Subject linkage calling method and system triggered based on attribute characteristics of target person

Abstract

The application relates to the technical field of linkage regulation, and discloses a subject linkage calling method and system triggered based on attribute characteristics of target personnel, the method comprising the following steps: carrying out cleaning and analysis on multidimensional attribute characteristic data of target personnel in a multi-subject business terminal to obtain structured data portraits; constructing a collaborative diagnosis and treatment knowledge graph of the target personnel; mapping key attribute characteristics in the structured data portraits to the collaborative diagnosis and treatment knowledge graph for matching analysis to obtain a set of to-be-linked subjects; performing priority identification linkage on the set of to-be-linked subjects to obtain a linkage calling sequence; encoding the linkage calling sequence into a control instruction to respond to service preparation and resource presetting operations of the multi-subject business terminal, and integrating the service preparation and the resource presetting into response state information of the control instruction; and according to the response state information, forming a closed-loop linkage state view of the target personnel; the application can improve the efficiency of subject linkage calling triggered based on attribute characteristics of target personnel.

Description

Technical Field

[0001] This invention relates to the field of linkage control technology, and in particular to a method and system for triggering subject linkage based on the attribute characteristics of target personnel. Background Technology

[0002] In the application of multi-disciplinary collaborative diagnosis and treatment technologies, existing technologies are unable to efficiently clean and accurately analyze the multi-source heterogeneous attribute data of target personnel. They are unable to effectively remove abnormal data, integrate duplicate information, and achieve standardized transformation of unstructured data. As a result, the constructed personnel data profile lacks completeness and accuracy, making it difficult to accurately extract key attribute features and provide reliable data support for subsequent subject linkage matching, thereby affecting the accuracy of linkage calls.

[0003] Existing technologies lack a mechanism for constructing collaborative diagnosis and treatment knowledge graphs based on clinical norms, expert knowledge, and historical cases. Furthermore, cross-institutional resource allocation strategies are inadequate, and the prioritization of multi-disciplinary collaboration lacks scientific quantitative basis, making it impossible to rationally prioritize subjects for collaboration. Simultaneously, the lack of effective integration and closed-loop monitoring of service preparation and resource pre-positioning during the collaboration process makes it difficult to grasp the progress and status differences of collaboration in real time, resulting in low efficiency in subject collaboration, unreasonable resource allocation, and an inability to quickly respond to diagnostic and treatment service needs. Summary of the Invention

[0004] This invention provides a method and system for triggering subject linkage based on the attribute characteristics of target personnel, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a subject linkage invocation method triggered by the attribute characteristics of target personnel, comprising:

[0006] S1. Clean and analyze the multi-dimensional attribute feature data of the target personnel in the multi-subject business terminal to obtain a structured data profile of the target personnel;

[0007] S2. Based on the clinical standard information, expert prior knowledge, and historical case literature in the multi-subject business terminal, and combined with the preset rule engine, construct the collaborative diagnosis and treatment knowledge graph of the target personnel;

[0008] S3. Map the key attribute features in the structured data profile to the collaborative diagnosis and treatment knowledge graph for matching analysis to obtain the set of subjects to be linked for the target person;

[0009] S4. Based on the cross-organizational resource scheduling strategy of the multi-subject business terminal, the set of subjects to be linked is linked by priority identification to obtain the linkage call sequence of the multi-subject business terminal.

[0010] S5. Encode the linked call sequence into a control command to respond to the service preparation and resource preset operations of the multi-subject business terminal, and integrate the service preparation and resource preset operations into the response status information of the control command;

[0011] S6. Based on the response status information, form a closed-loop linkage status view of the target personnel.

[0012] In a preferred embodiment, the step of cleaning and parsing the multi-dimensional attribute feature data of the target personnel in the multi-subject business terminal to obtain a structured data profile of the target personnel includes:

[0013] In the multi-subject business terminal, multi-source heterogeneous attribute data of the target personnel are collected in parallel. The multi-source heterogeneous attribute data includes at least text-formatted medical records, numerical-formatted vital sign time-series data, and structured form-formatted personal basic information.

[0014] The multi-source heterogeneous attribute data is subjected to data cleaning operations, which include filling in missing data values, removing outliers that exceed a preset reasonable range, and merging duplicate data records for the same semantic entity to obtain a standard dataset of the multi-source heterogeneous attribute data.

[0015] The unstructured text data in the standard dataset is subjected to key information entity extraction and standardized encoding operations. The extracted entity information is mapped to a preset standardized medical terminology encoding set to generate a standardized feature vector of the unstructured text data.

[0016] The standardized feature vector is associated and aggregated with the standard dataset, and the aggregated data is labeled to obtain the feature identifier of the multi-source heterogeneous attribute data.

[0017] Based on the feature identifiers, construct a structured data profile of the target personnel.

[0018] In a preferred embodiment, the step of constructing a collaborative diagnosis and treatment knowledge graph for the target personnel based on clinical normative information, expert prior knowledge, and historical case literature in the multi-subject business terminal, combined with a preset rule engine, includes:

[0019] The clinical standard information documents, expert prior knowledge base, and historical case literature set in the multi-subject business terminal are used as diverse knowledge sources for constructing the knowledge graph.

[0020] Natural language processing is performed on the historical case literature set, and the case entities, diagnosis and treatment operation entities, and the case entities and diagnosis and treatment operation entities recorded in the processed historical case literature set are collaboratively analyzed and integrated to obtain the historical case triplet set of the historical case literature set.

[0021] The relevant infectious disease prevention and control guidelines, mental disorder diagnosis and treatment guidelines, in-hospital emergency procedures, and medical consortium referral agreements for the target personnel are compiled into the rule engine of the multi-subject business terminal.

[0022] Based on the rule engine, the mandatory treatment path defined in the clinical normative information document, the conditional reasoning branch in the expert prior knowledge base, and the historical case triple set are logically associated and conflict resolved to obtain the treatment knowledge triple set of the target personnel.

[0023] Using the disease type and treatment stage of the target person as the core nodes, the treatment knowledge triples are organized and stored in a graph structure to construct a collaborative treatment knowledge graph for the target person.

[0024] In a preferred embodiment, the step of performing natural language processing on the historical case literature set, and then performing collaborative analysis and integration on the case entities, treatment operation entities, and the case entities and treatment operation entities recorded in the processed historical case literature set to obtain the historical case triplet set of the historical case literature set, including:

[0025] Semantic parsing is performed on the unstructured text in the historical case literature collection to obtain the text sequence of the unstructured text;

[0026] The text sequence is analyzed and read to obtain the case entities and treatment operation entities of the historical case literature set;

[0027] Dependency parsing and semantic role labeling are performed on the case entity and the diagnosis and treatment operation entity to obtain the diagnosis and treatment relationship between the case entity and the diagnosis and treatment operation entity;

[0028] By binding the case entity, the treatment operation entity, and the treatment relationship, a historical case triplet set of the historical case literature set is obtained.

[0029] In a preferred embodiment, the step of mapping key attribute features in the structured data profile to the collaborative diagnosis and treatment knowledge graph for matching analysis to obtain the set of subjects to be linked for the target person includes:

[0030] Extract key attribute features from the structured data profile that are relevant to the current diagnosis and treatment decision. The key attribute features include the main diagnosis, core symptoms, and key examination and testing indicators.

[0031] Using the key attribute features as query entities, entity matching and positioning are performed in the collaborative diagnosis and treatment knowledge graph to obtain the medical entity nodes of the target person;

[0032] Based on the collaborative diagnosis and treatment knowledge graph, the subject entity nodes of the structured data profile are obtained by multi-hop graph traversal and identification starting from the medical entity nodes;

[0033] Based on the preset linkage triggering rules, the subject entity nodes are filtered and aggregated to obtain the type weight and path depth of the association relationship edges of the collaborative diagnosis and treatment knowledge graph;

[0034] The type weights and path depths are prioritized to obtain the set of subjects to be linked for the target personnel.

[0035] In a preferred embodiment, the step of prioritizing and linking the set of subjects to be linked based on the cross-institutional resource scheduling strategy of the multi-subject business terminals to obtain the linkage call sequence of the multi-subject business terminals includes:

[0036] Obtain resource status data from the real-time resource monitoring interface in the multi-subject business terminal;

[0037] Based on the rules defined in the cross-organizational resource scheduling strategy of the multi-subject business terminal, the resource status data of different subjects in the multi-subject business terminal are compared and analyzed with the standard service capability baseline to obtain the dynamic scheduling parameters of the multi-subject business terminal.

[0038] Based on the dynamic scheduling parameters, the urgency of the associated treatment paths between subjects in the set of subjects to be linked is coupled and evaluated to obtain a comprehensive linkage priority score for the set of subjects to be linked.

[0039] Based on the comprehensive linkage priority score, the linkage subjects in the set of subjects to be linked are associated, sorted, and regularized to obtain the linkage call sequence of the subjects to be linked.

[0040] In a preferred embodiment, the formula for calculating the comprehensive linkage priority score is as follows:

[0041] ;

[0042] In the formula, For the first The comprehensive linkage priority score of each subject to be linked This refers to the index identifier of a specific subject in the set of subjects to be linked. For the first The quantification value of the urgency level of the treatment pathways associated with each subject to be linked. This is a preset urgency amplification factor. It is a natural exponential function. The dynamic resource load rate of the multi-subject business terminal corresponding to the i-th subject to be linked. This is the preset resource load sensitivity coefficient.

[0043] In a preferred embodiment, encoding the linked call sequence into control commands to respond to the service preparation and resource pre-setting operations of the multi-subject business terminal, and integrating the service preparation and resource pre-setting operations into response status information of the control commands, includes:

[0044] The core elements of the linked call sequence are parsed to obtain the target business terminal network address, business interface protocol type and execution time window suggestion of the linked call sequence, and a basic scheduling parameter set is generated.

[0045] Based on the business interface protocol type, determine the standard instruction template for the linked call sequence;

[0046] The scheduling information from the basic scheduling parameter set is filled into the standard instruction template to obtain the initial format control instruction of the basic scheduling parameter set;

[0047] Based on the target service terminal network address, the initial format control instruction is encapsulated to obtain the final control instruction of the multi-subject service terminal;

[0048] Receive execution response information from the final control instruction in the target business terminal, and update the execution status of the subject to be linked in the linkage call sequence based on the execution response information;

[0049] The service preparation and resource pre-configuration operations in the execution state are integrated into the response status information of the control command.

[0050] In a preferred embodiment, the step of forming a closed-loop linkage status view of the target person based on the response status information includes:

[0051] The execution status identifier and resource preparation progress data of the target terminal in the multi-subject business terminal are used as response status information;

[0052] The response status information is formatted to obtain a standard status dataset for the control command.

[0053] The standard state dataset is compared and analyzed with the expected linkage state in the linkage call sequence to obtain the linkage state difference set of the standard state dataset;

[0054] Based on the set of linkage state differences and the set of standard state datasets, a visualization data structure is constructed according to preset view construction rules to obtain the initial linkage state view of the target personnel.

[0055] The initial linkage status view is associated and integrated with the structured data profile and the collaborative diagnosis and treatment knowledge graph to obtain the closed-loop linkage status view of the target person.

[0056] To address the aforementioned problems, this invention also provides a subject-linked recall system triggered by the attribute characteristics of target personnel, the system comprising:

[0057] The data profiling module is used to clean and parse the multi-dimensional attribute feature data of target personnel in multi-subject business terminals to obtain a structured data profile of the target personnel.

[0058] The knowledge graph construction module is used to construct a collaborative diagnosis and treatment knowledge graph for the target personnel based on clinical norms information, expert prior knowledge, and historical case literature in the multi-subject business terminal, combined with a preset rule engine.

[0059] The intelligent matching and analysis module is used to map the key attribute features in the structured data profile to the collaborative diagnosis and treatment knowledge graph for matching and analysis, so as to obtain the set of subjects to be linked for the target person;

[0060] The dynamic scheduling decision module is used to prioritize and link the set of subjects to be linked based on the cross-organizational resource scheduling strategy of the multi-subject business terminals, so as to obtain the linkage call sequence of the multi-subject business terminals.

[0061] The instruction generation and distribution module is used to encode the linkage call sequence into control instructions to respond to the service preparation and resource pre-setting operations of the multi-subject business terminal, and to integrate the service preparation and resource pre-setting operations into the response status information of the control instructions;

[0062] The status monitoring and view generation module is used to form a closed-loop linkage status view of the target personnel based on the response status information.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] 1. This invention generates a complete and standardized structured data profile by cleaning and parsing multi-source heterogeneous attribute data, extracting key information entities, and standardizing and encoding them; at the same time, it integrates clinical norms, expert prior knowledge, historical cases, and relevant rules to construct a collaborative diagnosis and treatment knowledge graph, achieving accurate matching of key attribute features and subject entities, effectively improving the accuracy of subject linkage and calling, and providing reliable support for diagnosis and treatment decisions.

[0065] 2. Based on cross-institutional resource scheduling strategies and a quantitative comprehensive linkage priority scoring mechanism, this invention can scientifically sort linkage subjects and generate an optimized linkage call sequence. By encoding the sequence into control instruction response service preparation and resource pre-setting, and constructing a closed-loop linkage status view, the progress of linkage and resource status can be monitored in real time, significantly improving the efficiency of subject linkage call, ensuring reasonable and orderly resource allocation, and strengthening the synergy and controllability of diagnosis and treatment services. Attached Figure Description

[0066] Figure 1 This is a flowchart illustrating a method for triggering subject linkage based on the attribute characteristics of a target person, as provided in an embodiment of the present invention.

[0067] Figure 2 A functional module diagram of a subject linkage calling system triggered by the attribute characteristics of a target person, provided in an embodiment of the present invention;

[0068] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0069] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0070] This application provides a method for triggering subject-based linkage invocation based on the attribute characteristics of target personnel. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for triggering subject-based linkage invocation based on the attribute characteristics of target personnel can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0071] Reference Figure 1 The diagram shown is a flowchart illustrating a subject-linked invocation method triggered by the attribute characteristics of a target person, according to an embodiment of the present invention. In this embodiment, the subject-linked invocation method triggered by the attribute characteristics of the target person includes:

[0072] S1. Clean and analyze the multi-dimensional attribute feature data of the target personnel in the multi-subject business terminal to obtain a structured data profile of the target personnel;

[0073] In this embodiment of the invention, the step of cleaning and parsing the multi-dimensional attribute feature data of the target personnel in the multi-subject business terminal to obtain a structured data profile of the target personnel includes:

[0074] In the multi-subject business terminal, multi-source heterogeneous attribute data of the target personnel are collected in parallel. The multi-source heterogeneous attribute data includes at least text-formatted medical records, numerical-formatted vital sign time-series data, and structured form-formatted personal basic information.

[0075] The multi-source heterogeneous attribute data is subjected to data cleaning operations, which include filling in missing data values, removing outliers that exceed a preset reasonable range, and merging duplicate data records for the same semantic entity to obtain a standard dataset of the multi-source heterogeneous attribute data.

[0076] The unstructured text data in the standard dataset is subjected to key information entity extraction and standardized encoding operations. The extracted entity information is mapped to a preset standardized medical terminology encoding set to generate a standardized feature vector of the unstructured text data.

[0077] The standardized feature vector is associated and aggregated with the standard dataset, and the aggregated data is labeled to obtain the feature identifier of the multi-source heterogeneous attribute data.

[0078] Based on the feature identifiers, construct a structured data profile of the target personnel.

[0079] For multi-subject business terminals, the HIS system, vital signs monitoring module, and EMR system simultaneously initiate data collection processes. The complete text-format medical records containing chief complaint, present medical history, examination orders, and treatment measures are extracted from the HIS system. The full numerical format time-series data of vital signs, including body temperature, heart rate, systolic blood pressure, diastolic blood pressure, and blood oxygen saturation, generated continuously according to timestamps, are extracted from the vital signs monitoring module. The complete personal basic information, including name, gender, age, place of origin, contact information, and past medical history, is extracted from the structured forms of the EMR system. During the collection process, data is transmitted synchronously from each terminal, and after aggregation, multi-source heterogeneous attribute data of the target personnel covering the three types of data is formed.

[0080] Layered cleaning was performed on each field of the multi-source heterogeneous attribute data. Missing values ​​in numerical vital sign time series data were filled with the average of two consecutive valid monitoring values ​​for the target person within the same time interval. Missing values ​​in text-format medical records were filled with standard medical records from the same department and similar medical scenarios for the target person. Missing values ​​in structured form-format personal basic information were filled with supplementary information filed with the business terminal. Numerical vital sign time series data were compared one by one with preset reasonable ranges, including body temperature exceeding 35.0℃-42.0℃ and heart rate exceeding 60 beats per minute. Values ​​with blood pressure ranging from -180 beats per minute, systolic blood pressure exceeding 80-200 mmHg, diastolic blood pressure exceeding 60-120 mmHg, and blood oxygen saturation exceeding 90%-100% are directly identified as outliers and removed. Text and form data are not subject to outlier removal. Records pointing to the same semantic entity are filtered using the unique identifier of the target individual as the search key. Records with completely identical content are merged and retained as single records. Records with partial content differences are merged, and duplicate records are retained and deleted after merging the complete information, ultimately forming a standard dataset of multi-source heterogeneous attribute data. For unstructured text medical records in the standard dataset, a sentence-by-sentence entity boundary recognition method is used to extract key information entities such as disease name, symptoms, examination items, medications, treatment methods, and body parts. All extracted key information entities are precisely matched with a pre-set standardized medical terminology code set, assigning a unique standard code to each key information entity. The codes are arranged according to the order of appearance of the key information entities in the medical record, forming an ordered code sequence as the standardized feature vector of the unstructured text data.

[0081] Using the unique identifier of the target person as the primary key, the standardized feature vector is fully integrated and aggregated with the numerical vital sign time series data and the structured form personal basic information in the standard dataset to form aggregated data. According to the preset labeling rules, unique feature labels are matched for each dimension of the aggregated data, and the label injection is completed at the corresponding data dimension position, finally generating feature identifiers for multi-source heterogeneous attribute data.

[0082] Using feature identifiers as the core data source, the construction work is carried out according to the preset structured data profile framework. Personal basic information data is extracted from the feature identifiers and filled into the corresponding dimensions. Numerical vital sign time series data are sorted by timestamp and filled into the vital sign time series dimension. Standardized feature vectors and related key diagnostic and treatment information entities are extracted and classified into the diagnostic and treatment information dimension. All feature labels are classified and sorted according to the label system and filled into the feature label dimension. After all four dimensions of data are filled, a structured data profile of the target personnel with clear dimensions, interconnected data, and complete content is formed.

[0083] The beneficial effects include improving the emergency treatment of infectious diseases and psychiatric medical services for patients with mental disorders, optimizing the diagnosis and treatment process and management efficiency, ensuring the safety of medical staff and patients, promoting the integration and sharing of medical resources and the implementation of hierarchical diagnosis and treatment, improving the patient's medical experience, reducing medical costs, strengthening business supervision and data security, and promoting the standardization and convenience of mental health services.

[0084] S2. Based on the clinical standard information, expert prior knowledge, and historical case literature in the multi-subject business terminal, and combined with the preset rule engine, construct the collaborative diagnosis and treatment knowledge graph of the target personnel;

[0085] In this embodiment of the invention, the step of constructing a collaborative diagnosis and treatment knowledge graph for the target personnel based on clinical normative information, expert prior knowledge, and historical case literature in the multi-subject business terminal, combined with a preset rule engine, includes:

[0086] The clinical standard information documents, expert prior knowledge base, and historical case literature set in the multi-subject business terminal are used as diverse knowledge sources for constructing the knowledge graph.

[0087] Natural language processing is performed on the historical case literature set, and the case entities, diagnosis and treatment operation entities, and the case entities and diagnosis and treatment operation entities recorded in the processed historical case literature set are collaboratively analyzed and integrated to obtain the historical case triplet set of the historical case literature set.

[0088] The relevant infectious disease prevention and control guidelines, mental disorder diagnosis and treatment guidelines, in-hospital emergency procedures, and medical consortium referral agreements for the target personnel are compiled into the rule engine of the multi-subject business terminal.

[0089] Based on the rule engine, the mandatory treatment path defined in the clinical normative information document, the conditional reasoning branch in the expert prior knowledge base, and the historical case triple set are logically associated and conflict resolved to obtain the treatment knowledge triple set of the target personnel.

[0090] Using the disease type and treatment stage of the target person as the core nodes, the treatment knowledge triples are organized and stored in a graph structure to construct a collaborative treatment knowledge graph for the target person.

[0091] The process involves performing natural language processing on the historical case literature set, and then performing collaborative analysis and integration on the case entities, treatment operation entities, and the case entities and treatment operation entities recorded in the processed historical case literature set to obtain a set of historical case triplets, including:

[0092] Semantic parsing is performed on the unstructured text in the historical case literature collection to obtain the text sequence of the unstructured text;

[0093] The text sequence is analyzed and read to obtain the case entities and treatment operation entities of the historical case literature set;

[0094] Dependency parsing and semantic role labeling are performed on the case entity and the diagnosis and treatment operation entity to obtain the diagnosis and treatment relationship between the case entity and the diagnosis and treatment operation entity;

[0095] By binding the case entity, the treatment operation entity, and the treatment relationship, a historical case triplet set of the historical case literature set is obtained.

[0096] We extract officially published clinical guidelines and information documents such as the guidelines for the prevention and control of infectious diseases and the guidelines for the diagnosis and treatment of mental disorders from the clinical information database of multi-subject business terminals. We collect expert prior knowledge bases formed by psychiatric and infectious disease experts on the diagnostic points, medication principles, and risk prevention and control points for the treatment of comorbidities. We compile medical records, treatment summaries, and academic research literature of patients with mental disorders and infectious diseases in the hospital and medical alliance over the past five years to form a historical case literature collection. We summarize these three types of data in their entirety as a multi-source knowledge base for constructing a collaborative diagnosis and treatment knowledge graph.

[0097] The historical case literature collection is subjected to sentence-by-sentence word segmentation, semantic analysis and entity recognition. First, the basic semantic units in the text are split, then the case entities and diagnosis and treatment operation entities are identified. Then, by comparing the contextual relationship of the entities in the literature, the correspondence between the case entities and the diagnosis and treatment operation entities is clarified. Each relationship is organized into a structured form of "case entity - relationship - diagnosis and treatment operation entity", and finally the historical case triplet set is formed.

[0098] By sorting out the various rules and regulations related to the target personnel in the multi-subject business terminals, the diagnostic criteria, isolation requirements, and treatment procedures in the national and local infectious disease prevention and control guidelines, the medication guidelines, disease assessment standards, and rehabilitation intervention pathways in the mental disorder diagnosis and treatment guidelines, the hospital's infectious disease emergency response procedures, and the referral conditions, referral procedures, and information sharing requirements specified within the medical consortium, etc., are classified and organized according to the structure of "triggering conditions - execution actions - judgment criteria" to form a logically clear and directly callable rule engine.

[0099] The relevant rules in the rule engine are invoked. First, the mandatory treatment paths defined in the clinical guidelines information document are matched with similar treatment scenarios in the historical case triplet set. Then, the conditional reasoning branches in the expert prior knowledge base are incorporated into the matching process. If a rule conflict occurs, the mandatory treatment path in the clinical guidelines information document takes priority. If there is no mandatory requirement, the consensus opinion in the expert prior knowledge base takes precedence. Logical association and conflict resolution are completed in this way, and finally, a treatment knowledge triplet containing the entire process of treatment for the target personnel is formed.

[0100] Using the specific disease type and current treatment stage of the target individuals as core nodes, the case entities and treatment operation entities in the treatment knowledge triples are used as associated nodes. The system is organized according to the structure of "core node - relationship - associated node". For example, "novel coronavirus infection complicated with schizophrenia - treatment period" is used as the core node, and associated triples such as "positive nucleic acid test - basis - antiviral medication" and "worsening of mental symptoms - trigger - adjustment of antipsychotic drugs" are linked. A graph database is used to store these nodes and relationships to ensure that the associations between nodes are clear and traceable, and finally a collaborative treatment knowledge graph of the target individuals is constructed.

[0101] Unstructured text from a historical case literature collection is segmented sentence by sentence, breaking down the text into basic semantic units. Redundant components such as modal particles and auxiliary words without actual diagnostic or treatment significance are removed. The remaining semantic units are then recombined according to the logical order of the text's narration, forming a clear and semantically coherent text sequence. For the obtained text sequence, key information is analyzed segment by segment using a pre-defined standardized medical terminology dictionary. Content describing the patient's health status is identified as case entity, and content describing medical interventions is identified as diagnostic and treatment operation entity. Each entity is also labeled with its corresponding category, completing the extraction of case entity and diagnostic and treatment operation entity.

[0102] Sentence component analysis was used to perform dependency parsing on text fragments containing case entities and treatment operation entities to clarify the grammatical dependencies between entities. Semantic role labeling was then used to determine the functional role of each entity in the treatment scenario. Combined with medical treatment logic, specific association types between case entities and treatment operation entities were extracted, forming clear treatment relationships such as "symptom-trigger-examination," "examination-guidance-medication," and "treatment-improvement-symptom." Following a fixed structure of "case entity-treatment relationship-treatment operation entity," each case entity, its corresponding treatment relationship, and associated treatment operation entity were bound one-to-one. Examples include "persistent fever-trigger-nucleic acid testing," "acute exacerbation of schizophrenia-guidance-adjustment of antipsychotic medication," and "positive nucleic acid test-trigger-isolation treatment." All the bound structured information was compiled and summarized to form a historical case ternary set in the historical case literature collection.

[0103] The beneficial effects include integrating multi-source diagnostic and treatment data such as clinical guidelines, expert experience, and historical cases; accurately separating and extracting case entities and diagnostic and treatment operation entities; clearly defining the diagnostic and treatment relationships between entities; forming standardized structured triples; and constructing a logically coherent, clearly related, and traceable collaborative diagnostic and treatment knowledge graph. This provides accurate evidence for the diagnosis and treatment of patients with mental disorders and infectious diseases, improves the standardization and efficiency of the diagnostic and treatment process, and ensures the scientific nature and safety of diagnostic and treatment decisions.

[0104] S3. Map the key attribute features in the structured data profile to the collaborative diagnosis and treatment knowledge graph for matching analysis to obtain the set of subjects to be linked for the target person;

[0105] In this embodiment of the invention, the step of mapping key attribute features in the structured data profile to the collaborative diagnosis and treatment knowledge graph for matching analysis to obtain the set of subjects to be linked for the target person includes:

[0106] Extract key attribute features from the structured data profile that are relevant to the current diagnosis and treatment decision. The key attribute features include the main diagnosis, core symptoms, and key examination and testing indicators.

[0107] Using the key attribute features as query entities, entity matching and positioning are performed in the collaborative diagnosis and treatment knowledge graph to obtain the medical entity nodes of the target person;

[0108] Based on the collaborative diagnosis and treatment knowledge graph, the subject entity nodes of the structured data profile are obtained by multi-hop graph traversal and identification starting from the medical entity nodes;

[0109] Based on the preset linkage triggering rules, the subject entity nodes are filtered and aggregated to obtain the type weight and path depth of the association relationship edges of the collaborative diagnosis and treatment knowledge graph;

[0110] The type weights and path depths are prioritized to obtain the set of subjects to be linked for the target personnel.

[0111] From the structured data profiles of the target personnel, key attribute features directly related to the current diagnosis and treatment decision are selected. The primary diagnosis must specify the specific disease combination, the core symptoms must be extracted to show the decisive role in the formulation of the treatment plan, and the key examination and test indicators must be selected as quantitative results with diagnostic specificity. During the screening process, redundant information that is not related to the current treatment decision is strictly excluded, and only the core data that affects the linkage between departments and the adjustment of the treatment plan are retained.

[0112] The extracted main diagnoses, core symptoms, and key examination and testing indicators are treated as independent query entities. They are then compared against the standardized entity database in the collaborative diagnosis and treatment knowledge graph using standardized rules such as ICD10 disease coding and SNOMED terminology coding. For example, "novel coronavirus infection" matches the corresponding disease entity node in the knowledge graph, "persistent fever above 38.5℃" matches the corresponding symptom entity node, and "positive nucleic acid test" matches the corresponding examination indicator entity node. This ensures that each query entity can be located in a unique and accurate medical entity node in the knowledge graph, forming a set of medical entity nodes for the target person.

[0113] Starting from the located medical entity nodes, a multi-hop graph traversal is performed according to the preset association relationships in the collaborative diagnosis and treatment knowledge graph. During the traversal, only the association paths directly related to diagnosis and treatment are retained. The maximum traversal depth is set to 3 hops. For example, starting from the "novel coronavirus infection" disease node, a 1-hop traversal yields the "infectious disease department" subject node; starting from the "auditory hallucinations and delusions" symptom node, a 1-hop traversal yields the "psychiatry department" subject node; starting from the "chest CT ground-glass opacity" examination indicator node, a 1-hop traversal yields the "radiology department" subject node. Finally, all traversed subject category nodes are collected to form a set of subject entity nodes corresponding to the structured data profile.

[0114] Pre-defined linkage trigger rules: Subject types that directly participate in disease diagnosis and treatment are weighted at 3 points, subject types that provide auxiliary examination and testing support are weighted at 2 points, and subject types that provide logistical support or technical support are weighted at 1 point; path depth is calculated based on the number of traversal jumps, with 1 jump counting as 1 point, 2 jumps as 2 points, and 3 jumps as 3 points, and the lower the path depth score, the higher the priority.

[0115] The association edges corresponding to subject entity nodes are weighted by type and their path depth is calculated. For example, "Infectious Diseases" corresponds to the relationship of "Disease - Corresponding Treatment Subject," with a type weight of 3 and a path depth of 1; "Psychiatry" corresponds to the relationship of "Symptom - Related Treatment Subject," with a type weight of 3 and a path depth of 1; "Radiology" corresponds to the relationship of "Examination Indicator - Related Auxiliary Subject," with a type weight of 2 and a path depth of 1; and "Laboratory" corresponds to the relationship of "Examination Indicator - Related Auxiliary Subject," with a type weight of 2 and a path depth of 1. Subject entity nodes are sorted according to the priority rule of "descending order of type weight, ascending order of path depth if weights are the same." Subjects with a type weight of 3 are prioritized over subjects with 2 or 1, and subjects with a path depth of 1 are prioritized over subjects with 2 or 3 if weights are the same. After sorting, subject entity nodes with a weight ≥ 1 are selected, and a set of subjects to be linked for the target personnel is formed according to the sorting results.

[0116] The beneficial effects include accurately extracting key attribute features related to diagnosis and treatment decisions from structured data profiles, achieving efficient entity matching and positioning with collaborative diagnosis and treatment knowledge graphs, comprehensively identifying related subject entity nodes, scientifically screening and aggregating and clarifying the type weight and path depth of the relationships, forming a set of subjects to be linked through reasonable priority sorting, providing a clear basis for multi-subject collaborative diagnosis and treatment, improving the accuracy and efficiency of diagnosis and treatment decisions, and ensuring the coordination and standardization of the diagnosis and treatment process for patients with mental disorders and infectious diseases.

[0117] S4. Based on the cross-organizational resource scheduling strategy of the multi-subject business terminal, the set of subjects to be linked is linked by priority identification to obtain the linkage call sequence of the multi-subject business terminal.

[0118] In this embodiment of the invention, the step of prioritizing and linking the set of subjects to be linked based on the cross-institutional resource scheduling strategy of the multi-subject business terminal to obtain the linkage call sequence of the multi-subject business terminal includes:

[0119] Obtain resource status data from the real-time resource monitoring interface in the multi-subject business terminal;

[0120] Based on the rules defined in the cross-organizational resource scheduling strategy of the multi-subject business terminal, the resource status data of different subjects in the multi-subject business terminal are compared and analyzed with the standard service capability baseline to obtain the dynamic scheduling parameters of the multi-subject business terminal.

[0121] Based on the dynamic scheduling parameters, the urgency of the associated treatment paths between subjects in the set of subjects to be linked is coupled and evaluated to obtain a comprehensive linkage priority score for the set of subjects to be linked.

[0122] Based on the comprehensive linkage priority score, the linkage subjects in the set of subjects to be linked are associated, sorted, and regularized to obtain the linkage call sequence of the subjects to be linked.

[0123] The formula for calculating the comprehensive linkage priority score is as follows:

[0124] ;

[0125] In the formula, For the first The comprehensive linkage priority score of each subject to be linked This refers to the index identifier of a specific subject in the set of subjects to be linked. For the first The quantification value of the urgency level of the treatment pathways associated with each subject to be linked. This is a preset urgency amplification factor. It is a natural exponential function. The dynamic resource load rate of the multi-subject business terminal corresponding to the i-th subject to be linked. This is the preset resource load sensitivity coefficient.

[0126] The urgency level quantification values ​​are derived from the diagnostic and treatment guidelines for mental disorders complicated with infectious diseases, the collaborative diagnosis and treatment process of medical consortia, and historical case data. First, the urgency level of the associated diagnosis and treatment pathways is divided into three levels: Level 1 is life-threatening and requires immediate treatment; Level 2 is disease progression and requires prompt treatment; and Level 3 is stable and can be treated routinely. Then, the influence weight of different urgency level subjects on the diagnosis and treatment effect in historical cases is combined with statistical analysis to determine that the quantification value of Level 1 is 8, Level 2 is 5, and Level 3 is 3, forming the urgency level quantification value of each subject to be linked.

[0127] The urgency amplification coefficient is based on the hospital's statistical data on the diagnosis and treatment of mental disorders combined with infectious diseases over the past three years. The analysis of the adverse event incidence rate when different urgency levels of departments fail to coordinate in a timely manner revealed that the adverse event incidence rate for Level 1 urgency departments is 1.5 times that of Level 2. Combining the collaborative diagnosis and treatment experience of psychiatric and infectious disease experts, the coefficient was determined to be 1.2 to highlight the priority of urgency departments and ensure that departments with higher urgency levels receive more significant weight in the scoring.

[0128] The dynamic resource load rate is derived from resource status data collected by the real-time resource monitoring interface of the multi-subject business terminal. This includes indicators such as the on-duty rate of medical staff, equipment idle rate, bed vacancy rate, and utilization rate of laboratory and examination equipment. The calculation method is to take the average of the ratio of the actual value of each indicator to the standard service capacity baseline. For example, if the on-duty rate of medical staff in a certain department is 62.5%, the equipment idle rate is 40%, and the bed vacancy rate is 25%, then the dynamic resource load rate is 0.725, which directly reflects the current resource occupancy of the subject.

[0129] The resource load sensitivity coefficient is based on historical data of cross-institutional resource scheduling in medical consortia. By analyzing the changing trend of the linkage response efficiency of subjects under different resource load rates, it was found that when the resource load rate exceeds 70%, the linkage response time is extended by an average of 2 times. Combined with experts' suggestions on resource scheduling balance, in order to avoid the linkage effect of high-load subjects being affected by insufficient resources, the coefficient was determined to be 0.8, so that the score decay of subjects with higher resource load rates is more obvious.

[0130] The significance of this formula lies in combining the urgency of the treatment of the subjects to be linked with the real-time resource load. It strengthens the priority of urgent subjects through an urgency amplification coefficient and reflects the attenuation effect of resource load on linkage feasibility through a natural exponential function. The final calculated comprehensive linkage priority score ensures that urgent subjects such as life-threatening and rapidly progressing diseases are prioritized for linkage, while avoiding the impact of over-saturated subjects on overall treatment efficiency due to their inability to respond in a timely manner. This achieves the scientific quantification of multi-subject linkage priority, provides an accurate basis for the sorting of subsequent linkage call sequences, and supports the reasonable scheduling and efficient collaboration of multi-subject business terminals across institutions' resources.

[0131] Through the cross-institutional resource monitoring interface of the multi-subject business terminal, real-time resource status data of the Municipal Mental Health Center and 16 district and county mental health centers is captured. This includes the number of medical staff on duty in each subject, the availability of diagnostic and treatment equipment, bed occupancy rate, utilization rate of laboratory and examination equipment, drug reserves, and the smoothness of referral channels. All data is categorized and organized by subject to ensure that each resource status data corresponds to a clear subject and institution without omissions or confusion. Based on the preset rules in the cross-institutional resource scheduling strategy of the multi-subject business terminal, the baseline service capacity of each subject is first defined: the baseline for the number of medical staff on duty is set according to the department size; the baseline for equipment vacancy rate is 30%; the baseline for bed vacancy rate is 20%; the baseline for equipment utilization rate is no more than 70%; the baseline for drug reserves is sufficient to meet 3 days of treatment needs; and the baseline for referral channel response time is no more than 30 minutes.

[0132] The real-time resource status data of each subject is compared with the corresponding baseline one by one. If the number of medical staff on duty is lower than the baseline, the resource sufficiency in the dynamic scheduling parameters is marked as "insufficient". If the equipment idle rate is lower than the baseline, it is marked as "stressed". If the bed vacancy rate is lower than the baseline, it is marked as "saturated". If the equipment utilization rate exceeds the baseline, it is marked as "overloaded". If the drug reserve is lower than the baseline, it is marked as "shortage". If the response time of the referral channel exceeds the baseline, it is marked as "congested". Otherwise, it is marked as "normal". Finally, dynamic scheduling parameters containing resource sufficiency, load level and response priority coefficient of each subject are formed.

[0133] Based on dynamic scheduling parameters, the associated treatment paths between the subjects to be linked are first sorted out, and then the treatment paths are divided into three levels according to their urgency: Level 1 is life-threatening and requires immediate treatment, Level 2 is disease progression and requires prompt treatment, and Level 3 is stable and can be treated routinely.

[0134] The evaluation is coupled with dynamic scheduling parameters: Level 1 urgency is scored as follows: 8 points for Level 1, 5 points for Level 2, and 3 points for Level 3; 2 points are added for "normal" resource sufficiency, no points are added for "insufficient" or "tight", and 2 points are deducted for "saturated", "overloaded", "shortage", and "congested"; 2 points are added for "normal" load level and deducted for "overloaded". Finally, the basic score and the points added or deducted by the resource parameters are added together to obtain the comprehensive linkage priority score for each subject to be linked.

[0135] All subjects to be linked are sorted from high to low according to the comprehensive linkage priority score. Subjects with the same score are arranged in the logical order of related diagnosis and treatment paths. During the sorting process, duplicate or unnecessary subjects are removed to ensure that the sorted subject sequence conforms to the continuity of the diagnosis and treatment process and the efficiency of cross-institutional resource scheduling, and finally forms a clear and orderly linkage call sequence of subjects to be linked.

[0136] The beneficial effects include real-time acquisition of multi-subject business terminal resource status data, generation of dynamic scheduling parameters by comparing with the standard service capability baseline, accurate assessment of the urgency of the associated diagnosis and treatment paths between subjects to be linked and obtaining a comprehensive linkage priority score, forming an orderly linkage call sequence according to the score, realizing reasonable cross-institutional resource scheduling and efficient multi-subject collaboration, improving the accuracy, orderliness and efficiency of diagnosis and treatment services for patients with mental disorders and infectious diseases, and ensuring a standardized and smooth diagnosis and treatment process.

[0137] S5. Encode the linked call sequence into a control command to respond to the service preparation and resource preset operations of the multi-subject business terminal, and integrate the service preparation and resource preset operations into the response status information of the control command;

[0138] In this embodiment of the invention, encoding the linked call sequence into control instructions to respond to the service preparation and resource pre-setting operations of the multi-subject business terminal, and integrating the service preparation and resource pre-setting operations into response status information of the control instructions, includes:

[0139] The core elements of the linked call sequence are parsed to obtain the target business terminal network address, business interface protocol type and execution time window suggestion of the linked call sequence, and a basic scheduling parameter set is generated.

[0140] Based on the business interface protocol type, determine the standard instruction template for the linked call sequence;

[0141] The scheduling information from the basic scheduling parameter set is filled into the standard instruction template to obtain the initial format control instruction of the basic scheduling parameter set;

[0142] Based on the target service terminal network address, the initial format control instruction is encapsulated to obtain the final control instruction of the multi-subject service terminal;

[0143] Receive execution response information from the final control instruction in the target business terminal, and update the execution status of the subject to be linked in the linkage call sequence based on the execution response information;

[0144] The service preparation and resource pre-configuration operations in the execution state are integrated into the response status information of the control command.

[0145] Each subject to be linked in the linkage call sequence is parsed one by one, the target business terminal network address corresponding to each subject is extracted, the business interface protocol type is determined, and the execution time window suggestion is set according to the urgency of diagnosis and treatment. This information is classified and organized by subject to form a basic scheduling parameter set containing terminal address, protocol type, execution time window and subject name.

[0146] Based on the parsed business interface protocol type, a corresponding standard instruction template is matched. If the protocol type is WEBSERVICE, an XML-formatted standard instruction template is selected, which includes fixed fields such as instruction identifier, target address, protocol version, execution time, and operation type. If it is an HL7 protocol, a text-formatted template conforming to the HL7v2.x standard is selected, which includes core elements such as field separators, message type, and triggering events. If it is a DICOM protocol, a dedicated instruction template for medical imaging equipment is selected, which includes fields such as device identifier, opcode, and resource type, ensuring that the template is fully compatible with the protocol type.

[0147] The scheduling information, such as the target service terminal network address, execution time window, and pending operation, from the basic scheduling parameter set is filled into the specified fields of the corresponding standard instruction template one by one. During the filling process, the template field format requirements are strictly followed to ensure that the information is accurately matched and that there are no missing fields or format errors. After the filling is completed, an initial format control instruction is formed, which clearly defines the execution subject, execution content, and time requirements.

[0148] Based on the target service terminal network address in the basic scheduling parameter set, the initial format control command is encapsulated using the TCP / IP protocol. During the encapsulation process, a terminal unique identifier, command check code, and data length field are added. The check code is generated by calculating the character combination of the command content according to fixed rules. After encapsulation, the reachability of the target network address also needs to be verified. If the address is unreachable, the command is marked as abnormal and feedback is provided. If it is reachable, a final control command that can be directly sent to the target terminal is formed.

[0149] The multi-subject business terminal receives the execution response message returned by the target terminal through the communication interface. The response message includes the instruction identifier, execution result, execution time, failure reason, etc. Based on the instruction identifier, the corresponding subject to be linked is matched, and the execution status of the subject is updated to "executed", "execution failed" or "execution in progress".

[0150] The execution status of each subject to be linked is broken down, and the service preparation status and resource pre-configuration status are extracted. These are then integrated according to the structure of subject name, execution status, service preparation details, resource pre-configuration details, and response time to form a unified format of control instruction response status information. This ensures that the information fully reflects the actual preparation status after the execution of each subject instruction.

[0151] The beneficial effects include standardizing the entire process of generating and executing multi-subject linkage control instructions. By accurately extracting key subject information, matching and adapting standard templates, and standardizing the filling and encapsulation of instructions, the system ensures that control instructions are accurate, compliant, and directly executable. It also clearly tracks the execution status of each subject, comprehensively presents service preparation and resource pre-positioning details, and guarantees the accuracy, orderliness, and traceability of cross-institutional linkage of multi-subject business terminals. This provides efficient instruction support for the collaborative diagnosis and treatment of patients with mental disorders and infectious diseases.

[0152] S6. Based on the response status information, form a closed-loop linkage status view of the target personnel.

[0153] In this embodiment of the invention, the step of forming a closed-loop linkage status view of the target person based on the response status information includes:

[0154] The execution status identifier and resource preparation progress data of the target terminal in the multi-subject business terminal are used as response status information;

[0155] The response status information is formatted to obtain a standard status dataset for the control command.

[0156] The standard state dataset is compared and analyzed with the expected linkage state in the linkage call sequence to obtain the linkage state difference set of the standard state dataset;

[0157] Based on the set of linkage state differences and the set of standard state datasets, a visualization data structure is constructed according to preset view construction rules to obtain the initial linkage state view of the target personnel.

[0158] The initial linkage status view is associated and integrated with the structured data profile and the collaborative diagnosis and treatment knowledge graph to obtain the closed-loop linkage status view of the target person.

[0159] The execution status identifier of each subject to be linked is extracted from the target terminal of the multi-subject business terminal. At the same time, the resource preparation progress data of each subject is collected. These data come from the execution response messages fed back by the target terminal and the real-time resource monitoring interface of the multi-subject business terminal. The two types of data are integrated as response status information.

[0160] The response status information is standardized according to the fixed field format of "subject name - execution status - resource type - preparation progress - response time". The execution status is uniformly labeled with three types of enumeration values: "executed", "execution failed" and "execution in progress". The resource preparation progress is presented in the form of percentage. The response time is uniformly converted to the format of "year-month-day hour:minute:second". Duplicate and invalid data are removed to form a standard status dataset of control instructions with a unified structure and standardized content.

[0161] Extract the expected linkage status of each subject to be linked from the linkage call sequence, including the expected execution status and expected resource preparation progress. Compare the actual execution status and resource preparation progress of each subject in the standard status dataset with the corresponding expected values ​​one by one to identify the content that is inconsistent with the actual and the expected, such as the execution status being "execution failed" or "in execution", the resource preparation progress being less than 100%, the remote channel not being opened, etc. Organize these inconsistent information according to the structure of "subject name - expected value - actual value - difference type" to form the linkage status difference set of the standard status dataset.

[0162] The default view construction rules are as follows: Subjects to be linked are displayed in the order of their linked call sequence; "Executed" status is marked in green, "Execution Failed" status in red, and "Executing" status in yellow; a progress bar visually displays the resource preparation progress, with the bar length proportional to the percentage; the differences in the linked status difference set are listed in the corresponding position for each subject; the view includes four core columns: "Subject Name," "Execution Status," "Resource Preparation Progress," and "Difference Description." Based on these rules, the subject information, execution status, and resource progress data from the standard status dataset are integrated with the difference descriptions from the linked status difference set to construct a visual initial linked status view, ensuring that the view clearly presents the linked execution status and resource preparation status of each subject.

[0163] By using the unique identifier of the target personnel, the initial linkage status view is linked and integrated with the structured data profile and the collaborative diagnosis and treatment knowledge graph: the main diagnosis, core symptoms, key examination and test indicators and other core information from the structured data profile are added to the top of the initial view to facilitate intuitive understanding of the patient's diagnosis and treatment background; the nodes of each subject in the initial view are connected with the corresponding subject entity nodes in the collaborative diagnosis and treatment knowledge graph by lines to mark the relationship between subjects; the integrated view simultaneously includes the patient's basic diagnosis and treatment information, the multi-subject linkage execution status, resource preparation progress, difference explanations and subject association logic, forming a closed-loop linkage status view of the target personnel covering diagnosis and treatment background, linkage execution, resource guarantee and association logic.

[0164] The beneficial effects include comprehensively integrating the execution status and resource preparation progress of the target terminal, standardizing the response status information format to form a standard dataset, accurately identifying the differences between the linkage status and the expectations, constructing a visual initial linkage status view, linking structured data profiles and collaborative diagnosis and treatment knowledge graphs, forming a closed-loop linkage status view covering the diagnosis and treatment background, linkage execution, resource guarantee, and linkage logic, clearly presenting the entire process of multi-subject linkage, providing comprehensive data support for diagnosis and treatment decisions, and ensuring the accuracy, orderliness, and traceability of multi-subject collaborative diagnosis and treatment for patients with mental disorders and infectious diseases.

[0165] like Figure 2The diagram shown is a functional module diagram of a subject linkage calling system triggered by the attribute characteristics of target personnel, provided in an embodiment of the present invention.

[0166] The subject linkage recall system 100 triggered by the attribute characteristics of target personnel described in this invention can be installed in an electronic device. Depending on the functions implemented, the subject linkage recall system 100 may include a data profile construction module 101, a knowledge graph construction module 102, an intelligent matching analysis module 103, a dynamic scheduling decision module 104, an instruction generation and distribution module 105, and a status monitoring and view generation module 106. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0167] In this embodiment, the functions of each module / unit are as follows:

[0168] The data profile building module 101 is used to clean and parse the multi-dimensional attribute feature data of the target personnel in the multi-subject business terminal to obtain the structured data profile of the target personnel.

[0169] The knowledge graph construction module 102 is used to construct a collaborative diagnosis and treatment knowledge graph of the target personnel based on clinical normative information, expert prior knowledge, and historical case literature in the multi-subject business terminal, combined with a preset rule engine.

[0170] The intelligent matching analysis module 103 is used to map the key attribute features in the structured data profile to the collaborative diagnosis and treatment knowledge graph for matching analysis, so as to obtain the set of subjects to be linked for the target person;

[0171] The dynamic scheduling decision module 104 is used to prioritize and link the set of subjects to be linked based on the cross-organizational resource scheduling strategy of the multi-subject business terminal, so as to obtain the linkage call sequence of the multi-subject business terminal.

[0172] The instruction generation and distribution module 105 is used to encode the linkage call sequence into control instructions to respond to the service preparation and resource preset operations of the multi-subject business terminal, and to integrate the service preparation and resource preset operations into the response status information of the control instructions.

[0173] The status monitoring and view generation module 106 is used to form a closed-loop linkage status view of the target personnel based on the response status information.

[0174] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0175] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0176] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0177] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0178] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0179] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for triggering subject-based linkage based on the attribute characteristics of target personnel, characterized in that: The method includes: S1. Clean and analyze the multi-dimensional attribute feature data of the target personnel in the multi-subject business terminal to obtain a structured data profile of the target personnel; S2. Based on the clinical standard information, expert prior knowledge, and historical case literature in the multi-subject business terminal, and combined with the preset rule engine, construct the collaborative diagnosis and treatment knowledge graph of the target personnel; S3. Map the key attribute features in the structured data profile to the collaborative diagnosis and treatment knowledge graph for matching analysis to obtain the set of subjects to be linked for the target person, including: Extract key attribute features from the structured data profile that are relevant to the current diagnosis and treatment decision. The key attribute features include the main diagnosis, core symptoms, and key examination and testing indicators. Using the key attribute features as query entities, entity matching and positioning are performed in the collaborative diagnosis and treatment knowledge graph to obtain the medical entity nodes of the target person; Based on the collaborative diagnosis and treatment knowledge graph, the subject entity nodes of the structured data profile are obtained by multi-hop graph traversal and identification starting from the medical entity nodes; Based on the preset linkage triggering rules, the subject entity nodes are filtered and aggregated to obtain the type weight and path depth of the association relationship edges of the collaborative diagnosis and treatment knowledge graph; The type weights and path depths are prioritized to obtain the set of subjects to be linked for the target personnel. S4. Based on the cross-institutional resource scheduling strategy of the multi-subject business terminal, prioritize and link the set of subjects to be linked to obtain the linkage call sequence of the multi-subject business terminal, including: Obtain resource status data from the real-time resource monitoring interface in the multi-subject business terminal; Based on the rules defined in the cross-organizational resource scheduling strategy of the multi-subject business terminal, the resource status data of different subjects in the multi-subject business terminal are compared and analyzed with the standard service capability baseline to obtain the dynamic scheduling parameters of the multi-subject business terminal. Based on the dynamic scheduling parameters, the urgency of the associated treatment paths between subjects in the set of subjects to be linked is coupled and evaluated to obtain a comprehensive linkage priority score for the set of subjects to be linked. Based on the comprehensive linkage priority score, the linkage call sequence of the linkage subjects in the set of linkage subjects is sorted and normalized to obtain the linkage call sequence of the linkage subjects. S5. Encode the linked call sequence into a control command to respond to the service preparation and resource preset operations of the multi-subject business terminal, and integrate the service preparation and resource preset operations into the response status information of the control command; S6. Based on the response status information, form a closed-loop linkage status view of the target personnel.

2. The subject linkage invocation method based on the attribute characteristics of the target personnel as described in claim 1, characterized in that, The step of cleaning and parsing the multi-dimensional attribute feature data of the target personnel in the multi-subject business terminal to obtain a structured data profile of the target personnel includes: In the multi-subject business terminal, multi-source heterogeneous attribute data of the target personnel are collected in parallel. The multi-source heterogeneous attribute data includes at least text-formatted medical records, numerical-formatted vital sign time-series data, and structured form-formatted personal basic information. The multi-source heterogeneous attribute data is subjected to data cleaning operations, which include filling in missing data values, removing outliers that exceed a preset reasonable range, and merging duplicate data records for the same semantic entity to obtain a standard dataset of the multi-source heterogeneous attribute data. The unstructured text data in the standard dataset is subjected to key information entity extraction and standardized encoding operations. The extracted entity information is mapped to a preset standardized medical terminology encoding set to generate a standardized feature vector of the unstructured text data. The standardized feature vector is associated and aggregated with the standard dataset, and the aggregated data is labeled to obtain the feature identifier of the multi-source heterogeneous attribute data. Based on the feature identifiers, construct a structured data profile of the target personnel.

3. The subject linkage invocation method based on the attribute characteristics of the target personnel as described in claim 1, characterized in that, The method of constructing a collaborative diagnosis and treatment knowledge graph for the target personnel based on clinical norms information, expert prior knowledge, and historical case literature in the multi-subject business terminal, combined with a preset rule engine, includes: The clinical standard information documents, expert prior knowledge base, and historical case literature set in the multi-subject business terminal are used as diverse knowledge sources for constructing the knowledge graph. Natural language processing is performed on the historical case literature set, and the case entities, diagnosis and treatment operation entities, and the case entities and diagnosis and treatment operation entities recorded in the processed historical case literature set are collaboratively analyzed and integrated to obtain the historical case triplet set of the historical case literature set. The relevant infectious disease prevention and control guidelines, mental disorder diagnosis and treatment guidelines, in-hospital emergency procedures, and medical consortium referral agreements for the target personnel are compiled into the rule engine of the multi-subject business terminal. Based on the rule engine, the mandatory treatment path defined in the clinical normative information document, the conditional reasoning branch in the expert prior knowledge base, and the historical case triple set are logically associated and conflict resolved to obtain the treatment knowledge triple set of the target personnel. Using the disease type and treatment stage of the target person as the core nodes, the treatment knowledge triples are organized and stored in a graph structure to construct a collaborative treatment knowledge graph for the target person.

4. The subject linkage invocation method based on the attribute characteristics of the target personnel as described in claim 3, characterized in that, The process involves performing natural language processing on the historical case literature set, and then performing collaborative analysis and integration on the case entities, treatment operation entities, and the case entities and treatment operation entities recorded in the processed historical case literature set to obtain a set of historical case triplets, including: Semantic parsing is performed on the unstructured text in the historical case literature collection to obtain the text sequence of the unstructured text; The text sequence is analyzed and read to obtain the case entities and treatment operation entities of the historical case literature set; Dependency parsing and semantic role labeling are performed on the case entity and the diagnosis and treatment operation entity to obtain the diagnosis and treatment relationship between the case entity and the diagnosis and treatment operation entity; By binding the case entity, the diagnosis and treatment operation entity, and the diagnosis and treatment relationship, the historical case triplet set of the historical case literature set is obtained.

5. The subject linkage invocation method based on the attribute characteristics of the target personnel as described in claim 1, characterized in that, The formula for calculating the comprehensive linkage priority score is as follows: ； In the formula, For the first The comprehensive linkage priority score of each subject to be linked This refers to the index identifier of a specific subject in the set of subjects to be linked. For the first The quantification value of the urgency level of the treatment pathways associated with each subject to be linked. This is a preset urgency amplification factor. It is a natural exponential function. The dynamic resource load rate of the multi-subject business terminal corresponding to the i-th subject to be linked. This is the preset resource load sensitivity coefficient.

6. The subject linkage invocation method based on the attribute characteristics of the target personnel as described in claim 1, characterized in that, The step of encoding the linked call sequence into control commands to respond to the service preparation and resource pre-setting operations of the multi-subject business terminal, and integrating the service preparation and resource pre-setting operations into response status information of the control commands, includes: The core elements of the linked call sequence are parsed to obtain the target business terminal network address, business interface protocol type and execution time window suggestion of the linked call sequence, and a basic scheduling parameter set is generated. Based on the business interface protocol type, determine the standard instruction template for the linked call sequence; The scheduling information from the basic scheduling parameter set is filled into the standard instruction template to obtain the initial format control instruction of the basic scheduling parameter set; Based on the target service terminal network address, the initial format control instruction is encapsulated to obtain the final control instruction of the multi-subject service terminal; Receive execution response information from the final control instruction in the target business terminal, and update the execution status of the subject to be linked in the linkage call sequence based on the execution response information; The service preparation and resource pre-configuration operations in the execution state are integrated into the response status information of the control command.

7. The subject linkage invocation method based on the attribute characteristics of the target personnel as described in claim 1, characterized in that, The step of forming a closed-loop linkage status view of the target person based on the response status information includes: The execution status identifier and resource preparation progress data of the target terminal in the multi-subject business terminal are used as response status information; The response status information is formatted to obtain a standard status dataset for the control command. The standard state dataset is compared and analyzed with the expected linkage state in the linkage call sequence to obtain the linkage state difference set of the standard state dataset; Based on the set of linkage state differences and the set of standard state datasets, a visualization data structure is constructed according to preset view construction rules to obtain the initial linkage state view of the target personnel. The initial linkage status view is associated and integrated with the structured data profile and the collaborative diagnosis and treatment knowledge graph to obtain the closed-loop linkage status view of the target person.

8. A subject-linked recall system triggered by the attribute characteristics of target personnel, characterized in that: The system is used to implement the subject linkage invocation method based on the attribute characteristics of the target personnel as described in claim 1, the system comprising: The data profiling module is used to clean and parse the multi-dimensional attribute feature data of target personnel in multi-subject business terminals to obtain a structured data profile of the target personnel. The knowledge graph construction module is used to construct a collaborative diagnosis and treatment knowledge graph for the target personnel based on clinical norms information, expert prior knowledge, and historical case literature in the multi-subject business terminal, combined with a preset rule engine. The intelligent matching and analysis module is used to map key attribute features in the structured data profile to the collaborative diagnosis and treatment knowledge graph for matching and analysis, thereby obtaining the set of subjects to be linked for the target personnel. Specifically, it is used for: Extract key attribute features from the structured data profile that are relevant to the current diagnosis and treatment decision. The key attribute features include the main diagnosis, core symptoms, and key examination and testing indicators. Using the key attribute features as query entities, entity matching and positioning are performed in the collaborative diagnosis and treatment knowledge graph to obtain the medical entity nodes of the target person; Based on the collaborative diagnosis and treatment knowledge graph, the subject entity nodes of the structured data profile are obtained by multi-hop graph traversal and identification starting from the medical entity nodes; Based on the preset linkage triggering rules, the subject entity nodes are filtered and aggregated to obtain the type weight and path depth of the association relationship edges of the collaborative diagnosis and treatment knowledge graph; The type weights and path depths are prioritized to obtain the set of subjects to be linked for the target personnel. The dynamic scheduling decision module is used to prioritize and link the set of subjects to be linked based on the cross-organizational resource scheduling strategy of the multi-subject business terminals, thereby obtaining the linkage call sequence of the multi-subject business terminals. Specifically, it is used for: Obtain resource status data from the real-time resource monitoring interface in the multi-subject business terminal; Based on the rules defined in the cross-organizational resource scheduling strategy of the multi-subject business terminal, the resource status data of different subjects in the multi-subject business terminal are compared and analyzed with the standard service capability baseline to obtain the dynamic scheduling parameters of the multi-subject business terminal. Based on the dynamic scheduling parameters, the urgency of the associated treatment paths between subjects in the set of subjects to be linked is coupled and evaluated to obtain a comprehensive linkage priority score for the set of subjects to be linked. Based on the comprehensive linkage priority score, the linkage call sequence of the linkage subjects in the set of linkage subjects is sorted and normalized to obtain the linkage call sequence of the linkage subjects. The instruction generation and distribution module is used to encode the linkage call sequence into control instructions to respond to the service preparation and resource pre-setting operations of the multi-subject business terminal, and to integrate the service preparation and resource pre-setting operations into the response status information of the control instructions; The status monitoring and view generation module is used to form a closed-loop linkage status view of the target personnel based on the response status information.

Images