A teaching method for clinical decision-making of warm disease based on big data knowledge graph

Abstract

This invention discloses a teaching method for clinical decision-making in febrile diseases based on a big data knowledge graph, specifically relating to the field of medical information processing technology. It involves constructing a knowledge graph of febrile diseases, introducing data on the transmission rules of Wei Qi Ying Xue and the San Jiao, and performing joint constraint modeling on the transmission constraints between syndrome elements to form a dynamic transmission relationship set of febrile disease syndromes. Based on this set, the initial syndrome state of teaching cases is temporally extrapolated to generate multiple attainable syndrome evolution paths. Path consistency analysis and path divergence identification are then performed to determine the key decision-making nodes in the clinical process of febrile diseases. Based on the key decision-making node locations and the transmission conditions of the syndrome evolution paths, treatment elements and prescription elements are matched and extrapolated to form clinical decision-making teaching paths. These paths are then organized according to the sequence of syndrome evolution, outputting extrapolable clinical decision-making process data for teaching and training.

Description

Technical Field

[0001] This invention relates to the field of medical information processing technology, and more specifically, to a teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs. Background Technology

[0002] Teaching clinical decision-making in febrile diseases involves syndrome differentiation, treatment selection, and prescription determination. The relevant knowledge is scattered across clinical case records, classical texts on febrile diseases, and standard teaching materials. While existing teaching aids include tools based on case databases, text databases, or structured knowledge compilation, they primarily present static knowledge points or explain fixed case conclusions. The teaching process fails to adequately express the temporal, directional, and uncertain aspects of syndrome progression in febrile disease clinical thinking, making it difficult to consistently cover the key judgment stages at different phases of clinical febrile disease treatment.

[0003] Existing clinical decision-making teaching methods for febrile diseases are insufficient to express the transmission process of syndrome elements in a deductive manner, and to form a reproducible decision-making path organization method when the syndrome evolution path diverges, so as to support the continuous training of clinical decision-making reasoning in the teaching process. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs to address the problems raised in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs includes the following steps: S1: Obtain clinical case data, literature data, and teaching standard data of febrile diseases, perform semantic normalization processing, and construct a knowledge graph of febrile diseases that includes syndrome elements, treatment elements, and prescription elements; S2: Based on the knowledge graph of febrile diseases, the transmission rules data of Wei Qi Ying Xue and the transmission rules data of San Jiao are introduced to jointly constrain the sequential relationship, evolution direction relationship and occurrence probability relationship among syndrome elements, and generate a set of dynamic transmission relationships of febrile disease syndromes. S3: Based on the dynamic transmission relationship set of febrile disease syndromes, the initial syndrome state of teaching cases is temporally deduced to generate multiple attainable syndrome evolution paths; S4: Conduct path consistency analysis and path divergence identification on multiple syndrome evolution paths to determine the key decision-making nodes in the clinical process of febrile diseases. S5: For key decision-making nodes, combined with the transmission conditions of syndrome evolution path, match and deduce the treatment elements and prescription elements to generate corresponding clinical decision-making teaching paths. S6: Organize the clinical decision-making teaching path according to the sequence of syndrome evolution, and output data on the inferable clinical decision-making process for teaching and training.

[0006] In a preferred embodiment, S1 specifically refers to: Acquire clinical case data, literature entries, and teaching standard data related to febrile diseases; Semantic normalization processing was performed on clinical case data, classical literature data, and teaching standard data of febrile diseases. Establish physical identifiers for syndrome elements, treatment elements, and prescription elements; Based on entity identifiers, the relationships between syndrome elements and treatment elements, between treatment elements and prescription elements, and between syndrome elements and prescription elements are extracted. A knowledge graph of febrile diseases is constructed based on syndrome elements, treatment elements, prescription elements, and their relationships.

[0007] In a preferred embodiment, the transmission constraint relationship includes the sequential relationship, evolution direction relationship, and occurrence probability relationship among the symptom elements.

[0008] In a preferred embodiment, S2 specifically refers to: Based on the knowledge graph of febrile diseases, the data on the transmission rules of Wei Qi Ying Xue and the data on the transmission rules of San Jiao are analyzed into transmission constraint rules between syndrome elements. Based on the rules of transmission constraint, the order of syndrome elements is constrained and labeled to form the temporal relationship data of syndrome elements; Based on the rules of transmission and transformation constraints, the evolution direction between syndrome elements is constrained and labeled to form directional relationship data of syndrome elements; Based on the transmission constraint rules, the occurrence probability of syndrome elements under different transmission conditions is marked to form the occurrence probability relationship data of syndrome elements; By associating temporal relationship data, directional relationship data, and probability relationship data with syndrome elements in the knowledge graph of febrile diseases, a set of dynamic transmission relationships of febrile disease syndromes is constructed.

[0009] In a preferred embodiment, the transmission constraint rules describe the constraints that should be satisfied between the symptom element entities before and after the transmission, as well as the clinicopathological feature changes allowed by the transmission.

[0010] In a preferred embodiment, S3 specifically refers to: The initial syndrome elements corresponding to the teaching cases are obtained as the initial syndrome state, and the initial syndrome state is mapped to the corresponding syndrome element node in the set of dynamic transmission relationship of febrile disease syndromes. Based on the temporal relationship data of syndrome elements, the order of occurrence of syndrome elements in the deduction process is limited; Based on the directional relationship data of syndrome elements, the evolution direction of syndrome elements in the deduction process is limited; Based on the probability relationship data of the occurrence of syndrome elements, conditional constraint judgments are made on the evolution path of syndrome elements that satisfy the temporal and directional relationships. Based on the joint constraint results, the initial syndrome state is deduced, generating multiple reachable syndrome evolution paths and corresponding transmission conditions.

[0011] In a preferred embodiment, S4 specifically refers to: Path consistency analysis was performed on multiple reachable symptom evolution paths to obtain path segments with consistent evolution order of symptom elements in multiple reachable symptom evolution paths, and these segments were marked as path consistency segments. Path divergence identification is performed on the syndrome element nodes at the termination position of the path consistency segment. The syndrome element nodes at the termination position of the path consistency segment with two or more different evolution directions are identified as key decision nodes in the clinical process of febrile diseases.

[0012] In a preferred embodiment, S5 specifically refers to: Match the syndrome element nodes corresponding to the key decision nodes and the transmission conditions recorded in multiple reachable syndrome evolution paths with syndrome elements, treatment elements, and prescription elements. Based on the syndrome element nodes and their corresponding transmission conditions, and according to the relationship between syndrome elements and treatment elements, the treatment elements that match the syndrome element nodes are determined. Based on the relationship between treatment principles and prescription elements, determine the prescription elements that match the treatment principles. Based on the matching relationship between syndrome element nodes, determined treatment elements, and prescription elements, a clinical decision-making teaching path corresponding to the key decision node positions is generated.

[0013] In a preferred embodiment, S6 specifically refers to: The clinical decision-making teaching path is sorted according to the temporal order of the syndrome element nodes in multiple accessible syndrome evolution paths. The syndrome element nodes, treatment element, and prescription element in the sorted clinical decision-making teaching path are organized into a continuous decision sequence according to the evolution order of the syndrome element nodes. The corresponding transmission conditions are associated with the corresponding syndrome element nodes in the decision sequence to generate clinical decision process data. Clinical decision-making process data is stored in a teaching case database, serving as deductive clinical decision-making process data for teaching and training.

[0014] In a preferred embodiment, the clinical decision-making process data includes syndrome element nodes, treatment element, prescription element, and transmission condition.

[0015] The technical effects and advantages of this invention's teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs: By acquiring clinical case data, literature data, and teaching standard data of febrile diseases and performing semantic normalization, a unified knowledge graph of febrile diseases is formed, incorporating syndrome elements, treatment elements, and prescription elements, thus reducing the knowledge disconnect caused by the dispersion of teaching resources. By introducing data on the transmission rules of Wei Qi Ying Xue and the San Jiao, joint constraint modeling is performed on the transmission constraints between syndrome elements, forming a dynamic transmission relationship set of febrile disease syndromes. This allows the temporal and directional nature of syndrome transmission to be structurally expressed in the teaching data. Furthermore, by temporally extrapolating the initial syndrome state of teaching cases based on the dynamic transmission relationship set of febrile disease syndromes, multiple attainable syndrome evolution paths are generated. This approach provides a traceable evolutionary trajectory for teaching cases; by analyzing the consistency and identifying divergences of multiple syndrome evolution paths, it identifies key decision-making nodes in the clinical process of febrile diseases, enabling the teaching process to organize reasoning training around these key judgment nodes; by combining the transmission conditions of syndrome evolution paths to match and deduce treatment and prescription elements and generate clinical decision-making teaching paths, it enables treatment and prescription elements to form a reproducible decision chain with syndrome evolution paths; by organizing clinical decision-making teaching paths according to the order of syndrome evolution and outputting data that can be deduced from clinical decision-making processes, it supports the continuous presentation and interactive deduction of clinical decision-making processes in teaching training. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of a clinical decision-making teaching method for febrile diseases based on big data knowledge graphs, according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention. Example

[0018] Figure 1 This invention presents a teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs, which includes the following steps: S1: Obtain clinical case data, literature data, and teaching standard data of febrile diseases, perform semantic normalization processing, and construct a knowledge graph of febrile diseases that includes syndrome elements, treatment elements, and prescription elements; S2: Based on the knowledge graph of febrile diseases, the transmission rules data of Wei Qi Ying Xue and the transmission rules data of San Jiao are introduced to jointly constrain the sequential relationship, evolution direction relationship and occurrence probability relationship among syndrome elements, and generate a set of dynamic transmission relationships of febrile disease syndromes. S3: Based on the dynamic transmission relationship set of febrile disease syndromes, the initial syndrome state of teaching cases is temporally deduced to generate multiple attainable syndrome evolution paths; S4: Conduct path consistency analysis and path divergence identification on multiple syndrome evolution paths to determine the key decision-making nodes in the clinical process of febrile diseases. S5: For key decision-making nodes, combined with the transmission conditions of syndrome evolution path, match and deduce the treatment elements and prescription elements to generate corresponding clinical decision-making teaching paths. S6: Organize the clinical decision-making teaching path according to the sequence of syndrome evolution, and output data on the inferable clinical decision-making process for teaching and training.

[0019] S1: Acquire clinical case data, literature data, and teaching standard data related to febrile diseases, perform semantic normalization processing, and construct a knowledge graph of febrile diseases containing syndrome elements, treatment elements, and prescription elements, including: Acquire clinical case data, literature entries, and teaching standard data related to febrile diseases; Specifically, clinical case data in the study of febrile diseases includes patient information from the four diagnostic methods, diagnostic results, case progression, physician prescription records, and patient outcomes; literature data includes descriptions of symptoms, treatment methods, and prescriptions in classic works on febrile diseases such as "Wenbing Tiaobian," "Wenre Jingwei," and "Wenbing Xue"; and teaching standard data includes teaching syllabi, teaching guides, and guidelines or standards for the diagnosis and treatment of febrile diseases in higher education institutions of traditional Chinese medicine.

[0020] Semantic normalization processing was performed on clinical case data, classical literature data, and teaching standard data of febrile diseases. Specifically, Chinese word segmentation was performed on clinical case data, literature entries, and teaching standard data related to febrile diseases. For example, a forward maximum matching Chinese word segmentation method based on a medical dictionary was used to decompose each original text data into a standard sequence of medical semantic terms. The segmentation results were then processed for synonym normalization. By constructing and using a thesaurus of febrile disease terminology, words with different expressions but semantically equivalent meanings were uniformly mapped to standardized medical terms. For example, terms such as "Wei Biao Re Xie" and "Wei Fen Re Xie" were uniformly mapped to "Wei Fen Re Xie". Finally, entity name standardization was performed on the data after the synonym normalization process, that is, for each semantic term sequence, the entity names appearing in the sequence were standardized. The system performs entity boundary identification and entity type labeling for syndrome names, treatment names, and prescription names. For example, it uses medical named entity recognition technology and pre-trained conditional random field algorithms to automatically label each entity as a syndrome element entity, treatment element entity, or prescription element entity. For ambiguities that exist after entity boundary identification and type labeling, a context window-based conditional probability model is used to resolve ambiguities. By calculating the conditional probability value between the context semantic feature vector and the candidate entity feature vector, the candidate entity with the highest probability value is selected as the final entity labeling result, thereby ensuring the accuracy and uniqueness of entity labeling.

[0021] Establish physical identifiers for syndrome elements, treatment elements, and prescription elements; Specifically, entity identifier data tables are constructed for the syndrome element entities, treatment element entities, and prescription element entities obtained after semantic normalization. Each entity identifier data table uses the entity name as the primary key and assigns a unique identifier code to each entity name. For example, the syndrome element entity Wei Fen Re Xie is assigned the identifier code ZH001, the treatment element entity Xin Liang Jie Biao is assigned the identifier code ZF001, and the prescription element entity Yin Qiao San is assigned the identifier code FY001, thereby achieving standardized identification of entities.

[0022] Based on entity identifiers, the relationships between syndrome elements and treatment elements, between treatment elements and prescription elements, and between syndrome elements and prescription elements are extracted. Specifically, entity relationship templates are established using semantically normalized document data and teaching standard data, combined with entity identifiers. These templates define the allowed relationship types between entities, such as the corresponding treatment method relationship between syndrome elements and treatment method elements, the recommended prescription relationship between treatment method elements and prescription elements, and the applicable prescription relationship between syndrome elements and prescription elements. Based on the entity relationship templates, association relationships are extracted through a combination of entity co-occurrence frequency statistics and conditional probability calculations between entities. Entity co-occurrence frequency statistics specifically involve counting the occurrences of each entity in the semantically normalized text data. If the frequency of a pair of entities appearing together exceeds a preset frequency threshold, a valid association is determined between the entities. For example, the preset frequency threshold is set to 5 times. The conditional probability calculation between entities is as follows: the conditional probability is calculated based on the number of times entity A and entity B appear together and the number of times entity A appears alone. If the conditional probability exceeds a preset conditional probability threshold, a valid association is determined between the entities. For example, the preset conditional probability threshold is set to 0.6. The above methods are used to extract and confirm the association between syndrome elements and treatment elements, treatment elements and prescription elements, and syndrome elements and prescription elements.

[0023] A knowledge graph of febrile diseases is constructed based on syndrome elements, treatment elements, prescription elements, and their relationships. Specifically, entity nodes for syndrome elements, treatment methods, and prescriptions are created separately. Each node is distinguished and identified by an identifier code and is accompanied by entity description information as a node attribute. For example, the syndrome element entity node ZH001 is described as having heat evil in the Wei level. Edges are created between entity nodes, and the relationship type of the edge is determined. For example, the edge from a syndrome element entity node to a treatment method entity node is marked as the corresponding treatment method, the edge from a treatment method entity node to a prescription element entity node is marked as the recommended prescription, and the edge from a syndrome element entity node to a prescription element entity node is marked as the applicable prescription. An association strength attribute is set for the edge, which is the entity co-occurrence frequency and conditional probability value. For example, when the entity co-occurrence frequency is 15 times and the conditional probability value is 0.8, the association strength attribute is marked as co-occurrence 15 | probability 0.8, thus forming a knowledge graph of febrile diseases that includes syndrome element nodes, treatment method element nodes, prescription element nodes, and corresponding association relationships.

[0024] S2: Based on the knowledge graph of febrile diseases, this study introduces the transmission rules data of Wei Qi Ying Xue and the transmission rules data of the San Jiao, and performs joint constraint modeling on the sequential relationship, evolution direction relationship, and occurrence probability relationship among syndrome elements to generate a dynamic transmission relationship set of febrile disease syndromes, including: Based on the knowledge graph of febrile diseases, the data on the transmission rules of Wei Qi Ying Xue and the data on the transmission rules of San Jiao are analyzed into transmission constraint rules between syndrome elements. Specifically, the data of the transmission rules of defensive qi, qi phase, nutrient phase, and blood phase are the theoretical text data of warm disease theory that record the syndrome transmission relationships at different stages among the defensive qi, qi phase, nutrient phase, and blood phase. The data of the transmission rules of the triple energizer are the theoretical text data of warm disease theory that record the syndrome transmission relationships at different stages among the upper energizer, middle energizer, and lower energizer. Semantic analysis is performed on each item of the data of the transmission rules of defensive qi, qi phase, nutrient phase, and blood phase and the data of the transmission rules of the triple energizer. The forward maximum matching Chinese word segmentation method based on a medical dictionary is used to segment the text of the transmission rule data. Then, through entity name standardization processing technology and medical named entity recognition technology, the syndrome element entities appearing in each text are identified, and the transmission constraint rules between the syndrome element entities included in each transmission rule text are constructed. That is, each transmission constraint rule describes the constraint conditions that should be satisfied between the syndrome element entities before and after transmission and the clinical pathological feature change conditions allowed for transmission. For example, the text of the data of the transmission rules of defensive qi, qi phase, nutrient phase, and blood phase "when pathogenic factors enter the defensive qi and are not resolved, they enter the qi phase" is parsed into a transmission constraint rule: the syndrome element entity of heat pathogen in the defensive qi transforms into the syndrome element entity of excessive heat in the qi phase, and the constraint condition is that the duration of the defensive qi syndrome is greater than 2 days and the heat symptoms are not relieved. The text of the data of the transmission rules of the triple energizer "if the syndrome of the upper energizer is not resolved, the disease can be transmitted downward to the middle energizer" is parsed into a transmission constraint rule: the syndrome element entity of lung heat in the upper energizer transforms into the syndrome element entity of stomach heat in the middle energizer, and the constraint condition is that the symptoms of the upper energizer persist and the clinical manifestation of thirst and desire to drink appears. All the parsed transmission constraint rules are stored in the transmission rule data table.

[0025] According to the transmission constraint rules, the order sequence between syndrome elements is constrained and marked to form the sequence relationship data of syndrome elements. Specifically, traverse all the recorded transmission constraint rules in the transmission rule data table, extract the two syndrome element entities before and after transmission in each transmission constraint rule, determine the pre-transmission syndrome element entity as the starting node in the sequence relationship data, and determine the post-transmission syndrome element entity as the successor node in the sequence relationship data. The starting node and the successor node are marked with a directed relationship. If the same pair of syndrome element entities appears in different sequence orders in multiple transmission constraint rules, the sequence order with the highest frequency of occurrence in the transmission rule data table of each transmission constraint rule is used as the marking of the sequence relationship data. For example, if the transmission of heat pathogen in the defensive qi to excessive heat in the qi phase appears 10 times in different rule data texts and the reverse transmission of excessive heat in the qi phase to heat pathogen in the defensive qi appears 2 times, then the heat pathogen in the defensive qi is used as the starting node and the excessive heat in the qi phase is used as the successor node to mark the directed relationship of the sequence relationship data. Through the above method, a clear sequence relationship between all syndrome element entities is established, and finally the sequence relationship data of syndrome elements is formed.

[0026] According to the transmission constraint rules, the evolution direction between syndrome elements is constrained and marked to form the direction relationship data of syndrome elements. Specifically, the process iterates through all the transmission constraint rules recorded in the transmission rule data table. For each transmission constraint rule, the directional relationship between the initial syndrome element and the subsequent syndrome element is determined. The directional relationship type recorded in the transmission constraint rule data is marked as one of four types: forward transmission, reverse transmission, combined disease, or concurrent disease. If the rule data describes the evolution from a shallower syndrome element to a deeper syndrome element, it is marked as forward transmission, such as heat evil in the Wei level to heat excess in the Qi level; otherwise, it is marked as reverse transmission, such as heat excess in the Qi level reverse transmission to heat evil in the Wei level. If it describes two or more syndrome elements existing simultaneously and their conditions influencing each other, it is marked as combined disease, such as heat evil in the Wei level and damp heat in the Qi level. If two or more syndrome elements appear simultaneously but their conditions are independent, it is marked as concurrent disease, such as heat evil in the Wei level and cold dampness in the middle Jiao existing independently. Through these markings, the directional relationship data between syndrome elements is formed.

[0027] Based on the transmission constraint rules, the occurrence probability of syndrome elements under different transmission conditions is marked to form the occurrence probability relationship data of syndrome elements; Specifically, the probability of occurrence is obtained by performing historical statistical analysis on all clinical case data of febrile diseases. Based on the temporal and directional relationship data of syndrome elements, the number of cases in which the initial syndrome element entity successfully transforms into the subsequent syndrome element entity under specific transmission constraint rules is counted. This number is then divided by the total number of cases in which the initial syndrome element entity appears to obtain the probability of occurrence of each pair of syndrome elements under different transmission conditions. For example, Wei-level heat evil appears in 500 cases under the condition that the duration is greater than 2 days and the heat symptoms have not been relieved. Among them, 400 cases successfully transform into Qi-level heat excess. The probability relationship data of Wei-level heat evil to Qi-level heat excess is 0.8. Through the above methods, statistical analysis and probability labeling are performed on all syndrome elements involved in the transmission constraint rules to form the probability relationship data of syndrome elements.

[0028] By associating temporal relationship data, directional relationship data, and occurrence probability relationship data with syndrome elements in the knowledge graph of febrile diseases, a set of dynamic transmission relationships of febrile disease syndromes is constructed. Specifically, the entity nodes of syndrome elements in the knowledge graph of febrile diseases are used as the main nodes of transmission relationship. The temporal relationship data is established with the corresponding syndrome element entity nodes in the form of directed edges. The directional relationship data and the occurrence probability relationship data are used as the attribute information of the edges. For example, the directed edge between the node Wei-level heat evil and the node Qi-level heat excess is marked as a forward transmission relationship, and the occurrence probability relationship data 0.8 is marked with the occurrence probability 0.8 attribute to the corresponding edge. Through the above method, the dynamic transmission relationship between syndrome element entity nodes with temporal, directional and occurrence probability constraints is recorded, and finally a set of dynamic transmission relationships of febrile disease syndromes is formed.

[0029] S3: Based on the dynamic transmission relationship set of febrile disease syndromes, the initial syndrome state of teaching cases is temporally deduced to generate multiple attainable syndrome evolution paths, including: The initial syndrome elements corresponding to the teaching cases are obtained as the initial syndrome state, and the initial syndrome state is mapped to the corresponding syndrome element node in the set of dynamic transmission relationship of febrile disease syndromes. Specifically, the initial syndrome elements include the initial symptoms and signs recorded in the patient's four diagnostic methods data of the teaching case, as well as the initial syndrome differentiation conclusion recorded in the diagnostic results data. For example, if the initial condition recorded in the teaching case is chills, fever, sore throat, and floating and rapid pulse, and the diagnosis is determined to be the syndrome element entity Wei-level heat evil, then the syndrome element entity Wei-level heat evil is taken as the initial syndrome state. The initial syndrome state is mapped to the corresponding syndrome element node in the set of dynamic transmission relationships of febrile disease syndromes, that is, the syndrome element entity node that matches the initial syndrome state is found in the set of dynamic transmission relationships of febrile disease syndromes. For example, the syndrome element entity Wei-level heat evil is mapped to the node Wei-level heat evil in the set of dynamic transmission relationships of febrile disease syndromes, thereby establishing a one-to-one correspondence between the initial syndrome state and the syndrome element node.

[0030] Based on the temporal relationship data of syndrome elements, the order of occurrence of syndrome elements in the deduction process is limited; Specifically, starting from the syndrome element node corresponding to the initial syndrome state, based on the temporal relationship data in the dynamic transmission relationship set of febrile disease syndromes, the syndrome element entity node is extended and deduced along the directed edge direction specified by the temporal relationship data, starting from the starting point. During the deduction process, if the syndrome element entity node does not have a successor node that satisfies the temporal relationship data, then the syndrome element entity node is taken as the endpoint of the current syndrome element evolution path. For example, if the syndrome element entity node Wei Fen Heat Evil has a successor node Qi Fen Heat Prosperity according to the temporal relationship data, then the syndrome element entity node is extended and deduced temporally along the directed edge Wei Fen Heat Evil - Qi Fen Heat Prosperity.

[0031] Based on the directional relationship data of syndrome elements, the evolution direction of syndrome elements in the deduction process is limited; Specifically, during the temporal extension deduction process, for each deduction step, the type of the next allowed syndrome element entity node is determined based on the directional relationship data in the dynamic transmission relationship set of febrile disease syndromes. When the directional relationship data is marked as forward transmission, deduction is only allowed from shallow syndrome element entity nodes to deep syndrome element entity nodes. When the directional relationship data is marked as reverse transmission, deduction is allowed from deep syndrome element entity nodes to shallow syndrome element entity nodes. If the directional relationship data is marked as combined disease, multiple mutually influential syndrome element entity nodes are allowed to exist simultaneously at the same temporal position. If the directional relationship data is marked as concurrent disease, multiple syndrome element entity nodes are allowed to exist simultaneously in parallel and independently. For example, if the directional relationship data of the syndrome element entity node Wei-level heat evil corresponding to Qi-level heat excess is marked as forward transmission, then the deduction process is limited to the forward transmission direction from Wei-level heat evil to Qi-level heat excess, and cannot be deduced in reverse or simultaneously in parallel.

[0032] Based on the probability relationship data of the occurrence of syndrome elements, conditional constraint judgments are made on the evolution path of syndrome elements that satisfy the temporal and directional relationships. Specifically, for each syndrome element entity node to be deduced, based on the occurrence probability relationship data recorded in the dynamic transmission relationship set of febrile disease syndromes, the occurrence probability between the syndrome element entity node and the previous syndrome element entity node in the current syndrome element evolution path is compared with a preset probability threshold to determine whether it can be deduced to the syndrome element entity node to be deduced. The probability threshold is set based on the statistical analysis of historical case deduction path data, and the weighted average method is used to calculate the weighted average of the historical occurrence probabilities of different syndrome element evolution paths. For example, if the preset probability threshold is set to 0.6, and the occurrence probability relationship data of the Wei level heat evil to Qi level heat excess in the current deduction is 0.8, which is greater than the preset probability threshold, then it is determined that the condition constraint requirement is met, and the deduction path is allowed to continue to the Qi level heat excess; if the occurrence probability is lower than 0.6, then it is determined that the condition constraint requirement is not met, and the current deduction path is terminated.

[0033] Based on the joint constraint results, the initial syndrome state is deduced to generate multiple reachable syndrome evolution paths and corresponding transmission conditions; Specifically, the joint constraint result is the set of constraints that simultaneously satisfy the temporal relationship data, directional relationship data, and occurrence probability relationship data of the syndrome elements. During the deduction process, the syndrome element node corresponding to the initial syndrome state is used as the initial node, and the path is traversed through a depth-first search algorithm. The above three types of constraints are satisfied at each path expansion. When the end point of the path is found or there is no feasible node that satisfies the joint constraint result, the current path is recorded as an attainable syndrome evolution path, and the transmission conditions involved in each step of the attainable syndrome evolution path are recorded. For example, starting from the initial syndrome state of Wei-level heat evil, passing through Qi-level heat excess to Ying-level heat blazing and satisfying all joint constraint results, the syndrome evolution path Wei-level heat evil - Qi-level heat excess - Ying-level heat blazing is generated and recorded. At the same time, the corresponding transmission conditions are recorded respectively: the duration of Wei-level syndrome is greater than 2 days and the heat symptoms have not been relieved, and the Qi-level heat symptoms persist and symptoms of irritability and restlessness appear. By traversing all paths that meet the joint constraint results, multiple attainable syndrome evolution paths and their corresponding transmission conditions are obtained.

[0034] S4: Conduct path consistency analysis and path divergence identification on multiple syndrome evolution paths to determine the key decision-making nodes in the clinical process of febrile diseases, including: Path consistency analysis was performed on multiple reachable symptom evolution paths to obtain path segments with consistent evolution order of symptom elements in multiple reachable symptom evolution paths, and these segments were marked as path consistency segments. Specifically, path consistency analysis involves comparing the sequence of symptom element entity nodes in each reachable symptom evolution path; using a sequence matching algorithm, determining node by node and sequence number whether the symptom element entity nodes between different reachable symptom evolution paths have the same evolution order; taking the starting symptom element entity node in each reachable symptom evolution path as the comparison starting point, and performing synchronous position comparison on multiple reachable symptom evolution paths according to the arrangement sequence number of the symptom element entity nodes in the reachable symptom evolution paths; when the symptom element entity nodes at the same arrangement sequence number are consistent in multiple reachable symptom evolution paths, the symptom element entity node at the corresponding sequence number is recorded as a consistent node, and the process continues for the next... The sequence numbers are compared; when the entity nodes of the syndrome elements corresponding to different reachable syndrome evolution paths are inconsistent at the sequence number, the consistency comparison is terminated. All consistent nodes recorded before the termination position constitute a path consistency segment, and the entity node of the syndrome element corresponding to the termination position and the nodes thereafter are determined as path divergence positions. For example, if there are three reachable syndrome evolution paths, namely, Wei-level heat evil - Qi-level heat excess - Ying-level heat blazing - Blood-level blood stasis and heat, Wei-level heat evil - Qi-level heat excess - Ying-level heat blazing - Blood-level depletion, and Wei-level heat evil - Qi-level heat excess - Qi-level bowel excess; at this time, the entity nodes of the first three syndrome elements are the same, and the fourth node is different, so the first three nodes are determined as a path consistency segment.

[0035] The method for marking path consistency segments is as follows: record the sequence of syndrome element entity nodes belonging to the path consistency segment, and add a marking symbol after the path consistency segment sequence to distinguish the boundary between the path consistency segment and the path divergence position; for example, mark the syndrome element entity node Wei Fen Heat Evil - Qi Fen Heat Excess - Ying Fen Heat Intensity as a path consistency segment, and add a path consistency end symbol after Ying Fen Heat Intensity.

[0036] Path divergence identification is performed on the syndrome element nodes at the termination position of the path consistency segment. The syndrome element nodes at the termination position of the path consistency segment with two or more different evolution directions are identified as key decision nodes in the clinical process of febrile diseases. Specifically, path divergence identification involves: obtaining the syndrome element node at the termination position of the path consistency segment and using this syndrome element node as the starting point for divergence identification analysis; traversing all possible next-step syndrome element entity nodes directly connected to the syndrome element node in the dynamic transmission relationship set of febrile disease syndromes, centered on the syndrome element node; using a breadth-first search algorithm, i.e., taking the syndrome element node at the termination position of the path consistency segment as the root node, sequentially searching all directly connected next-level child nodes and recording the evolution direction corresponding to all child nodes; then determining whether a path divergence exists based on the number of child nodes; for example, if the syndrome element node at the termination position of the path consistency segment is Ying-Fen Re-Chi (nutritive heat), and the next-step nodes directly connected to the syndrome element node are Xue-Fen Yu-Re (blood stasis and heat) and Xue-Fen Hua-Shang (blood depletion), then the Ying-Fen Re-Chi node satisfies the judgment condition for path divergence; if the termination node has only a single successor node, then the path divergence identification requirement is not met.

[0037] The criteria for path divergence identification are defined as follows: during the path divergence identification process, if the number of subsequent syndrome element entity nodes connected to the syndrome element node at the termination position of the path consistency segment is greater than or equal to two, then the syndrome element node is determined to be a location with path divergence. Nodes with path divergence characteristics are identified as key decision nodes in the clinical process of febrile diseases. Key decision nodes can evolve into multiple different syndrome element entity nodes during the syndrome evolution process, representing locations where branch selection is required in clinical decision-making. For example, the Ying-Fen Re-Chi node satisfies the condition of having two or more subsequent syndrome element entity nodes, thus determining Ying-Fen Re-Chi as a key decision node in the clinical process of febrile diseases. Using the above method, path divergence is identified for each syndrome element node at the termination position of the path consistency segment, and the number and name of the subsequent syndrome element entity nodes for each node are recorded.

[0038] To improve the accuracy of determining the location of key decision nodes, the differences in the evolutionary direction between the subsequent syndrome element entity nodes connected to each key decision node location are analyzed. Specifically, each subsequent syndrome element entity node connected to the key decision node location is marked according to the directional relationship data recorded in the dynamic transmission relationship set of febrile disease syndromes, and the subsequent syndrome element entity nodes are classified into forward transmission nodes, reverse transmission nodes, combined disease nodes, or concurrent disease nodes. For example, if the nodes Blood Stasis and Blood Consumption connected to the key decision node location of Yingfen Rechirping and Blood Stasis are both forward transmission nodes according to the directional relationship data analysis, then the directional type of each subsequent node is recorded as forward transmission, and it is recorded as a different branch of the same evolutionary direction; if the subsequent nodes are forward transmission nodes and reverse transmission nodes respectively, they are recorded as branches with different evolutionary directions, and the directional difference information is marked.

[0039] Construct a key decision node location data table, including the name of the syndrome element entity node at the key decision node location, the name of the subsequent syndrome element entity node connected to the syndrome element entity node, the evolution direction relationship type corresponding to each subsequent node, the sequence of syndrome element entity nodes in the path consistency segment, and the analysis results that meet the divergence conditions when identifying path divergences; for example, the key decision node location data table records that the key decision node location is Yingfen Rechirping, the corresponding subsequent nodes are Xuefen Yu Re (forward transmission) and Xuefen Hua Shang (forward transmission), the path consistency segment is Weifen Rexie - Qifen Resheng - Yingfen Rechirping, and the divergence identification condition is met when the number of subsequent nodes is 2.

[0040] S5: For key decision-making nodes, combined with the transmission conditions of syndrome evolution paths, match and deduce the treatment elements and prescription elements to generate corresponding clinical decision-making teaching paths, including: Match the syndrome element nodes corresponding to the key decision nodes and the transmission conditions recorded in multiple reachable syndrome evolution paths with syndrome elements, treatment elements, and prescription elements. Specifically, based on the key decision node location data table, the names of the syndrome element entity nodes corresponding to the key decision node locations are extracted. These extracted syndrome element entity nodes are then compared and matched with the syndrome element nodes stored in the febrile disease knowledge graph to determine the accurate entity identifier of each key decision node location's corresponding syndrome element node in the febrile disease knowledge graph. Combining the transmission conditions recorded in multiple reachable syndrome evolution paths—that is, the transmission constraint rules recorded in the dynamic transmission relationship set of febrile disease syndromes, including syndrome duration, clinical symptoms, and signs—condition matching is performed on the syndrome element entity nodes at each key decision node location. For example, the transmission conditions corresponding to the syndrome element entity node "Yingfen Rechi" include persistent fever for more than three days, a red tongue, and a thready and rapid pulse. The relationship between the transmission conditions and the syndrome element nodes is matched. The matching results are recorded uniformly to obtain the matching relationship between the key decision node location and the corresponding syndrome element entity node and transmission conditions.

[0041] Based on the syndrome element nodes and their corresponding transmission conditions, and according to the relationship between syndrome elements and treatment elements, the treatment elements that match the syndrome element nodes are determined. Specifically, based on the predefined associations between syndrome element nodes and treatment element nodes in the knowledge graph of febrile diseases, each syndrome element entity node corresponding to the key decision node position is traversed. According to the matching transmission conditions, the most suitable treatment element node is determined by conditional probability calculation. Utilizing the association strength attribute between syndrome elements and treatment elements, i.e., the association strength obtained by entity co-occurrence frequency statistics and conditional probability calculation, the treatment element node with the strongest association strength is selected as the best matching treatment element node for the syndrome element node under the transmission conditions. For example, if the syndrome element entity node "Yingfen Rechi" is under the transmission conditions of persistent fever for more than three days, red tongue, and thready and rapid pulse, the association strength with the treatment element entity node "Qingying Jiedu" is determined to be the strongest, and "Qingying Jiedu" is then identified as the corresponding treatment element node. Through the above methods, treatment element node matching is performed on the syndrome element entity nodes at all key decision node positions to form a set of matching relationships between syndrome elements and treatment elements.

[0042] Based on the relationship between treatment principles and prescription elements, determine the prescription elements that match the treatment principles. Specifically, the association between predefined treatment element nodes and prescription element nodes in the knowledge graph of febrile diseases is used, with the association strength attribute as the matching basis. The determined treatment element nodes are obtained. All prescription element nodes connected to each treatment element node in the knowledge graph of febrile diseases are traversed, and the association strength corresponding to each prescription element node is calculated and compared. The association strength is determined based on entity co-occurrence frequency statistics and conditional probability calculation. For example, among the prescription element nodes corresponding to the treatment element node Qingying Jiedu, if Qingying Tang has the highest association strength, with an entity co-occurrence frequency of 30 times and a conditional probability of 0.85, then Qingying Tang is determined as the best matching prescription element node for the treatment element node. Through comparison and calculation, the best prescription element node corresponding to each treatment element node is finally determined, generating a set of matching relationships between treatment elements and prescription elements.

[0043] Based on the matching relationship between syndrome element nodes, determined treatment elements and prescription elements, a clinical decision-making teaching path corresponding to the key decision node position is generated. Specifically, based on the names of the syndrome element entity nodes corresponding to the key decision node locations recorded in the key decision node location data table, and combining the matching relationship sets between syndrome elements and treatment elements, as well as the matching relationship sets between treatment elements and prescription elements, a hierarchical matching path is constructed from the syndrome element entity node to the treatment element node and then to the prescription element node. The matching paths of multiple key decision node locations are integrated and sorted according to the chronological order of the syndrome element entity nodes in clinical decision-making to form a clinical decision-making teaching path sequence. For example, if the determined key decision node location is "Yingfen Rechi" (nutritive heat), the treatment element node matched by the syndrome element node "Yingfen Rechi" is "Qingying Jiedu" (clearing nutritive detoxification), and the prescription element node matched by the treatment element node "Qingying Jiedu" is "Qingying Tang" (clearing nutritive decoction), then the clinical decision-making teaching path sequence is explicitly represented as "Yingfen Rechi" - "Qingying Jiedu" - "Qingying Tang". And so on, recording the clinical decision-making teaching path sequence data corresponding to all key decision node locations constitutes the clinical decision-making teaching path data table.

[0044] Record the names of the syndrome element entities corresponding to the key decision-making node locations, the clinicopathological characteristics of the transmission conditions, the names of the matched treatment element entities, the names of the matched prescription element entities, and the strength of the matching association between the nodes. For example, taking the key decision-making node location of Yingfen Rechi (nutritive heat) as an example, record: the name of the syndrome element entity node is Yingfen Rechi, the transmission conditions are continuous fever for more than three days, red tongue, and thready and rapid pulse, the treatment element entity node is Qingying Jiedu (clearing nutritive heat and detoxifying), the prescription element entity node is Qingying Tang (clearing nutritive heat decoction), the association strength between the syndrome element and the treatment element node is a co-occurrence frequency of 30 times and a conditional probability of 0.85, and the association strength between the treatment element and the prescription element node is a co-occurrence frequency of 25 times and a conditional probability of 0.8.

[0045] Based on the actual needs of clinical teaching applications, the clinical decision-making teaching path data table was screened and adjusted. Specifically, the path sequences were sorted based on the association strength attribute, and clinical decision-making teaching paths with association strength attributes exceeding a preset threshold were selected as priority display paths. The preset threshold was determined based on statistical analysis of practical experience in teaching febrile diseases, for example, paths with entity co-occurrence frequency ≥ 20 times and conditional probability ≥ 0.7 were selected as priority display paths. At the same time, according to the needs of clinical teaching applications and the richness of teaching cases, several suboptimal clinical decision-making teaching paths were appropriately selected and supplemented to ensure the diversity of path data and coverage of clinical scenarios. Through screening and adjustment, the final clinical decision-making teaching path for teaching febrile diseases was formed.

[0046] S6: Organize the clinical decision-making teaching pathway according to the sequence of syndrome evolution, and output deducible clinical decision-making process data for teaching and training, including: The clinical decision-making teaching path is sorted according to the temporal order of the syndrome element nodes in multiple accessible syndrome evolution paths. Specifically, matching relationship data for all syndrome element nodes, treatment element nodes, and prescription element nodes are extracted from the clinical decision-making teaching path data table. A path-by-path analysis is performed on all syndrome element entity nodes across multiple reachable syndrome evolution paths to obtain the position sequence number of each syndrome element entity node within its respective reachable syndrome evolution path. The method for determining the position sequence number is as follows: during the deduction of reachable syndrome evolution paths, based on the temporal relationship data of syndrome elements, the order of appearance of each syndrome element entity node within the reachable syndrome evolution path is recorded, and a sequence number is assigned to each syndrome element entity node based on this order. For example, the syndrome element entity node "Wei Fen Heat Evil" is located at the beginning of the reachable syndrome evolution path and is assigned the sequence number 1. The "Qi-type Heat Excess" is located in the next time sequence position and assigned the sequence number 2, and so on. The position sequence data of each syndrome element entity node in multiple reachable syndrome evolution paths are comprehensively analyzed to calculate the average occurrence sequence number of each syndrome element entity node. For example, if the sequence number of the syndrome element entity node "Qi-type Heat Excess" is 1 in reachable syndrome evolution path 1 and 3 in reachable syndrome evolution path 2, then the average sequence number of "Qi-type Heat Excess" is 2. Based on the average sequence number of the syndrome element entity nodes, all clinical decision teaching paths are uniformly sorted, that is, syndrome element nodes with smaller average sequence numbers are arranged first, and syndrome element nodes with larger average sequence numbers are arranged last, thereby determining the uniform time sequence order of all syndrome element entity nodes.

[0047] The syndrome element nodes, treatment element, and prescription element in the sorted clinical decision-making teaching path are organized into a continuous decision sequence according to the evolution order of the syndrome element nodes. Specifically, based on the determined unified temporal order, the sorted syndrome element nodes are extracted sequentially. Then, based on the matching relationship set between syndrome elements and treatment elements, and the matching relationship set between treatment elements and prescription elements, the corresponding treatment element node and prescription element node for each syndrome element node are determined. For example, if the average sequence number of the syndrome element entity node "Wei-level Heat Evil" is 1, and the corresponding treatment element node is determined to be "Acrid-Cool Exterior Relief" based on the matching relationship set, and the corresponding prescription element node is "Yin-Qiao Powder," then the first-ranked continuous decision sequence is determined to be "Wei-level Heat Evil - Acrid-Cool Exterior Relief - Yin-Qiao Powder." Then, based on the ranking order, the second-ranked syndrome element entity node "Qi-level Heat Excess" is extracted, and so on, forming a continuous decision sequence containing all sorted syndrome element nodes and their corresponding treatment element nodes and prescription element nodes.

[0048] The corresponding transmission conditions are associated with the corresponding syndrome element nodes in the decision sequence to generate clinical decision process data. Specifically, the process iterates through each syndrome element node in the continuous decision sequence, extracts the transmission conditions corresponding to each syndrome element entity node based on the transmission condition matching relationship, that is, the transmission constraint rules recorded in the dynamic transmission relationship set of febrile disease syndromes; the transmission conditions include the duration of the syndrome, clinical symptoms, signs and other clinical and pathological feature changes; for example, for the syndrome element entity node Wei-level heat evil, the corresponding transmission conditions are fever, chills, thin white tongue coating, and floating and rapid pulse. Combined with the determined treatment element of pungent and cool exterior-releasing and the prescription element Yinqiao powder, all the above information is associated and recorded as a set of clinical decision process data; for all syndrome element nodes in the decision sequence, the association and recording of transmission conditions are completed, generating clinical decision process data containing syndrome element nodes, treatment elements, prescription elements and transmission conditions.

[0049] Clinical decision-making process data is stored in a teaching case database as deductive clinical decision-making process data for teaching and training. Specifically, a teaching case database is established, including the names of syndrome element nodes, treatment element nodes, prescription element nodes, transmission conditions, sequence number sorting, and path identifiers. For example, the syndrome element node records heat evil in the Wei level, the treatment element node records pungent and cool exterior-releasing herbs, the prescription element node records Yinqiao Powder, the transmission conditions record fever, chills, thin white tongue coating, and floating and rapid pulse, the sequence number sorting field records 1, and the path identifier field records path 1. All node data in the decision sequence is recorded in this way. An indexing method is set up, using the syndrome element node and sequence number sorting field as a joint index to improve the efficiency of fast retrieval and real-time deduction in the teaching and training process. A data update and maintenance mechanism for the teaching case database is set up, including regularly supplementing and updating new teaching case data to maintain the accuracy of data in the teaching case database in a timely manner.

[0050] In addition, to facilitate the intuitive visualization and interactive training application of clinical decision-making teaching pathway data, a corresponding clinical decision-making teaching interactive interface can be constructed based on the clinical decision-making process data. The interface design includes visual diagrams of syndrome element nodes, treatment element nodes, prescription element nodes, and transmission conditions. Each node is distinguished by different colors or shapes, and transmission conditions are displayed by connecting lines and dynamic pop-ups. The interactive interface allows users to select syndrome nodes and prompts the corresponding treatment element and prescription element nodes as well as transmission conditions.

[0051] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0052] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0053] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0054] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0055] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0056] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0057] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0058] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0059] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs, characterized in that, Includes the following steps: S1: Obtain clinical case data, literature data, and teaching standard data of febrile diseases, perform semantic normalization processing, and construct a knowledge graph of febrile diseases that includes syndrome elements, treatment elements, and prescription elements; S2: Based on the knowledge graph of febrile diseases, the transmission rule data of Wei Qi Ying Xue and the transmission rule data of San Jiao are introduced to jointly constrain the transmission constraint relationship between syndrome elements and generate a set of dynamic transmission relationship of febrile disease syndromes. S3: Based on the dynamic transmission relationship set of febrile disease syndromes, the initial syndrome state of teaching cases is temporally deduced to generate multiple attainable syndrome evolution paths; S4: Conduct path consistency analysis and path divergence identification on multiple syndrome evolution paths to determine the key decision-making nodes in the clinical process of febrile diseases. S5: For key decision-making nodes, combined with the transmission conditions of syndrome evolution path, match and deduce the treatment elements and prescription elements to generate corresponding clinical decision-making teaching paths. S6: Organize the clinical decision-making teaching path according to the sequence of syndrome evolution, and output data on the inferable clinical decision-making process for teaching and training.

2. The teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs according to claim 1, characterized in that, S1, specifically: Acquire clinical case data, literature entries, and teaching standard data related to febrile diseases; Semantic normalization processing was performed on clinical case data, classical literature data, and teaching standard data of febrile diseases. Establish physical identifiers for syndrome elements, treatment elements, and prescription elements; Based on entity identifiers, the relationships between syndrome elements and treatment elements, between treatment elements and prescription elements, and between syndrome elements and prescription elements are extracted. A knowledge graph of febrile diseases is constructed based on syndrome elements, treatment elements, prescription elements, and their relationships.

3. The teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs according to claim 2, characterized in that, Transmission constraints include the sequential relationship, evolutionary direction relationship, and occurrence probability relationship among symptom elements.

4. The teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs according to claim 3, characterized in that, S2, specifically: Based on the knowledge graph of febrile diseases, the data on the transmission rules of Wei Qi Ying Xue and the data on the transmission rules of San Jiao are analyzed into transmission constraint rules between syndrome elements. Based on the rules of transmission constraint, the order of syndrome elements is constrained and labeled to form the temporal relationship data of syndrome elements; Based on the rules of transmission and transformation constraints, the evolution direction between syndrome elements is constrained and labeled to form directional relationship data of syndrome elements; Based on the transmission constraint rules, the occurrence probability of syndrome elements under different transmission conditions is marked to form the occurrence probability relationship data of syndrome elements; By associating temporal relationship data, directional relationship data, and probability relationship data with syndrome elements in the knowledge graph of febrile diseases, a set of dynamic transmission relationships of febrile disease syndromes is constructed.

5. A teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs according to claim 4, characterized in that, Transmission constraint rules describe the constraints that should be met between the symptom elements before and after transmission, as well as the conditions for changes in clinicopathological features that are allowed during transmission.

6. A teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs according to claim 5, characterized in that, S3, specifically: The initial syndrome elements corresponding to the teaching cases are obtained as the initial syndrome state, and the initial syndrome state is mapped to the corresponding syndrome element node in the set of dynamic transmission relationship of febrile disease syndromes. Based on the temporal relationship data of syndrome elements, the order of occurrence of syndrome elements in the deduction process is limited; Based on the directional relationship data of syndrome elements, the evolution direction of syndrome elements in the deduction process is limited; Based on the probability relationship data of the occurrence of syndrome elements, conditional constraint judgments are made on the evolution path of syndrome elements that satisfy the temporal and directional relationships. Based on the joint constraint results, the initial syndrome state is deduced, generating multiple reachable syndrome evolution paths and corresponding transmission conditions.

7. A teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs according to claim 6, characterized in that, S4, specifically: Path consistency analysis was performed on multiple reachable symptom evolution paths to obtain path segments with consistent evolution order of symptom elements in multiple reachable symptom evolution paths, and these segments were marked as path consistency segments. Path divergence identification is performed on the syndrome element nodes at the termination position of the path consistency segment. The syndrome element nodes at the termination position of the path consistency segment with two or more different evolution directions are identified as key decision nodes in the clinical process of febrile diseases.

8. A teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs according to claim 7, characterized in that, S5, specifically: Match the syndrome element nodes corresponding to the key decision nodes and the transmission conditions recorded in multiple reachable syndrome evolution paths with syndrome elements, treatment elements, and prescription elements. Based on the syndrome element nodes and their corresponding transmission conditions, and according to the relationship between syndrome elements and treatment elements, the treatment elements that match the syndrome element nodes are determined. Based on the relationship between treatment principles and prescription elements, determine the prescription elements that match the treatment principles. Based on the matching relationship between syndrome element nodes, determined treatment elements, and prescription elements, a clinical decision-making teaching path corresponding to the key decision node positions is generated.

9. A teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs according to claim 8, characterized in that, S6, specifically: The clinical decision-making teaching path is sorted according to the temporal order of the syndrome element nodes in multiple accessible syndrome evolution paths. The syndrome element nodes, treatment element, and prescription element in the sorted clinical decision-making teaching path are organized into a continuous decision sequence according to the evolution order of the syndrome element nodes. The corresponding transmission conditions are associated with the corresponding syndrome element nodes in the decision sequence to generate clinical decision process data. Clinical decision-making process data is stored in a teaching case database, serving as deductive clinical decision-making process data for teaching and training.

10. A teaching method for clinical decision-making in febrile diseases based on big data knowledge graphs according to claim 9, characterized in that, Clinical decision-making process data includes syndrome elements, treatment elements, prescription elements, and transmission conditions.

Images