Outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP

Abstract

The present invention relates to the field of medical system management technology, and in particular, provides an outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP. The system includes a knowledge graph construction component configured to integrate multi-source heterogeneous data such as medical records, medical insurance and physical signs, and to clean, normalize and structure the data through NLP technology to construct a chronic disease medical data knowledge graph; the diagnosis and treatment plan matching component is configured to perform semantic analysis on medical insurance policy documents and diagnosis and treatment guidelines through NLP technology, and match the diagnosis and treatment plan and medical insurance reimbursement path in combination with the patient's actual condition and medical insurance reimbursement record; the analysis and early warning component is configured to analyze the long-term health data of chronic disease patients, identify potential health risks and development trends, and provide personalized health management suggestions and early warning prompts. The present invention realizes multi-source data integration and knowledge graph construction through NLP technology, and solves the problem of data heterogeneity and semantic association requirements.

Description

Technical Field

[0001] The present invention relates to the technical field of medical system management, and in particular to an outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP (Natural Language Processing). Background Art

[0002] As the population ages, chronic diseases (such as hypertension, diabetes, cardiovascular and cerebrovascular diseases, etc.) have become a major burden on the global medical system. Chronic disease management requires long-term tracking of patient vital signs, medication compliance and medical insurance reimbursement data, but the traditional medical system has the following problems: Data is scattered and unstructured: Patient medical records, medical insurance records, and vital sign monitoring data are scattered in different systems, and most of them are unstructured texts (such as handwritten medical records and voice medical orders), which are difficult to integrate and analyze efficiently. Low interaction efficiency: Chronic disease patients need frequent follow-up visits and prescriptions, but outpatient registration, prescription management, medical insurance review and other processes rely on manual operations, resulting in uneven resource allocation and long waiting times. Lack of personalized services: The existing system is difficult to dynamically adjust the treatment plan based on the patient's individualized data (such as medical history and medication history), and the medical insurance reimbursement process is cumbersome, resulting in poor patient experience. Insufficient intelligence: Traditional technology has limited ability to parse medical texts and cannot extract key information (such as disease risks and medication conflicts) from massive unstructured data, which limits the realization of precision medicine.

[0003] At present, the existing technology has the problem that medical data (such as medical records, medical insurance, and physical signs) are scattered on different platforms and lack unified knowledge graph support, resulting in information fragmentation; for example, the medical insurance review system cannot retrieve the patient's chronic disease medication records in real time and requires manual review. Therefore, the present invention provides an outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP. Summary of the invention

[0004] In order to achieve the above object, the present invention adopts the following technical scheme:

[0005] One aspect of the present invention provides an outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP, comprising:

[0006] The knowledge graph construction component is configured to integrate multi-source heterogeneous data of medical records, medical insurance and physical signs, clean, normalize and structure the data through NLP technology, and build a knowledge graph of chronic disease medical data; it links the patient's basic information, diagnosis and treatment records, medication status and medical insurance reimbursement information to form a medical knowledge network;

[0007] The treatment plan matching component is configured to use NLP technology to perform semantic analysis on medical insurance policy documents and treatment guidelines, and match treatment plans and medical insurance reimbursement paths based on the patient's actual condition and medical insurance reimbursement records;

[0008] The analysis and early warning component is configured to analyze the long-term health data of patients with chronic diseases and identify potential health risks and development trends. It can detect abnormal changes through semantic analysis of patients' medical records, vital signs data and medication status, and provide personalized health management suggestions and early warning prompts.

[0009] In an optional implementation, the knowledge graph construction component includes:

[0010] The multi-source data integration module is configured to use semantic parsing technology to identify and extract key entities of disease names, symptom descriptions and medication records in text data from multi-source heterogeneous data, and extract reimbursement items and expense details from medical insurance forms; normalize the extracted entity information and convert unstructured or semi-structured data into a unified structured format; and mine the relationship between entities through semantic association technology;

[0011] The knowledge fusion application module is configured to integrate medical data with information on social security policies, living habits, and environmental factors; it uses a graph data structure for storage and supports efficient node and path queries;

[0012] The dynamic update optimization module is configured to capture and update multi-source heterogeneous changes through a real-time monitoring mechanism after the knowledge graph is built; and to continuously optimize semantic association rules and reasoning logic through a feedback mechanism.

[0013] In an optional implementation, the multi-source data integration module includes:

[0014] The text entity strength calculation submodule is configured to calculate the term frequency-inverse document frequency TF-IDF value for each entity in the text, and evaluate the importance of a word to a document set or one of the documents in a corpus; divide the TF-IDF value of each entity by its maximum value, perform normalization, and multiply it by the preset weight in the knowledge graph to assign different weights to different entities;

[0015] The heterogeneous data compensation item calculation submodule is configured to calculate the difference between the table data and the text entity; measure the difference between the heterogeneous data by calculating the sum of squares of the Euclidean distances between the two; process the sum of squares through a hyperbolic tangent function to adjust the difference compensation value between the heterogeneous data;

[0016] The weighted summation submodule is configured to add the text entity strength and the heterogeneous data compensation term and multiply by the adjustment factor to obtain a score of cross-modal entity alignment of entity importance and heterogeneous data differences in a text; the text entity strength evaluates the importance of a word to a document set or corpus by calculating the term frequency-inverse document frequency TF-IDF value for each entity in the text; the TF-IDF value of each entity is divided by the maximum value, normalized, and multiplied by the preset weight in the knowledge graph; the preset weight is set based on the experience of domain experts or the results obtained through data training;

[0017] The heterogeneous data compensation item is to calculate and make up for the difference between the two types of data. The difference between the heterogeneous data is measured by calculating the sum of the squares of the Euclidean distances between the table data and the text entities. The square sum is processed using the hyperbolic tangent function to adjust the difference compensation value between the heterogeneous data.

[0018] The weighted summation submodule adds the text entity strength and the heterogeneous data compensation term and multiplies it by an adjustment factor to obtain a score for cross-modal entity alignment based on the importance of entities in the text and the differences in heterogeneous data; the setting of the adjustment factor is based on repeated trials and adjustments based on experimental data.

[0019] In an optional implementation, the cross-modal entity alignment formula of the multi-source data integration module is:

[0020]

[0021] In the formula, represents the entity alignment function, which accepts the input entity set E and outputs the alignment result; ∏ represents the product symbol, which means to perform the product operation on the following expression; i=1 to n: represents all i Traverse, n is the total number; Representing Entities t No. i attribute value; It means to find the maximum TF-IDF value of all attributes k; represents the weight parameter; represents the balance parameter, which is used to adjust the influence of attribute similarity and feature vector difference in entity alignment; represents the hyperbolic tangent function; j=1 to m means traversing all features j, and m is the total number of features; represents the square of the Euclidean distance between the feature vector in the table and the feature vector in the text; Is the first j The vector representation of features, is the vector representation of the corresponding text data.

[0022] In an optional implementation, the diagnosis and treatment plan matching component includes:

[0023] The first matching degree calculation module is configured to obtain the number of disease characteristics, determine the weight of each disease characteristic according to its influence on the diagnosis and treatment plan, and calculate the matching degree with the diagnosis and treatment guidelines by defining a function to represent the matching degree with the diagnosis and treatment guidelines; multiply the matching degrees of all disease characteristics by their weights, and sum them to obtain the matching degree of the entire disease characteristic with the diagnosis and treatment guidelines; Indicates Symptoms Diagnosis and treatment guidelines The matching function between

[0024] The second matching degree calculation module is configured to determine the weight of each combination of medical insurance policy and reimbursement path according to the economic impact of its reimbursement, define a function to represent its matching degree, and calculate the matching; multiply the matching degree of all medical insurance policy and reimbursement path combinations by their weights, and sum them to obtain the matching degree of the entire medical insurance policy and reimbursement path;

[0025] The plan matching module is configured to multiply the matching degree of the disease characteristics with the diagnosis and treatment guidelines, the matching degree of the medical insurance policy and the reimbursement path, and then multiply it by the risk adjustment factor. The final result is the matching degree of the diagnosis and treatment plan.

[0026] In an optional implementation, the first matching degree calculation module includes:

[0027] The first feature weight calculation submodule is configured to identify all features of an individual's condition, classify and quantify them; obtain the characteristics of the condition by collecting patient information, including symptoms, signs, and laboratory test results; evaluate the impact of each feature on the diagnosis and treatment results, and assign a weight value to each feature;

[0028] The first function definition submodule is configured to define a function for each disease feature to represent the degree of match between the disease feature and the diagnosis and treatment guideline; the function will quantify the consistency between the disease feature and the diagnosis and treatment guideline, and calculate the degree of match between each disease feature and the diagnosis and treatment guideline; Indicates Symptoms Diagnosis and treatment guidelines The matching function between

[0029] The first matching degree weighted submodule is configured to multiply the matching degree of each disease feature by its corresponding weight value, and sum the weighted matching degrees of all features; and obtain the overall matching degree of the entire disease feature set with the diagnosis and treatment guidelines. The overall matching degree is a comprehensive indicator that reflects the overall degree of compliance between the patient's disease characteristics and the diagnosis and treatment guidelines.

[0030] In an optional implementation, wherein: represents the set of all the patient's disease characteristics, where m is the total number of features; Represents the set of weight values ​​corresponding to each disease feature; Represents the set of matching functions corresponding to each disease feature; Indicates diagnosis and treatment guidelines; Indicates the overall matching degree between the disease feature set and the diagnosis and treatment guidelines;

[0031] Define the calculation formula for the matching degree between the disease characteristics and the diagnosis and treatment guidelines:

[0032]

[0033] In the formula, Indicates Symptoms Diagnosis and treatment guidelines The matching function between the disease characteristics and the diagnosis and treatment guidelines is quantified; specifically, further defined for:

[0034]

[0035] In the formula, Indicates treatment guidelines Symptoms and symptoms The similarity function between them, which quantifies the degree of similarity between them; It is The saturation parameter of each feature is used to control the steepness of the matching curve; and It is The linear transformation parameters of the features are used to adjust the output range of the similarity function; finally, the weight Indicates The influence of each feature on the diagnosis and treatment results is calculated by the following formula:

[0036]

[0037] In the formula, Indicates The original quantized value of each feature; , and It is The weight adjustment parameter of each feature.

[0038] In an optional implementation, the second matching degree calculation module includes:

[0039] The second feature weight calculation submodule is configured to obtain all medical insurance policies and their corresponding reimbursement paths, and to assign weights to each medical insurance policy and reimbursement path combination according to its impact on the patient's economic burden; the economic impact is reflected in the dimensions of reimbursement ratio, self-paid amount, and reimbursement scope; through comprehensive analysis, a weight value is assigned to each policy-path combination;

[0040] The second function definition submodule is configured to define a matching function for each combination of medical insurance policy and reimbursement path to quantify the degree of fit between the combination and the patient's economic affordability and treatment needs; the input of the matching function includes the economic attributes of the policy, the patient's economic conditions, and the feasibility of the treatment path; the matching function can output a value indicating the specific matching degree of the policy-path combination;

[0041] The second matching degree weighted submodule is configured to combine the matching degree with its corresponding weight value after completing the matching degree calculation of each combination of medical insurance policy and reimbursement path, multiply the matching degree of each combination by its weight value, and sum up the weighted matching degrees of all combinations to comprehensively reflect the economic impact and matching degree of all medical insurance policies and reimbursement paths, and finally obtain an overall matching degree value.

[0042] In an optional implementation, the solution matching module includes:

[0043] The matching degree calculation submodule is configured to multiply the matching degree between each disease characteristic and the diagnosis and treatment guideline and the economic impact of all medical insurance policies and reimbursement pathways with the matching degree to obtain a preliminary comprehensive matching degree, which reflects the comprehensive consideration of disease characteristics and economic factors;

[0044] an adjustment factor definition submodule configured to employ a risk adjustment factor adjusted according to a risk assessment of the patient;

[0045] The scheme matching degree calculation submodule is configured to multiply the preliminary comprehensive matching degree by the risk adjustment factor to obtain the matching degree of the diagnosis and treatment scheme.

[0046] In an optional implementation, the analysis and early warning component includes:

[0047] The data collection module is configured to collect various medical records, physical data and medication status of the patient, including the patient's medical history, blood pressure, heart rate, blood sugar physiological indicators, and all medications currently in use;

[0048] The data recognition module is configured to conduct longitudinal analysis of various medical records, vital sign data and medication status, identify patterns and trends through statistics, and find out the changing patterns of patients' health status;

[0049] The suggestion output module is configured to give personalized health management suggestions based on the analysis results, including adjusting medication regimens, changing lifestyle habits and exercising.

[0050] This invention uses NLP technology to achieve multi-source data integration and knowledge graph construction, solving data heterogeneity problems and semantic association requirements; based on knowledge graphs and semantic analysis, it provides diagnosis and treatment plan matching and health risk assessment to assist medical decision-making and management; it uses data analysis technology to achieve long-term dynamic monitoring and early warning of patients with chronic diseases, and improve the initiative and accuracy of medical services. It not only improves the management and application efficiency of medical data, but also provides technical support for the precision medicine and personalized management of chronic diseases, and promotes the intelligent transformation of medical services. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0052] Figure 1 This is a block diagram of the outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP provided in Example 1 of the present invention;

[0053] Figure 2 A block diagram of a component for constructing a knowledge graph provided in Example 2 of the present invention;

[0054] Figure 3 This is a block diagram of a multi-source data integration module provided in Embodiment 3 of the present invention;

[0055] Figure 4 This is a block diagram of a diagnosis and treatment plan matching component provided in Example 4 of the present invention;

[0056] Figure 5 This is a block diagram of a first matching degree calculation module provided in Embodiment 5 of the present invention;

[0057] Figure 6 This is a block diagram of a second matching degree calculation module provided in Embodiment 6 of the present invention;

[0058] Figure 7 This is a block diagram of the scheme matching module provided in Example 7 of the present invention;

[0059] Figure 8 This is a block diagram of the analysis and warning components provided in Example 8 of the present invention;

[0060] Fig. 9 A block diagram of an electronic device provided by the present invention;

[0061] Fig.10 A block diagram of a computer-readable storage medium provided for the present invention. DETAILED DESCRIPTION

[0062] The technical solutions in the embodiments of the present invention will be described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0063] In the following, the terms "first", "second", etc. are used only for convenience of description and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first", "second", etc. may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0064] In the present invention, unless otherwise clearly specified and limited, the term "connection" should be understood in a broad sense, for example, "connection" can be a fixed mechanical connection, or a detachable mechanical connection, or integrated; or, "connection" can be a direct connection, or an indirect connection through an intermediate medium. In addition, unless otherwise clearly specified and limited, the term "coupling" should be understood in a broad sense, for example, "coupling" can be a direct electrical connection, such as physical contact and electrical conduction between two components, and can also be understood as electrical connection between different components in a circuit structure through physical lines such as copper foil or wires on a printed circuit board (PCB) that can transmit electrical signals to transmit electrical signals; or, "coupling" can be an indirect electrical connection between two components through an intermediate medium; or, "coupling" can be an electrical connection between two components in an air-spaced / non-contact manner, for example, two components are electrically connected by capacitive coupling to transmit electrical signals.

[0065] In the embodiments of the present invention, directional terms such as "up", "down", "left" and "right" may be defined including but not limited to the orientation relative to the schematic placement of the components in the drawings. It should be understood that these directional terms may be relative concepts, which are used for relative description and clarification, and may change accordingly according to the change of the orientation of the components in the drawings.

[0066] Embodiment 1:

[0067] like Figure 1 As shown, the embodiment of the present invention provides an outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP, including:

[0068] The knowledge graph construction component is configured to integrate multi-source heterogeneous data such as medical records, medical insurance and physical signs, clean, normalize and structure the data through NLP technology, and build a knowledge graph of chronic disease medical data; it links the patient's basic information, diagnosis and treatment records, medication status and medical insurance reimbursement information to form a medical knowledge network;

[0069] The treatment plan matching component is configured to use NLP technology to perform semantic analysis on medical insurance policy documents and treatment guidelines, and match treatment plans and medical insurance reimbursement paths based on the patient's actual condition and medical insurance reimbursement records;

[0070] The analysis and early warning component is configured to analyze the long-term health data of patients with chronic diseases and identify potential health risks and development trends. It can detect abnormal changes through semantic analysis of patients' medical records, vital signs data and medication status, and provide personalized health management suggestions and early warning prompts.

[0071] In the above embodiments, the knowledge graph construction component uses NLP technology (such as entity recognition, relationship extraction, text classification, etc.) to clean, normalize and structure heterogeneous data such as medical records, medical insurance, and physical signs to achieve data standardization; uses graph databases and graph algorithms to model entities such as patient basic information, medical records, medication conditions, medical insurance reimbursement, and their relationships into knowledge graphs to build a chronic disease medical knowledge network; uses disambiguation, entity alignment and other technologies to solve semantic inconsistencies in multi-source data and achieve cross-system data association. Significance: Eliminate data silos and achieve cross-system and cross-domain data integration; transform medical data into understandable and computable knowledge to provide support for subsequent analysis and decision-making; improve data query and reasoning capabilities to lay the foundation for precision medicine. The diagnosis and treatment plan matching component uses NLP technology (such as text understanding, keyword extraction, and semantic similarity calculation) to parse medical insurance policy documents and diagnosis and treatment guidelines, extract key rules and constraints; through rule engines and reasoning algorithms, combined with the patient's actual condition (such as disease type, medication records) and medical insurance reimbursement, match the optimal diagnosis and treatment plan and medical insurance reimbursement path; calculate the patient's out-of-pocket expenses based on the matching results and medical insurance reimbursement ratio, and provide detailed cost details. Significance: Convert complex medical insurance policies into executable rules to improve policy implementation efficiency; provide targeted diagnosis and treatment plans for patients and doctors, optimize medical decision-making; reduce medical expenses and improve the efficiency of medical insurance funds. The analysis and early warning component uses NLP technology (such as time series analysis, anomaly detection, and trend prediction) to mine patients' long-term health data (such as medical records, physical signs, and medication) to identify potential health risks and development trends; through semantic analysis of patients' key indicators (such as blood pressure, blood sugar, and heart rate), abnormal changes are found and early warning mechanisms are triggered; based on the analysis results, personalized health management suggestions (such as diet, exercise, and medication optimization) are generated, and early warning prompts are provided. Significance: Timely identification of patients' health risks and reduction of the probability of complications; providing a scientific basis for chronic disease management and improving management effectiveness; enhancing patients' self-management capabilities and improving health outcomes through personalized recommendations.

[0072] In summary, this embodiment uses NLP technology to achieve multi-source data integration and knowledge graph construction, solves data heterogeneity problems and semantic association needs; based on knowledge graphs and semantic analysis, provides diagnosis and treatment plan matching and health risk assessment, assists medical decision-making and management; uses data analysis technology to achieve long-term dynamic monitoring and early warning of chronic disease patients, and improves the initiative and accuracy of medical services. It not only improves the management and application efficiency of medical data, but also provides technical support for the precision medicine and personalized management of chronic diseases, and promotes the intelligent transformation of medical services.

[0073] Embodiment 2:

[0074] like Figure 2As shown, based on Example 1, the knowledge graph construction component provided by the embodiment of the present invention includes:

[0075] The multi-source data integration module is configured to use semantic parsing technology to identify and extract key entities in text data (such as disease names, symptom descriptions, and medication records) from multi-source heterogeneous data, and extract reimbursement items and expense details from medical insurance forms; normalize the extracted entity information and convert unstructured or semi-structured data into a unified structured format; and mine the relationship between entities through semantic association technology;

[0076] The knowledge fusion application module is configured to integrate medical data with information on social security policies, living habits, and environmental factors. It uses a graph data structure for storage and supports efficient node and path queries.

[0077] The dynamic update optimization module is configured to capture and update multi-source heterogeneous changes through a real-time monitoring mechanism after the knowledge graph is built; and to continuously optimize semantic association rules and reasoning logic through a feedback mechanism.

[0078] Among them, the graph structure semantic constraint equation of the knowledge fusion application module is:

[0079]

[0080]

[0081] In the formula, It represents the graph structure semantic constraint result of the knowledge fusion application module and the comprehensive score of the knowledge graph; It represents the lifestyle factor, such as smoking is 0.6, BMI>30 is 0.8, which indicates the weight of the impact of lifestyle habits on the patient's health; Represents the edge weight calculation function, which is used to measure the patient node and disease nodes The strength of the association between represents the patient node, which represents the patient's information; represents a disease node, which indicates the information of the disease; The Social Security Administration The frequency of changes in drug reimbursement policies reflects the impact of policies on medical expenses; The environmental pollution index indicates the impact of The impact weight represents the contribution of environmental factors to health; Represents the total number of lifestyle factors; represents the total number of drug reimbursement policies; , Represents the 32-dimensional patient sign encoding generated based on GraphSAGE, representing the patient nodes and disease nodes The embedding vector of It represents the Euclidean distance between the patient's physical sign code and the disease code, reflecting the similarity between the two; Indicates patient and disease The matching degree of social security policies is used to measure the support of policies for patient treatment; Indicates patient medical expenses; Represents the threshold of medical expenses, which is used to control the calculation range of edge weights; represents the Sigmoid function, which is used to map the cost difference to the interval [0,1];

[0082] The composite weight optimization equation of the dynamic update optimization module is expressed as:

[0083]

[0084] The parameter update strategy is expressed as:

[0085]

[0086] In the formula, Indicates the weight change at the t+1th iteration; represents the momentum factor, which controls the retention ratio of historical weight changes; Indicates t The weight change at iterations; Indicates k The semantic change intensity of a time window reflects the timeliness change of the knowledge graph; Indicates k The delay time of a time window indicates the lag of data update; Indicates k Feedback information of each time window reflects the optimization requirements of semantic association and reasoning logic; A mapping function representing feedback information, used to convert feedback into contribution value of weight change; K Indicates the total number of time windows; Represents the updated weight value; Indicates the current weight value; Represents the learning rate, which controls the magnitude of weight updates; Indicates t +1 weight change for each iteration; Represents the attenuation coefficient, which controls the convergence speed when the weight is updated; Indicates the Euclidean distance between the current weight value and the reference weight value, reflecting the degree of deviation of the weight; Represents the reference weight value, which represents the weight value under ideal conditions.

[0087] In the above embodiments, the multi-source data integration module solves the format and semantic inconsistency problems of multi-source heterogeneous data, providing basic support for knowledge fusion and application; improving the computability and comprehensibility of medical data, laying a semantic foundation for the construction of knowledge graphs; establishing logical relationships between entities, and providing support for reasoning and query of medical knowledge. The knowledge fusion application module enriches the information dimension of the knowledge graph through cross-domain knowledge fusion and enhances its value in practical applications; using the characteristics of the graph data structure, it supports fast and accurate query and reasoning to meet the needs of complex medical scenarios; provides comprehensive support for medical decision-making, and optimizes diagnosis and treatment plans and health management strategies. The dynamic update optimization module ensures that the knowledge graph can reflect the updates and changes of medical data in real time and adapt to the rapidly developing medical scenarios; improves the intelligence level of the knowledge graph through the feedback mechanism and enhances its application capabilities in complex scenarios; improves the accuracy and adaptability of the knowledge graph through iterative optimization, and provides long-term support for medical decision-making.

[0088] To summarize, this embodiment solves the problems of information dispersion and semantic inconsistency through a multi-source data integration module, expands the cross-domain application capabilities of the knowledge graph through a knowledge fusion application module, and ensures the dynamic and intelligent nature of the knowledge graph through a dynamic update optimization module; through modular design and innovative technology, a comprehensive, dynamic, and intelligent chronic disease medical data knowledge graph is constructed, which provides strong technical support for the management, analysis, and decision-making of medical data and promotes the intelligent development of medical services.

[0089] Embodiment 3:

[0090] like Figure 3 As shown, based on Example 2, the multi-source data integration module provided by the embodiment of the present invention includes:

[0091] The text entity strength calculation submodule is configured to calculate the term frequency-inverse document frequency TF-IDF value for each entity in the text, and evaluate the importance of a word to a document set or one of the documents in a corpus; divide the TF-IDF value of each entity by its maximum value, perform normalization, and multiply it by the preset weight in the knowledge graph to assign different weights to different entities;

[0092] The heterogeneous data compensation item calculation submodule is configured to calculate the difference between the table data and the text entity; measure the difference between the heterogeneous data by calculating the sum of squares of the Euclidean distances between the two; process the sum of squares through a hyperbolic tangent function to adjust the difference compensation value between the heterogeneous data;

[0093] The weighted summation submodule is configured to add the text entity strength and the heterogeneous data compensation term and multiply it by the adjustment factor to obtain a score of cross-modal entity alignment based on the entity importance in the text and the heterogeneous data difference.

[0094] Among them, the cross-modal entity alignment formula of the multi-source data integration module is:

[0095]

[0096] In the formula, represents the entity alignment function, which accepts the input entity set E and outputs the alignment result; ∏ represents the product symbol, which means to perform the product operation on the following expression; i=1 to n: represents all i Traverse, n is the total number; Representing Entities t No. i attribute value; It means to find the maximum TF-IDF value of all attributes k; represents the weight parameter; represents the balance parameter, which is used to adjust the influence of attribute similarity and feature vector difference in entity alignment; represents the hyperbolic tangent function; j=1 to m means traversing all features j, and m is the total number of features; represents the square of the Euclidean distance between the feature vector in the table and the feature vector in the text; Is the first j The vector representation of features, is the vector representation of the corresponding text data.

[0097] In the above embodiment, the text entity strength calculation submodule: uses the TF-IDF algorithm to evaluate the importance of each entity in the text, TF-IDF = word frequency / (total number of documents * number of occurrences of the word in all documents), which reflects the contribution and discrimination of a word to a document set or a specific document; divides the TF-IDF value of the entity by the maximum value, so that its range is between 1 and 1, eliminating the impact of dimensions between different entities; combines the preset weights of the knowledge graph to give different entities differentiated importance, so that the model pays more attention to high-value entities. Significance: Improve the accuracy and discrimination of entity recognition, weight important entities, and provide a basis for subsequent cross-modal alignment; integrate the prior information of the knowledge graph into the entity weight calculation to enhance the model's personalized perception of different entities. The heterogeneous data compensation item calculation submodule measures the difference between tabular data and text entities by the sum of the squares of the Euclidean distance, capturing multi-dimensional difference information; the hyperbolic tangent function processes the sum of squares, maps the difference compensation value to the (-1, 1) interval, smoothes the difference, and makes its impact on the final score more controllable. Significance: It effectively captures the heterogeneous differences between heterogeneous data and provides a compensation mechanism for cross-modal alignment; it smoothes the differences through the hyperbolic tangent function to enhance the generalization and robustness of the model. The weighted summation submodule adds the text entity strength and the heterogeneous data compensation term to achieve the fusion of cross-modal features; through the adjustment factor, it dynamically balances the contribution of different submodules to make the final score more reasonable and stable. Significance: Through weighted summation, the effective fusion of heterogeneous data features is achieved, providing a comprehensive feature representation for cross-modal entity alignment; the introduction of the adjustment factor enhances the flexibility and controllability of the model, making the scoring results more in line with actual application needs.

[0098] In the multi-source data integration module, text entity strength and heterogeneous data compensation items work together through calculation methods to ultimately achieve cross-modal entity alignment; text entity strength evaluates the importance of a word to a document set or corpus by calculating the term frequency-inverse document frequency (TF-IDF) value for each entity in the text; words that appear frequently in a specific document but are relatively rare in the entire corpus are given higher TF-IDF values ​​to highlight their unique value; to further enhance this importance, the TF-IDF value of each entity is divided by its maximum value, normalized, and multiplied by the preset weight in the knowledge graph; the preset weights ensure that each entity can play its due role in the entity alignment process; in actual operations, the setting of preset weights is usually based on the experience of domain experts or the results obtained through training with a large amount of data.

[0099] The main task of the heterogeneous data compensation item is to calculate and compensate for the difference between the two types of data to ensure that they can dance harmoniously on the stage of entity alignment. The difference between the heterogeneous data is measured by calculating the sum of the squares of the Euclidean distances between the table data and the text entities. Then, the hyperbolic tangent function is used to process this sum of squares to adjust the difference compensation value between the heterogeneous data. In this way, the heterogeneous data compensation item can sensitively capture the subtle differences between the data;

[0100] The weighted summation submodule adds the two together and multiplies them by an adjustment factor to finally obtain a score for cross-modal entity alignment based on the importance of entities in the text and the difference in heterogeneous data. The setting of this adjustment factor is based on repeated trials and adjustments based on experimental data or optimization algorithms. Its purpose is to find the most appropriate balance point so that the text entity strength and heterogeneous data compensation items can play their respective advantages in the entity alignment process and jointly achieve accurate and efficient cross-modal entity alignment.

[0101] The text entity strength highlights the importance of the entity in the text by comprehensively evaluating the TF-IDF value and preset weight of the entity; while the heterogeneous data compensation term achieves the fusion of different data by calculating and adjusting the differences between the data. When they are weighted and summed and multiplied by the adjustment factor, they together constitute the cross-modal entity alignment.

[0102] In summary, the multi-source data integration module of this embodiment effectively integrates data features of different modalities through the collaborative work of three sub-modules: text entity strength calculation, heterogeneous data compensation, and weighted summation, and realizes the scoring calculation of cross-modal entity alignment.

[0103] Embodiment 4:

[0104] like Figure 4 As shown, based on Example 1, the diagnosis and treatment plan matching component provided in this embodiment of the present invention includes:

[0105] The first matching degree calculation module is configured to obtain the number of disease characteristics, determine the weight of each disease characteristic according to its influence on the diagnosis and treatment plan, and calculate the matching degree with the diagnosis and treatment guidelines by defining a function to represent the matching degree with the diagnosis and treatment guidelines; multiply the matching degrees of all disease characteristics by their weights, and sum them to obtain the matching degree of the entire disease characteristic with the diagnosis and treatment guidelines;

[0106] The second matching degree calculation module is configured to determine the weight of each combination of medical insurance policy and reimbursement path according to the economic impact of its reimbursement, define a function to represent its matching degree, and calculate the matching; multiply the matching degree of all medical insurance policy and reimbursement path combinations by their weights, and sum them to obtain the matching degree of the entire medical insurance policy and reimbursement path;

[0107] The plan matching module is configured to multiply the matching degree of the disease characteristics with the diagnosis and treatment guidelines, the matching degree of the medical insurance policy and the reimbursement path, and then multiply it by the risk adjustment factor. The final result is the matching degree of the diagnosis and treatment plan.

[0108] In the above embodiment, the first matching degree calculation module can comprehensively evaluate the overall impact of the disease characteristics on the diagnosis and treatment plan by weighted summation, and provide a quantitative basis for decision-making; it realizes the basis for personalized matching of the diagnosis and treatment plan, ensures the high consistency between the treatment plan and the patient's specific condition, and improves the pertinence and effect of diagnosis and treatment. The weighted summation method of the second matching degree calculation module can comprehensively evaluate the economic rationality of different medical insurance policy combinations; by considering economic factors, it helps to achieve economic optimization of the diagnosis and treatment plan, ensure that patients can afford the treatment plan, and make rational use of medical resources. The scheme matching degree module can evaluate the comprehensive performance of the treatment plan in terms of disease adaptation and economic rationality through multiplication operations; the final matching degree result provides a decision-making basis for selecting the optimal diagnosis and treatment plan, taking into account both treatment effect and economic cost, which is conducive to the efficient allocation and utilization of medical resources.

[0109] In summary, the diagnosis and treatment plan matching component of this embodiment realizes the matching degree calculation of disease characteristics, medical insurance policies and diagnosis and treatment guidelines through modular design, and provides a quantitative and objective diagnosis and treatment plan selection tool for doctors, patients and medical insurance decision makers through comprehensive evaluation. Such technical implementation not only improves the scientificity and rationality of diagnosis and treatment plans, but also helps to optimize the allocation of medical resources, reduce medical costs, and ultimately improve the overall efficiency and quality of medical services.

[0110] Embodiment 5:

[0111] like Figure 5 As shown, based on Example 4, the first matching degree calculation module provided in this embodiment of the present invention includes:

[0112] The first feature weight calculation submodule is configured to identify all features of an individual's condition, classify and quantify them; obtain the characteristics of the condition by collecting patient information, including symptoms, signs, laboratory test results, etc.; evaluate the impact of each feature on the diagnosis and treatment results, and assign a weight value to each feature;

[0113] The first function definition submodule is configured to define a function for each disease feature to represent the degree of match between the disease feature and the diagnosis and treatment guideline; the function will quantify the consistency between the disease feature and the diagnosis and treatment guideline, and calculate the degree of match between each disease feature and the diagnosis and treatment guideline;

[0114] The first matching degree weighted submodule is configured to multiply the matching degree of each disease feature by its corresponding weight value, and sum the weighted matching degrees of all features; and obtain the overall matching degree of the entire disease feature set with the diagnosis and treatment guidelines. The overall matching degree is a comprehensive indicator that reflects the overall degree of compliance between the patient's disease characteristics and the diagnosis and treatment guidelines.

[0115] Among them, the definition is: represents the set of all the patient's disease characteristics, where m is the total number of features; Represents the set of weight values ​​corresponding to each disease feature; Represents the set of matching functions corresponding to each disease feature; Indicates diagnosis and treatment guidelines; Indicates the overall matching degree between the disease feature set and the diagnosis and treatment guidelines;

[0116] Define the calculation formula for the matching degree between the disease characteristics and the diagnosis and treatment guidelines:

[0117]

[0118] In the formula, Indicates Symptoms Diagnosis and treatment guidelines The matching function between the disease characteristics and the diagnosis and treatment guidelines is quantified; specifically, further defined for:

[0119]

[0120] In the formula, Indicates treatment guidelines Symptoms and symptoms The similarity function between them, which quantifies the degree of similarity between them; It is The saturation parameter of each feature is used to control the steepness of the matching curve; and It is The linear transformation parameters of the features are used to adjust the output range of the similarity function; finally, the weight Indicates The influence of each feature on the diagnosis and treatment results is calculated by the following formula:

[0121]

[0122] In the formula, Indicates The original quantized value of each feature; , and It is The matching degree calculation formula comprehensively considers the similarity between the disease characteristics and the diagnosis and treatment guidelines, the feature weights and the quantitative values ​​of the features, and can more comprehensively evaluate the overall compliance of the patient's disease characteristics with the diagnosis and treatment guidelines.

[0123] In the above embodiment, the first feature weight calculation submodule can automatically identify and classify the characteristics of individual illnesses, including symptoms, signs, and laboratory test results, to ensure the comprehensiveness of the illness data; quantify each illness characteristic to provide basic data for matching calculation; assign weight values ​​to each characteristic according to the degree of influence of the characteristic on the diagnosis and treatment results to ensure that important characteristics occupy a larger proportion in the matching calculation. Significance: Ensure the accuracy and comprehensiveness of the illness characteristic data and provide a reliable basis for matching calculation; through weight allocation, make the matching calculation closer to the actual diagnosis and treatment needs and improve the accuracy of matching. The first function definition submodule defines a function for each illness characteristic, which can represent the degree of matching between the characteristic and the diagnosis and treatment guidelines; through function calculation, quantify the consistency between each illness characteristic and the diagnosis and treatment guidelines. Significance: Make the matching calculation operational, quantify the matching quantity in the form of a function, and facilitate subsequent calculation and comparison; improve the flexibility and adaptability of the matching calculation, and adjust the function definition according to different diagnosis and treatment guidelines. The first matching degree weighted submodule multiplies the matching degree of each disease feature by its corresponding weight value, and sums the weighted matching degrees of all features; the overall matching degree of the entire disease feature set and the diagnosis and treatment guidelines is obtained as a comprehensive indicator. Significance: Through weighted summation, the overall matching degree can reflect the overall compliance of the disease features with the diagnosis and treatment guidelines, which improves the representativeness and accuracy of the matching degree; the comprehensive evaluation results can provide important references for clinical decision-making, help optimize the diagnosis and treatment plan, and improve the diagnosis and treatment effect.

[0124] In summary, the first matching degree calculation module of this embodiment improves the personalization and accuracy of the diagnosis and treatment plan through technical means, helps to achieve precision medicine, and improves treatment effects and patient satisfaction.

[0125] Embodiment 6:

[0126] like Figure 6 As shown, based on Example 4, the second matching degree calculation module provided in this embodiment of the present invention includes:

[0127] The second feature weight calculation submodule is configured to obtain all medical insurance policies and their corresponding reimbursement paths, and to assign weights to each medical insurance policy and reimbursement path combination according to the degree of its impact on the patient's economic burden; the economic impact is reflected in dimensions such as reimbursement ratio, self-paid amount, and reimbursement scope; through comprehensive analysis, a weight value is assigned to each policy-path combination;

[0128] The second function definition submodule is configured to define a matching function for each combination of medical insurance policy and reimbursement path to quantify the degree of fit between the combination and the patient's economic affordability and treatment needs; the input of the matching function includes the economic attributes of the policy (such as reimbursement ratio, reimbursement scope), the patient's economic conditions (such as payment ability, medical budget), and the feasibility of the treatment path (such as reimbursement speed, process complexity); the matching function can output a value to indicate the specific matching degree of the policy-path combination;

[0129] The second matching degree weighted submodule is configured to combine the matching degree with its corresponding weight value after completing the matching degree calculation of each combination of medical insurance policy and reimbursement path, multiply the matching degree of each combination by its weight value, and sum up the weighted matching degrees of all combinations to comprehensively reflect the economic impact and matching degree of all medical insurance policies and reimbursement paths, and finally obtain an overall matching degree value.

[0130] in,

[0131]

[0132] In the formula, It represents the overall matching value, which is a comprehensive reflection of the economic impact and matching degree of all combinations of medical insurance policies and reimbursement paths; represents the total number of combinations of medical insurance policies and reimbursement pathways; Indicates The weight value of a policy-path combination reflects the impact of the combination on the economic burden of patients; represents the matching function, The fit between the policy-pathway combination and the patient's economic affordability and treatment needs; Indicates The economic attributes of individual medical insurance policies, including reimbursement ratios, reimbursement scope, etc.; Indicates The individual patient's financial situation, including ability to pay, medical budget, etc.; Indicates The feasibility of a diagnosis and treatment pathway, including reimbursement speed, process complexity, etc. Indicates The economic impact measure of each policy-pathway combination; It represents the maximum economic impact measurement value among all policy-path combinations. It comprehensively considers the economic attributes of the medical insurance policy, the economic conditions of the patient, and the feasibility of the diagnosis and treatment path, and reflects the overall matching degree of the combination of medical insurance policy and reimbursement path through multi-level and multi-dimensional calculations.

[0133] In the above embodiment, the second feature weight calculation submodule collects and analyzes medical insurance policies and their corresponding reimbursement paths, and conducts quantitative evaluation based on the degree of impact of these policies and paths on the economic burden of patients; by considering dimensions such as reimbursement ratio, self-paid amount, and reimbursement scope, a weight value is assigned to each policy-path combination. Significance: The calculation of weight values ​​helps to identify which policy-path combinations have the greatest economic impact on patients, thereby giving these combinations a higher priority in the matching calculation. The second function definition submodule defines a matching function that quantifies the fit between the combination of medical insurance policies and reimbursement paths and the economic affordability and treatment needs of patients; the matching function considers the economic attributes of the policy, the economic conditions of the patient, and the feasibility of the treatment path, and outputs a numerical value indicating the degree of matching of the policy-path combination. Significance: Through the matching function, the matching degree of different policy-path combinations with the actual situation of the patient can be evaluated, providing a scientific basis for patients to choose the most appropriate medical insurance policy. After completing the matching calculation for each combination of medical insurance policy and reimbursement path, the second matching weighted submodule combines the matching with the corresponding weight value, and obtains a comprehensive matching value by multiplying the matching of each combination by its weight value and summing them up. Significance: The calculation of weighted matching can more comprehensively reflect the economic impact and matching degree of all medical insurance policies and reimbursement paths, provide an overall matching value, and help decision makers or patients choose the most economical and appropriate medical insurance policies and reimbursement paths.

[0134] To sum up, the technical significance of the second matching degree calculation module of this embodiment is to provide a scientific and quantitative evaluation tool for the selection of medical insurance policies and the matching of reimbursement paths, so that patients can choose the most suitable medical insurance plan according to their own economic conditions and treatment needs, thereby maximizing the benefits of medical insurance policies and patient satisfaction.

[0135] Embodiment 7:

[0136] like Figure 7 As shown, based on Example 4, the scheme matching degree module provided in this embodiment of the present invention includes:

[0137] The matching degree calculation submodule is configured to multiply the matching degree between each disease characteristic and the diagnosis and treatment guideline and the economic impact of all medical insurance policies and reimbursement pathways with the matching degree to obtain a preliminary comprehensive matching degree, which reflects the comprehensive consideration of disease characteristics and economic factors;

[0138] an adjustment factor definition submodule configured to employ a risk adjustment factor adjusted according to a risk assessment of the patient;

[0139] The scheme matching degree calculation submodule is configured to multiply the preliminary comprehensive matching degree by the risk adjustment factor to obtain the matching degree of the diagnosis and treatment scheme.

[0140] In the above embodiment, the matching degree calculation submodule realizes the matching degree calculation between the disease characteristics and the diagnosis and treatment guidelines; realizes the economic impact and matching degree calculation of the medical insurance policy and reimbursement path; and performs weighted multiplication operation on the two sets of matching degrees to obtain a preliminary comprehensive matching degree. Significance: It reflects the matching degree between the disease characteristics and the diagnosis and treatment plan; comprehensively considers the impact of economic factors on the selection of diagnosis and treatment plans; and provides a preliminary comprehensive evaluation index to provide a reference for adjustment. The adjustment factor definition submodule calculates the risk adjustment factor according to the patient risk assessment; and performs risk adjustment on the preliminary comprehensive matching degree. Significance: It takes into account the individual differences of patients and improves the personalization of the diagnosis and treatment plan; reduces the risk of high-risk patients accepting inappropriate diagnosis and treatment plans; and improves the safety and effectiveness of the diagnosis and treatment plan. The scheme matching degree calculation submodule performs weighted multiplication operation on the preliminary comprehensive matching degree and the risk adjustment factor to obtain the final diagnosis and treatment plan matching degree. Significance: It comprehensively considers the disease characteristics, economic factors and patient risks; and provides a comprehensive evaluation index to guide the selection and optimization of diagnosis and treatment plans; improves the rationality and pertinence of the diagnosis and treatment plan, which is conducive to improving the diagnosis and treatment effect.

[0141] In summary, this embodiment comprehensively considers the characteristics of the disease, economic factors and patient risks, and realizes the personalization, safety and effectiveness of the diagnosis and treatment plan.

[0142] Embodiment 8:

[0143] like Figure 8 As shown, based on Example 1, the analysis and warning component provided by the embodiment of the present invention includes:

[0144] The data collection module is configured to collect various medical records, physical data and medication status of the patient, including the patient's medical history, blood pressure, heart rate, blood sugar and other physiological indicators, as well as all the medications currently being used;

[0145] The data recognition module is configured to conduct longitudinal analysis of various medical records, vital sign data and medication status, identify patterns and trends through statistics, and find out the changing patterns of patients' health status;

[0146] The suggestion output module is configured to give personalized health management suggestions based on the results of the analysis, including adjusting medication plans, changing lifestyle habits, and exercising.

[0147] In the above embodiment, the data collection module* collects the patient's medical records, physiological indicators (such as blood pressure, heart rate, blood sugar) and medication data in real time; uses sensors, data interfaces and database technology to ensure the integrity and accuracy of the data; the data collection module serves as the input end of the entire system to provide basic data support for analysis and early warning. The data identification module conducts longitudinal analysis on the collected data. This module uses statistical methods and machine learning algorithms to identify patterns and trends in the data, thereby revealing the changing laws of the patient's health status; the data identification module provides a scientific basis for predicting and warning of possible health problems of patients through in-depth analysis of the data, which can improve the accuracy of health management and reduce false positives and false negatives. According to the analysis results of the data identification module, the suggestion output module can generate personalized health management suggestions, including adjusting medication regimens, improving living habits and increasing moderate exercise; the suggestion output module converts the analysis results into specific action guidelines to help patients and medical personnel make more reasonable decisions and improve treatment effects and quality of life.

[0148] In summary, this embodiment realizes real-time monitoring, accurate analysis and effective early warning of the patient's health status, thereby improving the quality and efficiency of medical services, reducing medical costs, and ultimately improving the patient's health level and quality of life. It can provide data support for clinical decision-making, provide patients with personalized health management plans, and provide reference for the optimal allocation of medical resources.

[0149] Fig. 9 A block diagram of an exemplary electronic device suitable for implementing embodiments of the present invention is shown.

[0150] The electronic device may include a central processing unit / microprocessor / main control chip, etc.; a storage medium, coupled to the central processing unit / microprocessor / main control chip, etc., and storing computer executable instructions therein, for performing the steps of each method of an embodiment of the present invention when executed by the processor.

[0151] The central processing unit / microprocessor / main control chip etc. may include but are not limited to, for example, one or more processors or microprocessors etc.

[0152] The storage medium may include, but is not limited to, for example, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, computer storage media (e.g., hard disk, floppy disk, solid-state drive, removable disk, CD-ROM, DVD-ROM, Blu-ray disc, etc.).

[0153] In addition, the electronic device may also include (but not limited to) a data bus, an input / output bus / external bus / device bus, a display, and input / output devices (eg, keyboard, mouse, speaker, etc.), etc.

[0154] The central processing unit / microprocessor / main control chip etc. can communicate with external devices via an I / O bus via a wired or wireless network (not shown).

[0155] The storage medium may also store at least one computer executable instruction for executing the various functions and / or method steps in the embodiments described in the present technology when executed by a central processing unit / microprocessor / main control chip, etc.

[0156] In one embodiment, the at least one computer executable instruction may also be compiled into or constitute a software product, wherein one or more computer executable instructions are executed by a processor to perform the various functions and / or method steps in the embodiments described in the present technology.

[0157] Fig.10 A schematic diagram of a computer-readable storage medium according to an embodiment of the present invention is shown.

[0158] like Fig.10 As shown, instructions are stored on a non-transitory computer-readable storage medium, and the instructions are, for example, computer-readable instructions. When the computer-readable instructions are executed by the processor, the various methods described above can be executed. The non-transitory computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may, for example, include random access memory (RAM) and / or cache memory (cache), etc. The non-transitory non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc. For example, the non-transitory computer-readable storage medium may be connected to a computing device such as a computer, and then, when the computing device runs the computer-readable instructions stored on the computer-readable storage medium, the various methods described above may be performed.

[0159] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0160] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0161] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0162] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for executing all or part of the steps of the various embodiments of the present invention through a computer device (which can be a personal computer, server, or network device, etc.). The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (full name in English: Read-Only Memory, English abbreviation: ROM), random access memory (full name in English: Random Access Memory, English abbreviation: RAM), disk or optical disk, etc. Various media that can store program codes.

[0163] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP, characterized by: Include: The knowledge graph construction component is configured to integrate multi-source heterogeneous data of medical records, medical insurance and physical signs, clean, normalize and structure the data through NLP technology, and build a knowledge graph of chronic disease medical data; it links the patient's basic information, diagnosis and treatment records, medication status and medical insurance reimbursement information to form a medical knowledge network; The treatment plan matching component is configured to use NLP technology to perform semantic analysis on medical insurance policy documents and treatment guidelines, and match treatment plans and medical insurance reimbursement paths based on the patient's actual condition and medical insurance reimbursement records; The analysis and early warning component is configured to analyze the long-term health data of patients with chronic diseases and identify potential health risks and development trends; Through semantic analysis of patients' medical records, physical data and medication status, abnormal changes can be detected and personalized health management suggestions and early warning prompts can be provided; The knowledge graph construction component of the outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP includes: The multi-source data integration module is configured to use semantic parsing technology to identify and extract key entities of disease names, symptom descriptions and medication records in text data from multi-source heterogeneous data, and extract reimbursement items and expense details from medical insurance forms; normalize the extracted entity information and convert unstructured or semi-structured data into a unified structured format; and mine the relationship between entities through semantic association technology; The knowledge fusion application module is configured to integrate medical data with information on social security policies, living habits, and environmental factors; it uses a graph data structure for storage and supports efficient node and path queries; The dynamic update optimization module is configured to capture and update multi-source heterogeneous changes through a real-time monitoring mechanism after the knowledge graph is built; and to continuously optimize semantic association rules and reasoning logic through a feedback mechanism; The multi-source data integration module of the outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP includes: The text entity strength calculation submodule is configured to calculate the term frequency-inverse document frequency TF-IDF value for each entity in the text, and evaluate the importance of a word to a document set or one of the documents in a corpus; divide the TF-IDF value of each entity by its maximum value, perform normalization, and multiply it by the preset weight in the knowledge graph to assign different weights to different entities; The heterogeneous data compensation item calculation submodule is configured to calculate the difference between the table data and the text entity; measure the difference between the heterogeneous data by calculating the sum of squares of the Euclidean distances between the two; process the sum of squares through a hyperbolic tangent function to adjust the difference compensation value between the heterogeneous data; The weighted summation submodule is configured to add the text entity strength and the heterogeneous data compensation term and multiply by the adjustment factor to obtain a score of cross-modal entity alignment of entity importance and heterogeneous data differences in a text; the text entity strength evaluates the importance of a word to a document set or corpus by calculating the term frequency-inverse document frequency TF-IDF value for each entity in the text; the TF-IDF value of each entity is divided by the maximum value, normalized, and multiplied by the preset weight in the knowledge graph; The preset weights are set based on the experience of domain experts or the results obtained through data training; The heterogeneous data compensation item is to calculate and make up for the difference between the two types of data. The difference between the heterogeneous data is measured by calculating the sum of the squares of the Euclidean distances between the table data and the text entities. The square sum is processed using the hyperbolic tangent function to adjust the difference compensation value between the heterogeneous data. The weighted summation submodule adds the text entity strength and the heterogeneous data compensation term and multiplies it by an adjustment factor to obtain a score of cross-modal entity alignment based on the importance of entities in the text and the heterogeneous data differences. The setting of the adjustment factor is based on repeated trials and adjustments based on experimental data.

2. The outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP as claimed in claim 1 is characterized in that: in, The cross-modal entity alignment formula of the multi-source data integration module is: In the formula, Represents the entity alignment function, which accepts the input entity set E and outputs the alignment result; ∏ represents the product symbol, which means to perform product operation on the following expressions; i=1 to n: means to perform product operation on all i Traverse, n is the total number; Representing Entities t No. i attribute value; It means to find the maximum TF-IDF value of all attributes k; represents the weight parameter; represents the balance parameter, which is used to adjust the influence of attribute similarity and feature vector difference in entity alignment; represents the hyperbolic tangent function; j=1 to m means traversing all features j, and m is the total number of features; represents the square of the Euclidean distance between the feature vector in the table and the feature vector in the text; Is the first j The vector representation of features, is the vector representation of the corresponding text data.

3. The outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP as claimed in claim 1 is characterized in that: Treatment plan matching components include: The first matching degree calculation module is configured to obtain the number of disease characteristics, determine the weight of each disease characteristic according to its influence on the diagnosis and treatment plan, and calculate the matching degree with the diagnosis and treatment guidelines by defining a function to represent the matching degree with the diagnosis and treatment guidelines; multiply the matching degrees of all disease characteristics by their weights, and sum them to obtain the matching degree of the entire disease characteristic with the diagnosis and treatment guidelines; Indicates Symptoms Diagnosis and treatment guidelines The matching function between The second matching degree calculation module is configured to determine the weight of each combination of medical insurance policy and reimbursement path according to the economic impact of its reimbursement, define a function to represent its matching degree, and calculate the matching; multiply the matching degree of all medical insurance policy and reimbursement path combinations by their weights, and sum them to obtain the matching degree of the entire medical insurance policy and reimbursement path; The plan matching module is configured to multiply the matching degree of the disease characteristics with the diagnosis and treatment guidelines, the matching degree of the medical insurance policy and the reimbursement path, and then multiply it by the risk adjustment factor. The final result is the matching degree of the diagnosis and treatment plan.

4. The outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP as claimed in claim 3 is characterized in that: The first matching degree calculation module includes: The first feature weight calculation submodule is configured to identify all features of an individual's condition, classify and quantify them; obtain the condition features by collecting patient information, including symptoms, signs and laboratory test results; Evaluate the impact of each feature on the diagnosis and treatment results and assign a weight value to each feature; The first function definition submodule is configured to define a function for each disease feature to represent the degree of match between the disease feature and the diagnosis and treatment guideline; the function will quantify the consistency between the disease feature and the diagnosis and treatment guideline, and calculate the degree of match between each disease feature and the diagnosis and treatment guideline; Indicates Symptoms Diagnosis and treatment guidelines The matching function between The first matching degree weighted submodule is configured to multiply the matching degree of each disease feature by its corresponding weight value, and sum the weighted matching degrees of all features; and obtain the overall matching degree of the entire disease feature set with the diagnosis and treatment guidelines. The overall matching degree is a comprehensive indicator that reflects the overall degree of compliance between the patient's disease characteristics and the diagnosis and treatment guidelines.

5. The outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP as claimed in claim 4 is characterized in that: in, definition: represents the set of all the patient's disease characteristics, where m is the total number of features; Represents the set of weight values ​​corresponding to each disease feature; Represents the set of matching functions corresponding to each disease feature; Indicates diagnosis and treatment guidelines; Indicates the overall matching degree between the disease feature set and the diagnosis and treatment guidelines; Define the calculation formula for the matching degree between the disease characteristics and the diagnosis and treatment guidelines: In the formula, Indicates Symptoms Diagnosis and treatment guidelines The matching function between the disease characteristics and the diagnosis and treatment guidelines is quantified; specifically, further defined for: In the formula, Indicates treatment guidelines Symptoms and symptoms The similarity function between them, which quantifies the degree of similarity between them; It is The saturation parameter of each feature is used to control the steepness of the matching curve; and It is The linear transformation parameters of the features are used to adjust the output range of the similarity function; finally, the weight Indicates The influence of each feature on the diagnosis and treatment results is calculated by the following formula: In the formula, Indicates The original quantized value of each feature; , and It is The matching degree calculation formula comprehensively considers the similarity between the disease characteristics and the diagnosis and treatment guidelines, the feature weights and the quantitative values ​​of the features, and can more comprehensively evaluate the overall compliance of the patient's disease characteristics with the diagnosis and treatment guidelines.

6. The outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP as claimed in claim 3 is characterized in that: The second matching degree calculation module includes: The second feature weight calculation submodule is configured to obtain all medical insurance policies and their corresponding reimbursement paths, and to assign weights to each medical insurance policy and reimbursement path combination according to its impact on the patient's economic burden; the economic impact is reflected in the dimensions of reimbursement ratio, self-paid amount, and reimbursement scope; through comprehensive analysis, a weight value is assigned to each policy-path combination; The second function definition submodule is configured to define a matching function for each combination of medical insurance policy and reimbursement path to quantify the degree of fit between the combination and the patient's economic affordability and treatment needs; the input of the matching function includes the economic attributes of the policy, the patient's economic conditions, and the feasibility of the treatment path; the matching function can output a value indicating the specific matching degree of the policy-path combination; The second matching degree weighted submodule is configured to combine the matching degree with its corresponding weight value after completing the matching degree calculation of each combination of medical insurance policy and reimbursement path, multiply the matching degree of each combination by its weight value, and sum up the weighted matching degrees of all combinations to comprehensively reflect the economic impact and matching degree of all medical insurance policies and reimbursement paths, and finally obtain an overall matching degree value.

7. The outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP as claimed in claim 4 is characterized in that: Solution matching module, including: The matching degree calculation submodule is configured to multiply the matching degree between each disease characteristic and the diagnosis and treatment guideline and the economic impact of all medical insurance policies and reimbursement pathways with the matching degree to obtain a preliminary comprehensive matching degree, which reflects the comprehensive consideration of disease characteristics and economic factors; an adjustment factor definition submodule configured to employ a risk adjustment factor adjusted according to a risk assessment of the patient; The scheme matching degree calculation submodule is configured to multiply the preliminary comprehensive matching degree by the risk adjustment factor to obtain the matching degree of the diagnosis and treatment scheme.

8. The outpatient medical insurance chronic disease medical data processing system based on artificial intelligence NLP as claimed in claim 1 is characterized in that: Analysis and early warning components, including: The data collection module is configured to collect various medical records, physical data and medication status of the patient, including the patient's medical history, blood pressure, heart rate, blood sugar physiological indicators, and all medications currently in use; The data recognition module is configured to conduct longitudinal analysis of various medical records, vital sign data and medication status, identify patterns and trends through statistics, and find out the changing patterns of patients' health status; The suggestion output module is configured to give personalized health management suggestions based on the results of the analysis, including adjusting medication plans, changing lifestyle habits and exercising.

Images