the memory is communicatively connected to the processor; the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory so that the terminal executes the knowledge graph construction method based on BNCT described above in any way, or executes the method for predicting the treatment effect of BNCT based on the knowledge graph described above.

[0023] In a fourth aspect, the present application provides a computer storage medium, the computer storage medium stores a computer program, and when the computer program is executed by a processor, it implements the knowledge graph construction method based on BNCT described above in any way, or executes the method for predicting the treatment effect of BNCT based on the knowledge graph described above.

[0024] As described above, the knowledge graph construction method based on BNCT, the method for predicting the treatment effect of BNCT, the terminal and the medium provided by the present application, by constructing a BNCT knowledge graph based on causal association relationships, realize the systematic integration of treatment plan data and patient data, not only realize the transformation of fragmented knowledge into a structured relationship network, but also can infer the treatment effect corresponding to the treatment plan through knowledge reasoning, so as to assist doctors in predicting the treatment effects corresponding to different treatment plans; and, by constructing a dynamic knowledge update mechanism, based on the latest treatment sample data, update the BNCT knowledge graph, further improving the accuracy of the knowledge graph, and thus greatly improving the accuracy of personalized treatment plan design. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is shown as a flowchart of the knowledge graph construction method based on BNCT provided by an embodiment of the present application;

[0026] Figure 2It shows a schematic flowchart of step S20 in an embodiment of the present application;

[0027] Figure 3 It shows a schematic flowchart of step S40 in an embodiment of the present application;

[0028] Figure 4 It shows a schematic flowchart of the acquisition method of the causal relationship probability corresponding to a single summation root node in an embodiment of the present application;

[0029] Figure 5 It shows a schematic flowchart of the acquisition method of the causal relationship probability corresponding to a single product root node in an embodiment of the present application;

[0030] Figure 6 It shows a logic diagram of the causal association relationship between the boron drug dose and the tumor volume change rate in an embodiment of the present application;

[0031] Figure 7 It shows a schematic flowchart of the method for predicting the BNCT treatment effect based on the knowledge graph provided in the embodiment of the present application;

[0032] Figure 8 It shows a schematic structural diagram of the terminal in the embodiment of the present application. Detailed implementation manners

[0033] The following uses specific specific examples to illustrate the implementation manners of the present application. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0034] It should be noted that in the following description, reference is made to the accompanying drawings, which describe several embodiments of the present application. It should be understood that other embodiments can also be used, and mechanical composition, structure, electrical, and operational changes can be made without departing from the spirit and scope of the present application. The following detailed description should not be considered restrictive, and the scope of the embodiments of the present application is only defined by the claims of the published patent. The terms used here are only for describing specific embodiments and are not intended to limit the present application. Spatially related terms, such as "upper", "lower", "left", "right", "below", "beneath", "lower part", "above", "upper part", etc., can be used in the text to facilitate the description of the relationship between one element or feature shown in the figure and another element or feature.

[0035] Furthermore, as used herein, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly dictates otherwise. It should be further understood that the terms "comprising", "including" indicate the presence of the stated features, operations, elements, components, items, kinds, and / or groups, but do not preclude the presence, occurrence, or addition of one or more other features, operations, elements, components, items, kinds, and / or groups. The terms "or" and "and / or" used herein are to be construed as inclusive, meaning either any one or any combination.

[0036] To solve the technical problems in the prior art, in a first aspect, the present application provides a method for constructing a knowledge graph based on BNCT, which is used to construct a target BNCT knowledge graph based on causal association relationships for a target tumor type, so as to obtain a treatment effect prediction result corresponding to a treatment plan based on the target BNCT knowledge graph.

[0037] Wherein, the target tumor type is the tumor type targeted by the BNCT treatment plan, including but not limited to melanoma, glioma, etc.

[0038] Please refer to Figure 1 , which shows a schematic flowchart of the method for constructing a knowledge graph based on BNCT provided by an embodiment of the present application; as Figure 1 shown, the method includes the following steps:

[0039] S10, obtaining a treatment data set of a target tumor in historical BNCT treatments;

[0040] Wherein, the treatment data set is various types of data included in the historical treatment process, including treatment plan data and treatment effect data;

[0041] The treatment plan data are various treatment parameters set or generated during the BNCT treatment process, including information such as neutron source type, neutron source dose, irradiation angle, irradiation time, boron drug dosage, and boron drug dose absorption rate, etc.;

[0042] The treatment effect data are data on tumor control and surrounding tissue effects after BNCT treatment, including: information such as tumor radiation absorption rate, tumor volume change rate, and probability of normal tissue carcinogenesis, etc.

[0043] S20, using a data mining method to perform data preprocessing on the treatment data set to obtain a processed treatment data set;

[0044] Specifically, when this step is executed, as Figure 2 shown, it includes the following sub-steps:

[0045] S21, performing data denoising processing on each data in the treatment data set;

[0046] Specifically, remove noise, errors, missing values, etc. from each piece of data in the treatment dataset.

[0047] S22. Perform standardization processing on each piece of data in the treatment dataset.

[0048] Specifically, for data of a single type, perform operations such as normalization, standardization, or discretization on each piece of data under this type to convert data with different physical dimensions to a preset physical dimension, so that data of the same type has the same physical dimension.

[0049] In this application, perform the above steps on each of the treatment datasets to obtain each preprocessed treatment dataset.

[0050] In some alternative embodiments, when the treatment dataset contains text-type data, then when performing the standardization processing in step S22, it further includes:

[0051] Perform data encoding on the text-type data so that the encoded data can be computationally processed.

[0052] S30. Based on the causal association relationships between the data in the treatment dataset, construct an initial BNCT knowledge graph and obtain a sample dataset corresponding to the initial BNCT knowledge graph;

[0053] Among them, the causal association relationship is the mathematical and physical association relationship that each piece of data has in BNCT treatment; the causal association relationship refers to the mathematical and physical dependence relationship reflected by each type of data in BNCT treatment. The mathematical and physical dependence relationship is the statistical dependence and influence mechanism between different variables (such as drug dosage, treatment method, etc.), which is used to characterize the causal action path between variables and provide a scientific basis for BNCT treatment effect evaluation and decision support.

[0054] In this application, the causal association relationship is calculated using Bayesian causal probability, which is used to efficiently characterize the causal influence and action mechanism between different variables during the BNCT treatment process.

[0055] It should be noted that in some other embodiments, the causal association relationship can also use other correlation probabilities such as the Pearson correlation coefficient, Spearman rank correlation coefficient, etc. to reveal the correlation or causal relationship between feature variables from different statistical perspectives.

[0056] Specifically, convert each data in the treatment dataset into a node in the initial BNCT knowledge graph; extract the causal association relationships between the nodes, and convert the causal association relationships into directed edges in the initial BNCT knowledge graph; use the directed graph method to graphically represent the nodes and the causal association relationships between the nodes to construct the initial BNCT knowledge graph; among them, represent the nodes in the directed graph as each data (entity) in the graph, and represent the causal association relationship as a directed edge in the graph.

[0057] After constructing the initial BNCT knowledge graph, based on the node information in the initial BNCT knowledge graph, extract the feature variables corresponding to each node information from the pre-extracted treatment dataset; take each feature variable belonging to the same treatment case as a sample data strip; based on this method, extract sample data strips from the treatment dataset to construct a sample dataset corresponding to the initial BNCT knowledge graph; that is, the sample dataset contains each sample data strip.

[0058] In an alternative embodiment, the implementation method for extracting the causal association relationships between the nodes includes:

[0059] Take a treatment effect data in the treatment dataset as the current child node,

[0060] Use the semantic analysis method to extract the treatment plan data that has a causal association relationship with the treatment effect data from the treatment dataset as the current root node corresponding to the current child node;

[0061] Construct the causal association relationship between the current root node and the current child node as a directed edge between the current root node and the current child node.

[0062] Exemplarily, the treatment plan nodes in the initial BNCT knowledge graph include:

[0063] 1) Boron drug dose: such as "BPA 80ppm";

[0064] 2) Neutron source dose: such as "dose rate 2.5 Gy-eq";

[0065] 3) Administration method: such as "intravenous injection";

[0066] 4) Initial concentration of boron drug: such as 60ppm;

[0067] 5) Irradiation angle: such as 30°;

[0068] 6) Irradiation time: such as 15 minutes.

[0069] The treatment result nodes in the initial BNCT knowledge graph include:

[0070] 1) Tumor volume change rate: such as "volume reduced by 50%".

[0071] 2) Side effects: such as "mild nausea".

[0072] 3) Survival time: such as "survived for 24 months".

[0073] 4) Boron drug absorbed dose: such as "12 ppm".

[0074] Based on the causal association relationship, construct directed edges between nodes with causal association relationships; Exemplarily, the directed edges include:

[0075] "Boron drug dosage" → "Boron drug absorbed dose"

[0076] "Boron drug dose" → "Tumor volume change rate";

[0077] "Neutron source dose" → "Side effects".

[0078] Exemplarily, take "tumor volume change rate" as a child node, and "boron drug dose" is a parent node under this child node.

[0079] Optionally, after constructing the initial BNCT knowledge graph, the Neo4j tool can also be used to visualize the initial BNCT knowledge graph, so as to more efficiently and conveniently obtain the causal association relationships between the nodes in the graph based on the visualized initial BNCT knowledge graph.

[0080] S40. Among the nodes corresponding to the treatment plan data, extract the nodes that meet the weighted causal relationship as summation nodes; and among the nodes corresponding to the treatment effect data, extract the nodes that meet the joint causal relationship as product nodes; According to the sample data set, use the sum-product network method to obtain the causal relationship probabilities corresponding to each summation node and obtain the causal relationship probabilities corresponding to the product nodes, so as to obtain the target BNCT knowledge graph.

[0081] Among them, the sum-product network is a directed acyclic graph (DAG) structure containing summation nodes and product nodes;

[0082] The summation node performs a weighted summation operation on the outputs of the child nodes to represent the mixture of latent variables or the hierarchical clustering of data distributions or the probability weighting under different conditions;

[0083] The product node (Product Nodes) performs a product operation on the outputs of the child nodes to represent the independence assumption between variables or the joint distribution of local features.

[0084] Exemplarily, when considering the influence of different boron drug doses on the tumor volume change rate, the tumor volume change rate is used as a summation node (which is also a child node), and different boron drug doses are used as the root node. Then, this summation node is used to represent the influence probability of different boron drug doses on the tumor volume change rate (i.e., the causal relationship probability).

[0085] Moreover, when considering the influence of "boron drug dose" and "neutron irradiation dose" on "tumor volume change rate", the "tumor volume change rate" is used as a product node (which is also a child node), and the "boron drug dose" and "neutron irradiation dose" are used as the root nodes respectively. Then, this product node can be used to characterize the influence probability of different boron drug doses under the influence of neutron irradiation dose on the tumor volume change rate, or can be used to characterize the influence probability of different neutron irradiation doses under the influence of boron drug dose on the tumor volume change rate.

[0086] Specifically, when step S40 is executed, as Figure 3 shown, it includes the following steps:

[0087] S41, according to the causal association relationship between nodes, divide the summation nodes and product nodes among the nodes of the initial BNCT knowledge graph;

[0088] Among them, the form of the causal association relationship at least includes weighted causal relationship and joint causal relationship; the weighted causal relationship is a non-linear causal relationship, corresponding to the summation node, and the joint causal relationship is a multi-factor mixed causal relationship, corresponding to the product node.

[0089] In this application, the nodes of the initial BNCT knowledge graph include nodes corresponding to treatment plan data and nodes corresponding to treatment effect data;

[0090] By determining the form of the causal association relationship between nodes, extract the nodes that satisfy the weighted causal relationship among the nodes corresponding to the treatment plan data as the summation nodes; and among the nodes corresponding to the treatment effect data, by determining the form of the causal association relationship between nodes, extract the nodes that satisfy the joint causal relationship as the product nodes, so as to quickly and efficiently determine the node types to which each node belongs and provide data preparation for the execution of subsequent steps.

[0091] Specifically, for a single node among the nodes corresponding to the treatment plan data, determine whether this node is a child node; when this node is a child node, obtain the root nodes corresponding to the current child node; determine whether the causal relationship between each root node and the current child node is a weighted causal relationship. If so, label this child node as a summation node; execute this process for each node corresponding to the treatment plan data, thereby extracting each summation node.

[0092] Similarly, for a single node among the nodes corresponding to the treatment effect data, it is determined whether the node is a child node; when the node is a child node, the root nodes corresponding to the current child node are obtained; it is determined whether there is a joint causal relationship between each root node and the current child node, and if so, the child node is labeled as a product node; this process is performed on each node corresponding to the treatment plan data, so as to extract each product node.

[0093] In a specific embodiment, according to expert experience or by using a preset causal association relationship model, the correlation / independence between each node in the initial BNCT knowledge graph network is determined; based on the correlation between each node, sum nodes and product nodes are divided among the nodes in each initial BNCT knowledge graph network; exemplarily, by using the causal association relationship model built in the BNCT treatment planning software (TPS), the input parameters of the TPS are used as variable nodes, and the output parameters of the TPS are used as sum nodes or product nodes;

[0094] By using expert experience or a preset causal association relationship model, the existing causal association relationships between each node can be fully utilized, so that the sum nodes and product nodes can be determined more quickly and conveniently.

[0095] S42. For each of the sum nodes, a calculation method of frequency weight is adopted to obtain the causal relationship probability of each sum root node corresponding to the sum node;

[0096] Wherein, the frequency weight is the occurrence frequency of the sample data strip corresponding to the root node in the total sample data strip corresponding to the sum node;

[0097] For a single sum root node under a sum node, the obtaining manner of the corresponding causal relationship probability is as Figure 4 shown, and includes:

[0098] S421. In the sample data set, the sample subset corresponding to the current sum node is extracted, and the number of sample data strips in the sample subset is obtained as the total sample number of the current sum node;

[0099] Specifically, according to the numerical range corresponding to the current sum node, in the sample data set, each sample data strip whose numerical value of the sum node feature variable is within this numerical range is extracted; the set of each sample data strip within this numerical range is used as the sample subset corresponding to the current sum node, and the number of sample data strips in the sample subset is counted as the total sample number of the current sum node;

[0100] Among them, the summation node feature variable is the feature variable corresponding to / included in the summation node; for example, when the summation node is "boron drug dose > 50", the feature variable corresponding to / included in this summation node is "boron drug dose".

[0101] Exemplarily, when the current summation node is "boron drug dose > 50", in the sample dataset, extract the sample data entries where the value of the summation node feature variable, i.e., "boron drug dose", is greater than 50, a total of 100 entries; use the 100 extracted sample data entries as the sample data subset of the current summation node; correspondingly, the total number of samples of the current summation node is 100.

[0102] S422, in the sample subset, extract the current sample quantity corresponding to the current summation root node;

[0103] Specifically, determine the value range corresponding to the current summation root node; according to this value range, count the number of sample data entries in the sample subset where the value of the summation root node feature variable is within this value range, and use this number as the current sample quantity corresponding to the current summation root node;

[0104] Among them, the summation root node feature variable is the feature variable corresponding to / included in the summation root node; for example, when the summation root node is "boron drug dosage 10 - 15", the feature variable corresponding to / included in this summation root node is "boron drug dosage".

[0105] In the present application, the value range corresponding to a single summation root node is a pre - defined numerical range, and it satisfies: the total value range formed by the value range intervals corresponding to each summation root node (for the same feature variable) is the total numerical range of the feature variable corresponding to the summation root node in the sample dataset.

[0106] Exemplarily, when the summation node is boron drug dose, its corresponding summation root nodes are "boron drug dosage" in different value range intervals, including: "boron drug dosage" in the first interval (10 - 15), "boron drug dosage" in the second interval (15 - 20), "boron drug dosage" in the third interval (20 - 25), and "boron drug dosage" in the fourth interval (25 - 30), corresponding to the first summation root node, the second summation root node, the third summation root node, and the fourth summation root node respectively; for the first summation root node, in the sample subset corresponding to the summation node (a total of 100 sample data entries), count the number of sample data entries where "boron drug dosage" is within the first interval range, which is 30 entries, then use 30 as the current sample quantity corresponding to the current summation root node.

[0107] It should be noted that when constructing the initial BNCT knowledge graph, for each feature variable, first, the total value range in the sample dataset is statistically calculated, that is, the total value range; then, according to the preset number of intervals n, the total value range is divided into n intervals, and a corresponding node is established for each interval to construct nodes corresponding to different value range intervals of the same feature variable, thereby realizing the structured modeling of different value range intervals of the feature variable, providing a basis for subsequent causal association relationship modeling and reasoning.

[0108] Exemplarily, the feature variable of the summation node is the boron drug dose. The total value range of the boron drug dose in the sample dataset is obtained as 30 - 70; the number of intervals is predetermined as 5; according to this number of intervals, the total value range of the boron drug dose is evenly divided into 5 value range intervals, corresponding to: 30 - 40, 40 - 50, 50 - 60, and 60 - 70 respectively; based on different value range intervals, corresponding summation nodes are constructed, namely the first summation node (boron drug dose 30 - 40), the second summation node (boron drug dose 40 - 50), the third summation node (boron drug dose 50 - 60), and the fourth summation node (boron drug dose 60 - 70).

[0109] Similarly exemplarily, the feature variable of the summation root node is the boron drug administration amount. The total value range of the boron drug administration amount in the sample dataset is obtained as 10 - 30; the number of intervals is predetermined as 4; according to this number of intervals, the total value range of the boron drug administration amount is evenly divided into 4 value range intervals, corresponding to: 10 - 15, 15 - 20, 20 - 25, and 25 - 30 respectively; based on different value range intervals, corresponding summation root nodes are constructed, namely the first summation root node (boron drug administration amount 10 - 15), the second summation root node (boron drug administration amount 15 - 20), the third summation root node (boron drug administration amount 20 - 25), and the fourth summation root node (boron drug administration amount 25 - 30).

[0110] S423, based on the current sample quantity and the total sample quantity, calculate the proportion between the current sample data and the total sample quantity as the causal relationship probability corresponding to the current summation root node.

[0111] Specifically, after obtaining the current sample quantity corresponding to the current summation root node and the total sample quantity of the current summation node, calculate the ratio of the current sample quantity to the total sample quantity, and use this ratio as the causal relationship probability corresponding to the current summation root node. Quantify the contribution probability of each administration amount interval to the total dose interval, providing a basis for subsequent causal inference and knowledge graph parameter optimization.

[0112] Exemplarily, as shown in Table 1 below, the sum node has a boron drug dose of 50 - 60, and the corresponding total number of samples is 50; the sum root nodes corresponding to this sum node include: the first sum root node ("boron drug dosage" is 10 - 15), the second sum root node ("boron drug dosage" is 15 - 20), the third sum root node ("boron drug dosage" is 20 - 25), and the fourth sum root node ("boron drug dosage" is 25 - 30); among them, the current number of samples corresponding to the first sum root node is 30, and the corresponding causal relationship probability is 30 / 50 = 0.6; the current number of samples corresponding to the second sum root node is 15, and the corresponding causal relationship probability is 10 / 50 = 0.2, and so on. The causal relationship probabilities corresponding to the third sum root node and the fourth sum root node are both 0.1.

[0113] Table 1 Causal relationship probabilities corresponding to each sum root node under the sum node (boron drug dose is 50 - 60)

[0114]

[0115] As can be seen from Table 1 above, for the sum node (boron drug dose is 50 - 60), when the boron drug dosage is 10 - 15, the influence probability on this sum node is the largest, which is 0.6.

[0116] S43. For each product node, using the calculation method of the association effect probability, obtain the causal relationship probabilities of the product root nodes corresponding to this product node;

[0117] Among them, the association effect probability is the sum of the probability distribution of the result variable S and the marginal probability of each confounding variable U under the condition of the given treatment variable T;

[0118] In this application, the calculation method of the association effect probability includes:

[0119] 1) For the given treatment variable T, obtain the values of all possible confounding variables U under this treatment variable T;

[0120] 2) For each confounding variable U, calculate its corresponding marginal probability respectively, that is, the probability of the result variable S given the treatment variable T and the current confounding variable U , multiplied by the probability of the current confounding variable U appearing under the same given T condition ;

[0121] 3) Accumulate the marginal probabilities corresponding to each confounding variable U to obtain the association effect probability.

[0122] For a single product root node under a product node, the way to obtain its corresponding causal relationship probability is as Figure 5 shown, including:

[0123] S431. In the initial BNCT knowledge graph, determine the node group corresponding to the current product node. According to the causal association relationship, determine the treatment variable (T), the confounding variable (U) associated with the treatment variable, and the outcome variable (S) in the node group.

[0124] In the initial BNCT knowledge graph, according to the causal association relationship, determine the node group corresponding to the current product node; the node group is a combination of several nodes that have a causal association relationship with the current product node.

[0125] In the node group, first determine the treatment variable (T), that is, the root node of the current product; then, identify all confounding variables (U) that interact with the treatment variable (T) and can affect the outcome variable (S); finally, determine the outcome variable (S), that is, the characteristic variable formed under the joint action of T and U, which includes the current product node. By clarifying the causal association relationship between variables in the node group, it provides a structural basis for subsequent causal inference and knowledge graph update.

[0126] In a specific embodiment, the initial BNCT knowledge graph includes the following nodes:

[0127] Drug administration time interval, tissue density, neutron beam irradiation time, neutron beam energy, boron drug dosage, tumor boron concentration, tumor volume, tumor volume change rate, and tumor volume change value;

[0128] When the tumor volume change rate is the product node and the boron drug dosage is the product root node, based on the causal association relationship between the boron drug dosage and the tumor volume change rate, that is, in the causal logic process of the boron drug dosage (product root node) for the tumor volume change rate (product node), characteristic variables such as neutron beam irradiation time, neutron beam energy, tumor boron concentration, and tumor volume will affect this causal logic process; based on this, divide the neutron beam irradiation time, neutron beam energy, boron drug dosage, tumor boron concentration, tumor volume, and tumor volume change rate in the initial BNCT knowledge graph into a node group; and set the boron drug dosage as the treatment variable (T), set the neutron beam irradiation time and neutron beam energy as the first confounding variable (U1) and the first confounding variable (U2) respectively, and set the tumor volume concentration, tumor volume, and tumor volume change rate as the first outcome variable (S0), the second outcome variable (S1), and the third outcome variable (S2) respectively;

[0129] For easy understanding, define the characteristic variables included in the above node group as shown in Table 2 below:

[0130]

[0131] It should be noted that in different causal relationship chains, the same characteristic variable can be used as a summation node or a product node.

[0132] S432, constructing an association probability function between the outcome variable (S), the treatment variable (T) and the confounding variable (U) based on the causal association relationship contained in the node group;

[0133] The relationship probability function is a function used to characterize the effect of the treatment variable (T) on the outcome variable (S) under the influence of each confounding variable (U).

[0134] According to the causal relationship, the causal relationship between the treatment variable (T), confounding variable (U) and outcome variable (S) is sorted out, and the association probability function between each variable is constructed.

[0135] In one embodiment, according to the causal relationship between the boron drug dose and the tumor volume change rate, the conduction process of the causal relationship is sorted out as follows: Figure 6 The logic diagram shown in Figure 6 As shown in the figure, the boron dose affects the dose received by the tumor by affecting the tumor boron concentration, which ultimately affects the tumor volume change rate; and in the causal logic transmission process from the boron dose to the tumor boron concentration, it is also affected by the dosing time interval and tissue density.

[0136] According to the calculation process of TPS, the associated probability function of each treatment variable under the confounding variable for the outcome variable is constructed as follows:

[0137] Where P is the probability of association effect; is the probability of the association effect under the influence of the confounding factor on the treatment variable, is the probability of the associated effect of the treatment variable on the outcome variable under the influence of multiple interfering confounding factors; U1 is the neutron beam irradiation time, and U2 is the neutron beam energy;

[0138] In the formula, is the tumor volume change rate, which satisfies:

[0139]

[0140] in, is the tumor boron concentration influence coefficient, is the irradiation time influence coefficient; both are pre-set parameters; exemplary, , , where a negative sign indicates a reduction in tumor volume.

[0141] S433. Based on the sample data in the sample data subset, perform relationship probability calculation according to the association probability function to obtain the causal probability value corresponding to the product node group.

[0142] Specifically, in the sample data set, extract the sample data entries that simultaneously contain the outcome variable (S), the treatment variable (T), and the confounding variable (U) to construct the sample data subset of the node group corresponding to the current product node.

[0143] In the sample data subset, extract the sample data corresponding to the outcome variable (S), the treatment variable (T), and the confounding variable (U) in each of the sample data entries.

[0144] Based on the sample data corresponding to the outcome variable (S), the treatment variable (T), and the confounding variable (U) in the sample data subset, perform relationship probability calculation according to the association probability function to obtain the relationship probability value corresponding to each sample data entry.

[0145] Integrate the relationship probability values corresponding to each sample data entry to obtain the integrated value of the relationship probability, and use this integrated value as the causal probability value corresponding to the current product node group.

[0146] In a more specific embodiment, the sample data subset of the node group corresponding to the current product node is as shown in Table 3 below.

[0147] Table 3 Sample Data Subset and Causal Probability Value

[0148]

[0149] Based on the above sample data subset, calculate the causal probability value corresponding to the current product node as:

[0150]

[0151] That is:

[0152]

[0153] Conclusion: The above causal probability value is used to represent that when the boron drug dose is 15 mg / kg, considering all confounding variables, the probability that the tumor volume shrinks by more than 20% is 90.95%.

[0154] It should be noted thirdly that in the above formula, represents the probability value obtained by performing probability weighted summation over all possible values of all confounding variables for the case where the tumor volume shrinks by more than 20% ( ), which is used to eliminate the confounding effect of the confounding variable U to obtain P(S∣T).

[0155] S44. Synthesize the causal relationship probabilities of each of the summation root nodes and the causal relationship probabilities of each of the product root nodes to obtain a target BNCT knowledge graph.

[0156] Specifically, after obtaining the causal relationship probabilities of each of the summation root nodes and the causal relationship probabilities of each of the product root nodes, add each causal relationship probability to the initial BNCT knowledge graph to obtain a target BNCT knowledge graph.

[0157] It should be noted that in some alternative embodiments, the influence of confounding factors can be adjusted through stratified analysis or regression models. For example, estimate P(Treatment Outcome | Boron Dose, Neutron Flux, Tumor Type, Patient Age); and calculate the intervention effect using causal inference methods (such as do-calculus), for example, P(Treatment Outcome | do(Boron Dose = High)).

[0158] Since the accuracy of the constructed knowledge graph is positively correlated with the number of extracted treatment datasets, that is, the more treatment datasets are extracted, the higher the accuracy of the constructed knowledge graph. Based on this, in a specific embodiment, the method for obtaining the treatment datasets includes:

[0159] Extract treatment datasets related to the BNCT treatment of the target tumor from each historical treatment record or research material;

[0160] Among them, the historical treatment record is the case record corresponding to each BNCT historical treatment performed on the target tumor type; that is, a single historical treatment has a corresponding single case record, and the case record contains various types of data generated during the corresponding historical treatment process.

[0161] The research materials include information such as clinical trial reports, basic research papers, and review articles. Extract knowledge texts on aspects such as the boron drug uptake mechanism of the target tumor, the principle of cell damage caused by the interaction between boron drugs and neutrons, and the dose distribution law of different neutron sources in tumor tissues from each information.

[0162] Specifically, collect the clinical data of patients receiving treatment in the hospital medical record database or extract each research material from the medical database. Extract the diagnostic information of BNCT from each clinical data or research data, including: patient basic information, tumor characteristics, various parameters during the treatment process (such as boron drug concentration, neutron irradiation time, etc.) and treatment results (such as tumor recurrence situation, survival time, etc.).

[0163] Based on the same inventive concept, the present application also provides a method for predicting the treatment effect of BNCT based on a knowledge graph, which is used to predict the tumor treatment effect for a given BNCT treatment plan and obtain the prediction result of the treatment effect.

[0164] Among them, the BNCT treatment plan includes treatment parameters corresponding to the treatment plan; for example, neutron source type, neutron source intensity, boron drug dosage, irradiation angle, irradiation time, etc.; the tumor treatment effect includes the degree of tumor shrinkage and the probability of canceration of normal tissues, etc.

[0165] Please refer to Figure 7 , which shows a schematic flow chart of the method for predicting the treatment effect of BNCT based on a knowledge graph provided by an embodiment of the present application; as Figure 7 shown, this method includes the following steps:

[0166] S100, obtain a pre-constructed target BNCT knowledge graph;

[0167] In the embodiment, the construction method of the target BNCT knowledge graph is the same as the implementation method in the above embodiment, and will not be elaborated here.

[0168] S200, use sample data to perform graph update on the target BNCT knowledge graph to obtain an updated knowledge graph;

[0169] Obtain the latest treatment data set, based on the latest treatment data set;

[0170] Use each sample data to update the pre-constructed target BNCT knowledge graph and optimize each node parameter (such as weight and probability distribution); during the optimization process, use the EM algorithm or the gradient descent method to achieve the convergence effect of the optimization process.

[0171] Specifically, the implementation method of using each sample data to update the pre-constructed target BNCT knowledge graph and optimize each node parameter (such as weight and probability distribution) includes:

[0172] Obtain that the number of historical sample data entries in the historical treatment data set is M, and the number of the latest sample data entries in the latest treatment data set is N; merge the historical treatment data set and the latest treatment data set to obtain an updated treatment data set; in the updated treatment data set, obtain the proportion of historical sample data entries as the first weight, and obtain the proportion of the latest sample data entries as the second weight; based on the first weight and the second weight, update the causal relationship probability between each node in the target BNCT knowledge graph to obtain the updated causal relationship probability, which is

[0173] W c =AWa +BW b

[0174] Wherein, W c is the updated probability of causal relationship; W a is the probability of causal relationship corresponding to the historical treatment data set; W b is the probability of causal relationship corresponding to the latest treatment data set; A is the first weight; B is the first weight.

[0175] Wherein, the probability of causal relationship corresponding to the latest treatment data set is the probability of causal relationship obtained by using the method for constructing a knowledge graph based on BNCT provided in the above embodiment based on the latest treatment data set.

[0176] S300, based on the updated knowledge graph, obtain the tumor treatment effect of the patient under a given treatment plan.

[0177] Wherein, the given treatment plan is a treatment plan for a target tumor formulated in advance;

[0178] Specifically, based on the given treatment plan, extract the treatment plan data included in the treatment plan; based on the treatment plan data and combined with the patient data, construct a given data group; input the given data group into the pre-trained knowledge graph, and based on the knowledge graph, through search based on causal association relationships, obtain the tumor treatment effect corresponding to the given data group.

[0179] Exemplarily, in the trained knowledge graph, input the question "What is the probability of achieving a complete remission in the treatment effect under different neutron source parameters given a specific tumor type and boron drug characteristics", and use the SPN method to obtain the probability result of the tumor treatment effect by calculating the probabilities between relevant variable nodes, sum nodes, and product nodes, thereby realizing the inference of the treatment effect;

[0180] This inference ability can help doctors more accurately evaluate the causal impact of different factor combinations on the treatment effect when formulating treatment plans.

[0181] Please refer to Figure 8 , which is an optional hardware structure diagram of a terminal provided by an embodiment of the present invention. The terminal may be a mobile phone, a computer device, a tablet device, a personal digital processing device, a factory background processing device, etc. The terminal 60 includes: at least one processor 61, a memory 62, at least one network interface 64, and a user interface 63. Each component in the device is coupled together through a bus system 65. It can be understood that the bus system 65 is used to realize the connection and communication between these components. In addition to the data bus, the bus system also includes a power bus, a control bus, and a status signal bus.

[0182] Among them, the user interface 63 may include a display, a keyboard, a mouse, a trackball, a click gun, a button, a button, a touchpad, or a touch screen, etc.

[0183] It can be understood that the memory 62 can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM, Read Only Memory), a programmable read-only memory (PROM, Programmable Read-Only Memory), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM, Static Random Access Memory), synchronous static random access memory (SSRAM, Synchronous Static Random Access Memory). The memory characterized in the embodiments of the present invention is intended to include but not limited to these and any other suitable categories of memory.

[0184] The memory 62 in the embodiments of the present invention is used to store various categories of data to support the operation of the terminal. Examples of these data include: any executable program for operating on the terminal 60, such as the operating system 621 and the application program 622; the operating system 621 contains various system programs, such as the framework layer, the core library layer, the driver layer, etc., for implementing various basic services and processing hardware-based tasks. The application program 622 can include various application programs, such as a media player (Media Player), a browser (Browser), etc., for implementing various application services. Implementing the BNCT-based knowledge graph construction method or the BNCT treatment effect prediction method based on the knowledge graph provided in the embodiments of the present invention can be included in the application program 622.

[0185] The method disclosed in the embodiments of the present invention above can be applied to or implemented by the processor 61. The processor 61 may be an integrated circuit chip with signal processing capabilities. During implementation, the steps of the above method can be completed by the integrated logic circuit in hardware or instructions in software form in the processor 61. The above processor may be a general-purpose processor, a digital signal processor (DSP, Digital Signal Processor), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 61 can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor 61 may be a microprocessor or any conventional processor, etc. Combining the steps of the accessory optimization method provided in the embodiments of the present invention can be directly embodied as being executed and completed by the hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software module may be located in the storage medium, and this storage medium is located in the memory. The processor reads the information in the memory and combines its hardware to complete the steps of the foregoing method.

[0186] In an exemplary embodiment, the terminal 60 may be one or more application-specific integrated circuits (ASICs, Application Specific Integrated Circuit), DSPs, programmable logic devices (PLDs, Programmable Logic Device), complex programmable logic devices (CPLDs, Complex Programmable Logic Device) for executing the foregoing method.

[0187] The embodiments of the present invention also provide a computer-readable storage medium, on which a computer program is stored. When the program is called by a processor, it implements the method for constructing a knowledge graph based on BNCT or the method for predicting the treatment effect of BNCT based on the knowledge graph provided by the present invention.

[0188] Among them, the computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. The computer-readable storage medium may be, for example (but not limited to), an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: portable computer disks, hard disks, random access memories (RAMs), read-only memories (ROMs), erasable programmable read-only memories (EPROMs or flash memories), static random access memories (SRAMs), portable compact disk read-only memories (CD-ROMs), digital versatile disks (DVDs), memory sticks, floppy disks, mechanical encoding devices.

[0189] The computer-readable programs characterized herein can be downloaded from a computer-readable storage medium to various computing / processing devices, or downloaded to an external computer or external storage device through a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network adapter or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0190] In summary, the method for constructing a BNCT-based knowledge graph, the method for predicting the BNCT treatment effect, the terminal, and the medium provided in this application integrate the treatment plan data and patient data systematically by constructing a target BNCT knowledge graph based on causal association relationships. It not only converts fragmented knowledge into a structured relationship network but also infers the treatment effect corresponding to the treatment plan through knowledge reasoning, thereby assisting doctors in predicting the treatment effects corresponding to different treatment plans. Moreover, by constructing a dynamic knowledge update mechanism to update the target BNCT knowledge graph based on the latest treatment sample data, the accuracy of the knowledge graph is further improved, and thus the accuracy of personalized treatment plan design is greatly enhanced.

[0191] The above embodiments merely illustrate the principles and effects of this application and are not used to limit this application. Any person familiar with this technology can modify or change the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or changes completed by those with ordinary knowledge in the technical field without departing from the spirit and technical ideas disclosed in this application should still be covered by the claims of this application.

Claims

1. A method for constructing a knowledge graph based on BNCT, comprising: Obtaining a treatment dataset of a target tumor in historical BNCT treatments; The treatment dataset contains treatment plan data and treatment effect data in each historical treatment; Based on the causal association relationships of the data in the treatment dataset in BNCT treatments, constructing an initial BNCT knowledge graph and constructing a corresponding sample dataset for the initial BNCT knowledge graph; wherein, the causal association relationships include weighted causal relationships and joint causal relationships; Among the nodes corresponding to the treatment plan data, extracting the nodes that satisfy the weighted causal relationship as summation nodes; and among the nodes corresponding to the treatment effect data, extracting the nodes that satisfy the joint causal relationship as product nodes; According to the sample dataset, using the sum-product network method, obtaining the causal relationship probabilities corresponding to each summation node and obtaining the causal relationship probabilities corresponding to the product nodes to obtain the target BNCT knowledge graph.

2. The method for constructing a knowledge graph based on BNCT according to claim 1, wherein The construction of the sample dataset corresponding to the initial BNCT knowledge graph includes: Based on the node information in the initial BNCT knowledge graph, extracting the feature variables corresponding to the node information from the treatment dataset; Constructing each feature variable belonging to the same treatment case into a sample data entry; Performing the construction of each sample data entry on the treatment dataset to obtain the sample dataset corresponding to the initial BNCT knowledge graph.

3. The method for constructing a knowledge graph based on BNCT according to claim 2, wherein The implementation manner of using the sum-product network method to obtain the causal relationship probabilities corresponding to each summation node and obtaining the causal relationship probabilities corresponding to the product nodes includes: Adopting a calculation method of frequency weights to obtain the causal relationship probability of the summation root node corresponding to the summation node; Adopting a calculation method of association effect probability to obtain the causal relationship probability of the product root node corresponding to the product node.

4. The method for constructing a knowledge graph based on BNCT according to claim 3, wherein The obtaining manner of adopting the calculation method of frequency weights to obtain the causal relationship probability of the summation root node corresponding to the summation node includes: In the sample dataset, extracting the sample subset corresponding to the current summation node and obtaining the number of sample data entries in the sample subset as the total number of samples of the current summation node; In the sample subset, extracting the current sample quantity corresponding to the current summation root node; Based on the current sample quantity and the total number of samples, calculating the ratio between the current sample data and the total number of samples as the causal relationship probability corresponding to the current summation root node.

5. The method for constructing a knowledge graph based on BNCT according to claim 3, wherein The obtaining manner of adopting the calculation method of association effect probability to obtain the causal relationship probability of the product root node corresponding to the product node includes: In the initial BNCT knowledge graph, determining the node group corresponding to the current product node; according to the causal association relationship, determining the treatment variable, confounding variable and result variable in the node group; Based on the causal association relationship, constructing an association probability function between the result variable and the treatment variable and the confounding variable; Based on the sample data in the sample data subset, performing relationship probability calculation according to the association probability function to obtain the causal probability value corresponding to the product node group.

6. The method for constructing a knowledge graph based on BNCT according to claim 1, wherein The extraction method of the summation node includes; For a single node among the nodes corresponding to the treatment plan data, determine whether it is a child node; When the node is a child node, determine whether there is a weighted causal relationship between each root node corresponding to the current child node and the current child node. If so, label the child node as a summation node; And, the extraction method of the product node includes: For a single node among the nodes corresponding to the treatment effect data, determine whether the node is a child node; When the node is a child node, determine whether there is a joint causal relationship between each root node corresponding to the current child node and the current child node. If so, label the child node as a product node.

7. The method for constructing a knowledge graph based on BNCT according to claim 5, wherein When the product node is the tumor volume change rate and the product root node is the boron drug dosage, the node group includes: neutron beam irradiation time, neutron beam energy, boron drug dosage, tumor boron concentration, tumor volume, and tumor volume change rate; Set the boron drug dosage processing variable, set the neutron beam irradiation time and neutron beam energy as confounding variables respectively, and set the tumor volume concentration, tumor volume, and tumor volume change rate as result variables respectively; Construct the association probability function between the result variable and the processing variable and the confounding variable, which is: Where P is the probability of the association effect; is the probability of the association effect under the influence of the interfering confounding factor on the treatment variable, is the probability of the association effect of the treatment variable on the outcome variable under the influence of multiple interfering confounding factors; U1 is the neutron beam irradiation time, and U2 is the neutron beam energy; S2 is the tumor volume change rate; the tumor volume change rate is represented by a weighted combination or functional relationship of the tumor volume concentration and the tumor volume, and satisfies: ； Among them, is the tumor boron concentration influence coefficient, is the irradiation time influence coefficient; S0 is the tumor boron concentration; S1 is the tumor volume.

8. A method for predicting the BNCT treatment effect based on a knowledge graph, including: Obtain a pre-constructed target BNCT knowledge graph; Use sample data to train the target BNCT knowledge graph to obtain a trained knowledge graph; Based on the trained knowledge graph, obtain the tumor treatment effect of a patient under a given treatment plan; Wherein, the target BNCT knowledge graph is obtained by using the knowledge graph construction method based on BNCT described in any one of claims 1 to 7.

9. A terminal, characterized in that, Including: A processor and a memory, and the memory is communicatively connected to the processor; The memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory so that the terminal executes the knowledge graph construction method based on BNCT described in any one of claims 1 to 7, or executes the method for predicting the BNCT treatment effect based on the knowledge graph described in claim 8.

10. A computer storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the knowledge graph construction method based on BNCT described in any one of claims 1 to 7, or implements the method for predicting the BNCT treatment effect based on the knowledge graph described in claim 8.

Images