Multi-center intelligent specialized hospital cooperation platform

Abstract

The invention relates to a multi-center intelligent specialized hospital cooperation platform. The platform comprises a big data management and information platform, an operation management platform and a hierarchical diagnosis and treatment platform. The big data management and information platform performs acquisition, convergence, quality detection, standardization and structured processing on heterogeneous data of multiple hospital areas, and is also used for reserving a standardized data interface and an index structure for the special disease scientific research database; the operation management platform performs index calculation and decision visualization based on the standardized data; the hierarchical diagnosis and treatment platform establishes multi-hospital-area patient index association and hierarchical diagnosis and treatment collaboration. Wherein the big data governance and information platform comprises a data platform courtyard end, a data acquisition and convergence platform, a big data quality detection system, a big data standardization system and a big data post-structuring system. According to the invention, intercommunication and sharing, standardized processing and intelligent application of medical data of multiple hospital areas are realized, and the intercommunication and sharing performance, data uniformity and medical resource configuration efficiency of the medical data of multiple hospital areas are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical informatization, and particularly to a multi-center intelligent specialized hospital collaboration platform. Background Art

[0002] With the expansion of the scale of medical institutions and the rise of the hospital group management mode, the medical institution mode of multiple campuses and multiple centers has become a trend. However, at present, there is a common phenomenon of information islands in medical information systems, resulting in difficult data interconnection and sharing between different campuses, which hinders the optimization of medical resources allocation and the improvement of collaborative service capabilities.

[0003] Most current hospital information systems are designed for a single campus. The systems of each campus are fragmented from each other and the data standards are inconsistent, making it difficult to effectively support cross-campus business collaboration and management decision-making. At the same time, the existing systems have limited processing capabilities for medical unstructured data and cannot fully explore the value of medical data. In addition, there is a lack of unified patient identification and hierarchical diagnosis and treatment mechanisms between different campuses, which affects the efficiency of medical resource allocation and the patient's medical experience.

[0004] Therefore, there is an urgent need in this field for a hospital management platform that can achieve multi-center intelligent collaboration. Summary of the Invention

[0005] The present invention provides a multi-center intelligent specialized hospital collaboration platform to solve the defects of the prior art.

[0006] The present invention provides a multi-center intelligent specialized hospital collaboration platform, including:

[0007] A big data governance and information platform, which is used to collect, converge, perform quality inspection, standardize and structure the heterogeneous data of multiple campuses, and is also used to reserve standardized data interfaces and index structures for the specialized disease research database;

[0008] An operation management platform, which is connected to the big data governance and information platform and is used to perform index calculation and decision visualization based on the standardized data;

[0009] A hierarchical diagnosis and treatment platform, which is connected to the big data governance and information platform and the operation management platform and is used to establish multi-campus patient index association and hierarchical diagnosis and treatment collaboration.

[0010] According to the multi-center intelligent specialized hospital collaboration platform provided by the present invention, the big data governance and information platform includes:

[0011] The hospital campus end of the data platform is used to execute the ETL process based on the J2EE framework. The hospital campus end of the data platform has a data resource management module, a metadata management module, a data service module, a transformation task module, a job process module, a scheduling plan module, and an execution monitoring module. Among them, the data resource management module supports the connection configuration of various relational databases, flat files, and special data sources. The metadata management module supports table structure import and lineage analysis. The data service module provides WebService interface publishing and receiving functions. The transformation task module supports functions such as variable setting, data extraction, SQL execution, stored procedure call, and external program execution. The job process module supports visual workflow design. The scheduling plan module supports automatic scheduling based on time and events;

[0012] The data collection and aggregation system is used to centrally aggregate the data of heterogeneous networks and heterogeneous data sources in each hospital campus to the central data platform. The data collection and aggregation system has a collection system status monitoring module, an interface management module, a data dictionary management module, a hospital file management module, a data period management module, and a data monitoring module. Among them, the collection system status monitoring module tracks the connection status of the data collection systems in each hospital campus. The interface management module monitors the call situation of the open interfaces of the platform to the outside world. The data dictionary management module maintains the field definitions of the data reported by the hospital campus. The hospital file management module records the file information of the data accessed by each hospital. The data period management module performs partition management on the data according to the time period;

[0013] The big data quality detection system is used to perform quality control detection on the aggregated data. The big data quality detection system is configured with a quality control rule management module, a quality control task management module, a quality control result management module, a quality control plan management module, and a plan monitoring module. Among them, the quality control rule management module supports multiple quality control rules such as data relevance, integrity, uniqueness, normativity, rationality, consistency, and continuity. The quality control task management module configures the execution parameters and scheduling plan of the quality control task. The quality control result management module records and displays the execution results of the quality control task. The quality control plan management module combines multiple quality control tasks into a plan and sets reference standards;

[0014] Big data standardization system, which is used to perform standardization processing on the data that has passed quality inspection. The big data standardization system includes a metadata management module, a data lineage management module, a data security permission management module, a master data management module, and a standardization model application module. Among them, the metadata management module uniformly manages the metadata of all data sources, the data lineage management module tracks the dependency relationships in the data processing process, the data security permission management module controls data access based on roles, and the master data management module maintains the dictionary and mapping relationships of key master data;

[0015] Big data post-structuring system, which is used to perform structuring processing on unstructured medical data. The big data post-structuring system is configured with a data development component management module, a task scheduling management module, an offline development management module, and a structuring model application module. Among them, the data development component management module provides a preprocessing model for information extraction, the task scheduling management module configures the execution plan and dependency relationships of structuring tasks, the offline development management module supports task types such as remote calls, Hive SQL, preprocessing, and data extraction, and the structuring model application module performs structuring processing on medical texts;

[0016] Special disease data indexing module, which is used to create special disease-related indexes for standardized data. The special disease data indexing module includes a form indexing unit, a field mapping unit, and a data association unit. Among them, the form indexing unit establishes an index structure for special disease-related forms, the field mapping unit provides the mapping definition between standardized fields and special disease research fields, and the data association unit provides the association paths between multi-table data.

[0017] A multi-center intelligent special hospital collaborative platform provided by the present invention. The conversion task module at the hospital area end of the data platform includes:

[0018] Variable setting task component, which is used to call the SELECT statement in the data connection to assign values to variables, and the variables can be referenced in subsequent conversion tasks;

[0019] Data extraction task component, which is used to complete data collection and conversion between two data sources or within the same data source. The data extraction task component includes a basic information setting unit, a source table query unit, a target table update unit, and a field mapping unit. Among them, the source table query unit supports multiple selection of source tables and writing SELECT query statements, the target table update unit supports multiple target table update methods including not deleting, emptying the target table, deleting existing data, deleting this batch of data, executing custom statements, inserting new data, and inserting and updating data, and the field mapping unit sets the target table fields corresponding to the source table fields;

[0020] An SQL execution task component, which is used to directly execute a batch of SQL statements or SQL code blocks. The SQL execution task component supports DDL statements and DML statements, as well as SQL code blocks including variable declarations, judgments, loops, and exception handling;

[0021] A stored procedure task component, which is used to call a database stored procedure to complete data processing. The stored procedure task component supports configuring input parameters, output parameters, and return value variables of the stored procedure;

[0022] An external command task component, which is used to execute programs and commands under the operating system command line. The external command task component supports configuring execution scripts, successful return values, and return value variable assignments;

[0023] A data profiling task component, which is used to perform data quality exploration on a target table or SQL statement. The data profiling task component supports profiling types such as numeric value analysis, value matching check, character value analysis, date value analysis, boolean value analysis, duplicate value check, expression matching, referential integrity check, and value distribution analysis;

[0024] A specialized disease data conversion component, which is used to convert standardized data according to the needs of specialized disease research. The specialized disease data conversion component supports specialized disease database model mapping, multi-level data aggregation, and fine-grained data extraction.

[0025] A multi-center intelligent specialized hospital collaborative platform provided by the present invention. The rule types of the quality control rule management module of the big data quality detection system include:

[0026] A data correlation rule, which is used to check the correlation relationship between cross-table fields;

[0027] A data integrity rule, which is used to check whether the field value is empty;

[0028] A data uniqueness rule, which is used to check whether the field value is repeated;

[0029] A data normalization rule, which is used to check whether the field value conforms to a predetermined format. The data normalization rule includes a date format check sub-rule, a numeric format check sub-rule, a text format check sub-rule, and a value range dictionary check sub-rule;

[0030] A data rationality rule, which is used to check whether the field value conforms to medical logic;

[0031] Data consistency rules, which are used to check the consistency between fields within a table or between fields of different tables;

[0032] Data continuity rules, which are used to check the continuity of time series data.

[0033] According to a multi-center intelligent specialized hospital collaboration platform provided by the present invention, the preprocessing models of the data opening component management module of the big data post-structuring system include:

[0034] Date format conversion model, which is used to convert date strings in various formats into a standard date format;

[0035] ID card digit conversion model, which is used to unify the digit format of ID card numbers;

[0036] Electronic case document extraction model, which is used to extract specific document content from electronic medical records. The electronic case document extraction model supports configuring start characters, end characters, and ignored character parameters;

[0037] Electronic case chapter extraction model, which is used to extract specified chapter content from electronic medical records. The electronic case chapter extraction model supports configuring expression, start characters, and end characters parameters;

[0038] Key cleaning model, which is used to clean keyword fields in data;

[0039] Chinese character removal model, which is used to remove Chinese characters from text;

[0040] ID card information extraction model, which is used to extract date of birth and gender information from ID card numbers;

[0041] Full-width to half-width model, which is used to convert full-width characters into half-width characters;

[0042] Sensitive word replacement model, which is used to desensitize key patient information;

[0043] Encryption and decryption model group, which includes CBC encryption model, CBC decryption model, MD5 encryption model, BASE64 encryption model, and BASE64 decryption model, and is used to encrypt and decrypt key patient information.

[0044] According to a multi-center intelligent specialized hospital collaboration platform provided by the present invention, the operation management platform includes:

[0045] Operation Management Decision System, which is used to conduct multi - angle index analysis on the standardized data. The Operation Management Decision System has a multi - level permission management module, a multi - dimensional screening module, an index card display module, a chart analysis module, a ranking analysis module, a detailed data query module, and a department and doctor performance module. Among them, the multi - level permission management module sets hierarchical permissions based on the roles of the general hospital president, branch hospital presidents, and department directors. The multi - dimensional screening module supports flexible data screening according to the dimensions of hospital area, department, and time. The index card display module intuitively displays the key index values and year - on - year and month - on - month growth conditions. The chart analysis module displays the index trends and distributions through visualizations such as line charts and bar charts. The ranking analysis module supports index ranking comparisons among hospital areas, departments, and doctors. The detailed data query module provides the function of querying and exporting detailed data related to the index. The department and doctor performance module automatically calculates the department and doctor performance indicators based on business data;

[0046] President Decision System, which is used to display the hospital operation status in the form of a cockpit. The President Decision System includes an overall overview module, an operation management module, a medical quality module, a medication analysis module, a medical expense module, an information data module, a cancer type screening module, and a full - screen display module. Among them, the overall overview module displays the key indicators of the overall hospital operation status. The operation management module focuses on the analysis of hospital operation - related indicators. The medical quality module displays the indicators related to medical quality control. The medication analysis module provides medication use and expense analysis. The medical expense module analyzes the composition and change trends of medical expenses. The information data module displays the indicators related to informatization construction and data quality. The cancer type screening module supports data screening by cancer type for specialized analysis. The full - screen display module supports the full - screen switching function in the conference display mode.

[0047] According to a multi - center intelligent specialized hospital collaboration platform provided by the present invention, the permission design of the multi - level permission management module of the Operation Management Decision System includes:

[0048] General Hospital President Permission, which supports viewing the index data and ranking of all hospital areas, all departments in any hospital area, and specific departments in any hospital area;

[0049] Branch Hospital President Permission, which supports viewing the index data and ranking of the affiliated hospital area, all departments in the affiliated hospital area, and specific departments in the affiliated hospital area;

[0050] Department Director Permission, which supports viewing the index data of the configured departments in the affiliated hospital area and the rankings of the doctors in the affiliated department.

[0051] A multi - center intelligent specialized hospital collaborative platform provided by the present invention, the hospital director decision - making system includes:

[0052] An overall overview theme module, which is used to display the core index data of the whole hospital, and the overall overview theme module supports screening by hospital area and cancer type;

[0053] An operation management theme module, which is used to display the indicators related to hospital operation, including outpatient volume data, inpatient volume data, operation volume data, income data and expenditure data;

[0054] A medical quality theme module, which is used to display the indicators related to medical quality, including average length of stay data, surgical complication rate data, hospital infection rate data and antibacterial drug use rate data;

[0055] A medication analysis theme module, which is used to display the analysis of drug use conditions, including drug cost composition data, drug proportion data, antibiotic use data and large - scale equipment use data;

[0056] A medical expense theme module, which is used to display the composition and changes of medical expenses, including outpatient average cost data, inpatient average cost data, medical insurance expense data and out - of - pocket expense data;

[0057] An information data theme module, which is used to display the hospital informatization construction and data quality conditions, including system online rate, data integrity rate, data accuracy rate and system utilization rate data.

[0058] A multi - center intelligent specialized hospital collaborative platform provided by the present invention, the hierarchical diagnosis and treatment platform includes:

[0059] A patient master index system, which is used to establish a unique identifier for patients within a multi - hospital - area medical system. The patient master index system is configured with a patient query module, a patient 360 - view module, a suspected patient management module, an operation history record module and a weight configuration module. Among them, the patient query module supports querying patient information in multiple ways through the master index number, name and ID number. The patient 360 - view module displays the complete diagnosis and treatment information of the patient in different hospitals. The suspected patient management module automatically identifies different patient records that may be the same person based on a weight algorithm and supports manual review and merging. The operation history record module records all index operations to ensure data traceability. The weight configuration module supports configuring the weight parameters of the patient matching algorithm;

[0060] Remote consultation module, which is used to support cross-hospital expert remote consultation based on the information associated with the patient master index. The remote consultation module includes a consultation application sub-module, an expert scheduling sub-module, a consultation arrangement sub-module, a consultation execution sub-module, and a consultation record sub-module. Among them, the consultation application sub-module supports the initiating hospital to fill in patient information and consultation requirements. The expert scheduling sub-module manages the outpatient schedules of experts in each hospital area. The consultation arrangement sub-module coordinates the consultation time and participants. The consultation execution sub-module supports real-time audio and video interaction and medical image sharing. The consultation record sub-module records the consultation process and conclusions;

[0061] Two-way referral module, which is used to support the orderly referral of patients between different-level medical institutions. The two-way referral module has a referral application sub-module, a referral review sub-module, a referral reservation sub-module, a referral record sub-module, and a follow-up management sub-module. Among them, the referral application sub-module supports the initiation of upward and downward referral applications and the filling in of basic information. The referral review sub-module reviews the referral application and assigns the receiving department and doctor. The referral reservation sub-module arranges the appointment time for the patient to visit the transferred hospital. The referral record sub-module records the whole process information of the patient's referral. The follow-up management sub-module supports the follow-up plan and implementation of the referred patients;

[0062] Remote medical module, which is used to provide cross-hospital remote medical services. The remote medical module is configured with a remote outpatient sub-module, a remote ward round sub-module, a remote pathology sub-module, and a remote imaging sub-module. Among them, the remote outpatient sub-module supports remote outpatient services via video. The remote ward round sub-module supports senior hospital experts to conduct remote ward rounds on inpatients in lower-level hospitals. The remote pathology sub-module supports the remote diagnosis of pathological sections. The remote imaging sub-module supports the remote diagnosis of medical images.

[0063] A multi-center intelligent specialized hospital collaboration platform provided by the present invention. The weight configuration module of the patient master index system includes the following functional units:

[0064] Field weight configuration unit, which is used to set the weight values of each field in the patient matching process. The configurable fields include name, gender, date of birth, ID number, address, and contact phone number;

[0065] Weight range configuration unit, which is used to set the weight range for the recommendation of suspected patients. Patient records with a weight calculation score within this range are recommended to the list of suspected patients;

[0066] Matching algorithm configuration unit, which is used to set the matching algorithms between fields, including exact matching algorithm, fuzzy matching algorithm, edit distance algorithm, and pronunciation similarity algorithm;

[0067] A weight testing unit, which is used to test the matching effect of the current weight configuration and generate a test report;

[0068] A weight optimization unit, which is used to automatically adjust weight parameters according to historical matching data and optimize the matching accuracy.

[0069] A multi-center intelligent special hospital collaborative platform provided by the present invention, based on a multi-level data governance architecture, realizes the unified collection, standardization, and structured processing of multi-source heterogeneous medical data. Through the collaborative work of the hospital area end of the data platform and the data collection and aggregation system, the information islands between hospitals are broken; through 19 quality control rules of the big data quality detection system, the data quality is significantly improved; through the big data standardization system and the big data post-structuring system, non-standardized and unstructured medical data are converted into computable and analyzable standard data assets.

[0070] A multi-center intelligent special hospital collaborative platform provided by the present invention also provides a reserved mechanism for a special disease research interface layer as a standardized data interface and an index structure, solving the connection problem between medical data and scientific research applications. This interface layer enables the platform to flexibly expand and support the functions of a special disease scientific research database without affecting the original architecture, improving the scalability and adaptability of the platform. Through the standardized interface, the special disease database can obtain high-quality medical data after governance, while reducing the cost and complexity of secondary development. For example, in the oncology special disease database, through the index structure, the patient group of a specific tumor type can be quickly located without traversing the entire database, significantly reducing the time complexity of data retrieval.

[0071] A multi-center intelligent special hospital collaborative platform provided by the present invention provides accurate decision-making support for managers at different levels based on the standardized medical data. The operation management decision-making system meets the management needs of different roles through multi-level permission design and multi-dimensional index analysis; intuitively displays the operation status of the hospital, enabling staff at multiple levels to quickly master the key information of hospital operation.

[0072] In addition, the hierarchical diagnosis and treatment platform of the present invention establishes a unique identifier for patients among multiple hospitals based on the patient master index system, solving the problem of fragmented patient information. Through remote consultation, two-way referral, and telemedicine modules, the cross-hospital collaborative of medical resources is realized, promoting the sinking of high-quality medical resources and the hierarchical diagnosis and treatment of patients.

[0073] A multi - center intelligent specialized hospital collaborative platform provided by the present invention adopts a closed - loop design. The data flow between each module is smooth, forming a complete chain of data collection, governance, application, and feedback. After the system is deployed, it can significantly improve the accuracy rate and standardization rate of hospital data, greatly enhance the management decision - making efficiency, optimize the cross - campus resource collaboration, improve the patient's medical experience, enhance the medical quality, and effectively control the medical cost. Generally speaking, through the integrated collaboration of the three major platforms of the present invention, it effectively solves the problems of difficult inter - hospital medical data sharing, inconsistent data standards, and low efficiency of medical resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0075] Figure 1 It is a schematic structural diagram of a multi - center intelligent specialized hospital collaborative platform provided by an embodiment of the present invention;

[0076] Figure 2 It is a schematic structural diagram of a conversion task module provided by an embodiment of the present invention;

[0077] Figure 3 It is a schematic diagram of the implementation of the dean's decision - making system provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0078] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention with reference to the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. They should not be construed as limitations on the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention. In the description of the present invention, it should be understood that the terms used are only for the purpose of description and cannot be construed as indicating or implying relative importance.

[0079] The following describes the embodiments of the present invention with reference to the drawings.

[0080] As Figure 1 shown, the present invention provides a multi - center intelligent specialized hospital collaborative platform, including:

[0081] Big Data Governance and Information Platform, which is used to collect, converge, perform quality inspection, standardize and structure multi-campus heterogeneous data, and is also used to reserve standardized data interfaces and index structures for disease-specific research databases.

[0082] Specifically, the Big Data Governance and Information Platform is the basic layer of the entire system, mainly responsible for collecting, converging, performing quality inspection, standardizing and structuring multi-campus heterogeneous data. The Big Data Governance and Information Platform first performs ETL (Extract, Transform, Load) processing on the original data of each campus through the campus end of the data platform.

[0083] Among them, the Big Data Governance and Information Platform includes:

[0084] The campus end of the data platform, which is used to execute the ETL process based on the J2EE framework. The campus end of the data platform has a data resource management module, a metadata management module, a data service module, a transformation task module, a job flow module, a scheduling plan module and an execution monitoring module. Among them, the data resource management module supports the connection configuration of various relational databases, flat files and special data sources. The metadata management module supports table structure import and lineage analysis. The data service module provides WebService interface publishing and receiving functions. The transformation task module supports functions such as variable setting, data extraction, SQL execution, stored procedure call and external program execution. The job flow module supports visual workflow design. The scheduling plan module supports automatic scheduling based on time and events.

[0085] As Figure 2 shown, the transformation task module of the campus end of the data platform includes:

[0086] A variable setting task component, which is used to call the SELECT statement in the data connection to assign values to variables, and the variables can be referenced in subsequent transformation tasks;

[0087] A data extraction task component, which is used to complete data collection and transformation between two data sources or within the same data source. The data extraction task component includes a basic information setting unit, a source table query unit, a target table update unit and a field mapping unit. Among them, the source table query unit supports multiple selection of source tables and writing SELECT query statements. The target table update unit supports multiple target table update methods, including not deleting, emptying the target table, deleting existing data, deleting this batch of data, executing custom statements, inserting new data, and inserting and updating data. The field mapping unit corresponds the target table fields with the source table fields for setting.

[0088] An SQL execution task component, which is used to directly execute a batch of SQL statements or SQL code blocks. The SQL execution task component supports DDL statements and DML statements, as well as SQL code blocks including variable declaration, judgment, loop and exception handling;

[0089] A stored procedure task component, which is used to call a database stored procedure to complete data processing, and which supports configuring input parameters, output parameters, and return value variables of a stored procedure;

[0090] An external command task component, which is used to execute programs and commands under the operating system command line, and supports configuration of execution scripts, successful return values, and return value variable assignment;

[0091] A data analysis task component, which is used to perform data quality exploration on a target table or SQL statement, and supports analysis types including numeric value analysis, value matching check, character value analysis, date value analysis, Boolean value analysis, duplicate value check, expression matching, reference integrity check, and value distribution analysis;

[0092] A disease-specific data conversion component is used to convert the data structure of standardized data according to the needs of disease-specific research. The disease-specific data conversion component supports disease-specific library model mapping, multi-level data aggregation and fine-grained data extraction.

[0093] Specifically, the campus end of the data platform is developed based on the J2EE framework and uses the JDBC connection mode to access data. It supports multiple database types such as Oracle, MS SQL Server, MySQL, as well as file formats such as CSV, XLS, and special data sources such as SAP and OLAP. In the data resource management module, the system will create a connection configuration for each data source. For example, for the Oracle database, you need to configure the host IP address, port (default 1521), service name (SID), user name and password to form a complete JDBC connection string. The metadata management module will import the table structure (excluding the original data) in the data source for subsequent conversion tasks. It also supports the lineage analysis function, which displays the upstream and downstream relationships between tables in a visual way to facilitate tracking of data sources.

[0094] The disease-specific data conversion component is a special component designed for the specific needs of disease-specific research. It is responsible for further converting standardized medical data into a data structure suitable for disease-specific research. The component supports three key functions: disease-specific library model mapping, multi-level data aggregation, and fine-grained data extraction. The disease-specific library model mapping function can map the standardized data model to the data model required for specific disease-specific research, such as mapping the general diagnostic data structure to the TNM staging structure of tumor disease-specific research; the multi-level data aggregation function supports data aggregation at different granularities, such as aggregating multiple examination results of patients into disease progression trends; the fine-grained data extraction function can accurately extract detailed data of concern to disease-specific research from complex data, such as extracting specific molecular marker information from pathology report text.

[0095] Through these functions, the disease-specific data conversion component enables the platform to meet the special requirements of different disease-specific studies on data structure and content while maintaining a standardized data architecture, providing customized data processing capabilities for disease-specific scientific research, and greatly improving the efficiency and accuracy of medical data utilization in disease-specific research.

[0096] The data collection and aggregation system is used to aggregate the data of heterogeneous networks and heterogeneous data sources of each hospital area to a central data platform. The data collection and aggregation system has a collection system status monitoring module, an interface management module, a data dictionary management module, a hospital archive management module, a data period management module and a data monitoring module. The collection system status monitoring module tracks the connection status of the data collection system of each hospital area, the interface management module monitors the call status of the platform's open interfaces, the data dictionary management module maintains the field definition of the data reported by the hospital area, the hospital archive management module records the archive information of the data accessed by each hospital, and the data period management module performs partition management of the data according to the time period.

[0097] Specifically, the data collection and aggregation system is responsible for aggregating the data of each hospital area to the central platform, which monitors the connection status of the data collection systems of each hospital area in real time. When a hospital area fails to report data for more than 1 hour, the system will display an abnormal status and send an alarm; the interface management module monitors the call of the platform's open interfaces, and records information such as interface access frequency, response time and error rate; the data dictionary management module maintains meta-information such as the form code, field definition, data type and constraints of the data reported by the hospital area to ensure the accuracy of data analysis; the hospital archive management module records information such as the source system, source manufacturer, collectability, real-time and start time of the data accessed by each hospital, providing a basis for data collection strategies.

[0098] Big data quality detection system, which is used to perform quality control detection on the aggregated data. The big data quality detection system is configured with a quality control rule management module, a quality control task management module, a quality control result management module, a quality control plan management module, and a plan monitoring module. Among them, the quality control rule management module supports various quality control rules for data relevance, integrity, uniqueness, normativity, rationality, consistency, and continuity. The quality control task management module configures the execution parameters and scheduling plan of the quality control task. The quality control result management module records and displays the execution results of the quality control task. The quality control plan management module combines multiple quality control tasks into a plan and sets reference standards.

[0099] Among them, the rule types of the quality control rule management module of the big data quality detection system include:

[0100] Data relevance rule, which is used to check the association relationship between cross-table fields;

[0101] Data integrity rule, which is used to check whether the field value is empty;

[0102] Data uniqueness rule, which is used to check whether the field value is repeated;

[0103] Data normativity rule, which is used to check whether the field value conforms to the predetermined format. The data normativity rule includes a date format check sub-rule, a numerical format check sub-rule, a text format check sub-rule, and a value range dictionary check sub-rule;

[0104] Data rationality rule, which is used to check whether the field value conforms to medical logic;

[0105] Data consistency rule, which is used to check the consistency between fields within a table or between tables;

[0106] Data continuity rule, which is used to check the continuity of time series data.

[0107] Specifically, the big data quality inspection system conducts a comprehensive quality inspection on the aggregated data to ensure that the data meets the requirements of standardization and structuring. The system supports 19 quality control rules, covering multiple dimensions such as data relevance, integrity, uniqueness, standardization, rationality, consistency, and continuity. Taking the data relevance rule as an example, this rule is used to check the association relationship between cross-table fields. For example, it checks whether the patient ID in the doctor's order record table of a certain case exists in the patient basic information table. The data standardization rules include sub-rules such as date format check, numerical format check, text format check, and value range dictionary check. The date format check verifies whether the date field conforms to the specified format (such as YYYY-MM-DD). The numerical format check verifies whether the numerical field is within the valid range and has the correct precision. The value range dictionary check verifies whether the field value belongs to the predefined value range (for example, the gender field must be "male" or "female").

[0108] The big data standardization system is used to perform standardization processing on the data that has passed the quality inspection. The big data standardization system includes a metadata management module, a data lineage management module, a data security permission management module, a master data management module, and a standardization model application module. Among them, the metadata management module uniformly manages the metadata of all data sources. The data lineage management module traces the dependency relationships in the data processing process. The data security permission management module controls data access based on roles. The master data management module maintains the dictionaries and mapping relationships of key master data.

[0109] Specifically, the big data standardization system is responsible for standardizing the data that has passed the quality inspection and solving the problem of inconsistent data standards in multiple hospital areas. First, the system uniformly manages the metadata of all data sources through the metadata management module, including metadata at the database, table, and field levels. The data lineage management module traces the dependency relationships in the data processing process to form a complete data lineage graph. The data security permission management module, based on the role-based access control mechanism, ensures data access security. The master data management module maintains the dictionaries and mapping relationships of key master data, such as diagnostic standard codes, drug codes, and surgical operation classifications, etc. The standardization model application module converts the original data into a unified standard through predefined mapping rules. For example, it maps different drug coding systems (such as medical insurance codes and hospital-customized codes) used in different hospital areas to the national drug standard coding system.

[0110] Post-structured big data system, which is used for structuring unstructured medical data. The post-structured big data system is configured with a data development component management module, a task scheduling management module, an offline development management module, and a structured model application module. Among them, the data development component management module provides a preprocessing model for information extraction, the task scheduling management module configures the execution plan and dependency relationship of structured tasks, the offline development management module supports task types such as remote call, Hive SQL, preprocessing, and data extraction, and the structured model application module structures the medical text.

[0111] The preprocessing models of the data development component management module of the post-structured big data system include:

[0112] Date format conversion model, which is used to convert date strings in various formats into a standard date format;

[0113] ID card digit conversion model, which is used to unify the digit format of ID card numbers;

[0114] Electronic medical record document extraction model, which is used to extract specific document content from electronic medical records. The electronic medical record document extraction model supports configuring start character, end character, and ignore character parameters;

[0115] Electronic medical record chapter extraction model, which is used to extract specified chapter content from electronic medical records. The electronic medical record chapter extraction model supports configuring expression, start character, and end character parameters;

[0116] Key cleaning model, which is used to clean keyword fields in data;

[0117] Chinese character removal model, which is used to remove Chinese characters from text;

[0118] ID card information extraction model, which is used to extract date of birth and gender information from ID card numbers;

[0119] Full-width to half-width model, which is used to convert full-width characters into half-width characters;

[0120] Sensitive word replacement model, which is used to desensitize patients' key information;

[0121] Encryption and decryption model group, which includes CBC encryption model, CBC decryption model, MD5 encryption model, BASE64 encryption model, and BASE64 decryption model, and is used to encrypt and decrypt patients' key information.

[0122] Specifically, the big data post-structuring system targets the unstructured data of hospitals (such as medical record texts, inspection reports, etc.) and converts it into structured data through natural language processing and machine learning technologies. The data opening component management module of the system provides a variety of preprocessing models, including date format conversion, electronic medical record document extraction, electronic medical record chapter extraction, ID number information extraction, sensitive word replacement, encryption and decryption, etc.; the electronic medical record document extraction model is used to extract specific document content from electronic medical records, supporting the configuration of start character, end character, and ignored character parameters. For example, it can extract all the content between "admission record" and "discharge summary". The electronic medical record chapter extraction model is more refined and can extract the content of specific chapters from electronic medical records, such as chapters like "past history", "present illness history", or "physical examination". The ID number information extraction model can extract information such as date of birth and gender from 18-digit ID numbers. For example, from "110101199001011234", the date of birth is extracted as "1990-01-01" and the gender is "male" (odd numbers represent male). The sensitive word replacement model is used to desensitize the key information of patients. For example, the patient's name is replaced with "Zhang**", and the mobile phone number is replaced with "138****1234".

[0123] The specialized disease data indexing module is used to create specialized disease-related indexes for standardized data. The specialized disease data indexing module includes a form indexing unit, a field mapping unit, and a data association unit. Among them, the form indexing unit establishes an index structure for specialized disease-related forms, the field mapping unit provides the mapping definition between standardized fields and specialized disease research fields, and the data association unit provides the association path between multi-table data.

[0124] Specifically, the specialized disease data indexing module is specifically used to create an index structure for specialized disease research for standardized medical data. By establishing an efficient data indexing and association mechanism, the module solves the problems of low retrieval efficiency and complex data association in traditional medical data systems, and significantly improves the performance and convenience of accessing specialized disease research data.

[0125] In practical applications, the value of the specialized disease data indexing module can be more intuitively demonstrated through specific cases. For example, in a multi-center breast cancer research project, researchers need to quickly screen out "patients who received radical mastectomy for breast cancer within the past three years and received endocrine therapy for more than 12 months after the operation". In the traditional system, this requires separately querying the operation record table and the medication record table, and then performing complex data association and filtering. The whole process is time-consuming and error-prone.

[0126] In a system equipped with a disease-specific data indexing module, the form indexing unit has already established indexes for the surgical record form and the medication record form related to breast cancer; the field mapping unit maps the standardized "surgical code" to the "breast cancer surgical procedure" in the breast specialty, and maps the "drug code" to the "type of endocrine therapy drug"; the data association unit defines the association path between the surgical record and the medication record through the patient ID and the order of visits.

[0127] Researchers only need to set retrieval conditions using professional terms: "breast cancer surgical procedure = radical mastectomy AND type of endocrine therapy drug = tamoxifen AND duration of medication > 12 months AND surgical date >= within 3 years". The system will automatically locate the relevant forms through the index, convert them to standardized fields for query according to the field mapping, and then complete cross-table data integration based on the association path, and finally return the patient cohort that meets the conditions within seconds, greatly improving the data retrieval efficiency and research efficiency.

[0128] An operation management platform, which is connected to the big data governance and information platform and is used for index calculation and decision visualization based on the data after standardized processing.

[0129] Among them, the operation management platform includes:

[0130] An operation management decision-making system, which is used for multi-angle index analysis of the data after standardized processing. The operation management decision-making system has a multi-level permission management module, a multi-dimensional screening module, an index card display module, a chart analysis module, a ranking analysis module, a detailed data query module, and a department and doctor performance module. Among them, the multi-level permission management module sets hierarchical permissions based on the roles of the general hospital president, branch hospital president, and department director. The multi-dimensional screening module supports flexible data screening according to the dimensions of hospital area, department, and time. The index card display module intuitively displays the key index values and year-on-year and month-on-month growth conditions. The chart analysis module displays the index trends and distributions through the visualization methods of line charts and bar charts. The ranking analysis module supports index ranking comparisons among hospital areas, departments, and doctors. The detailed data query module provides the function of querying and exporting detailed data related to the index. The department and doctor performance module automatically calculates the department and doctor performance indexes based on business data.

[0131] Among them, the permission design of the multi-level permission management module of the operation management decision-making system includes:

[0132] The general hospital president permission, which supports viewing the index data and ranking of all hospital areas, all departments in any hospital area, and specific departments in any hospital area;

[0133] The authority of the branch hospital director, which supports viewing the data and ranking of various indicators of the affiliated hospital area, all departments in the affiliated hospital area, and specific departments in the affiliated hospital area;

[0134] The authority of the department director, which supports viewing the data of various indicators of the configured departments in the affiliated hospital area and the ranking of doctors in the affiliated department.

[0135] Specifically, the operation management platform is the decision-making layer of the system. Based on the standardized data processed by the big data governance and information platform, it provides decision-making support for hospital managers. The operation management decision-making system of the operation management platform adopts a multi-level permission design, and different data access permissions are set according to user roles (general hospital director, branch hospital director, department director). The general hospital director's authority can view the data and ranking of various indicators of all hospital areas, all departments in any hospital area, and specific departments in any hospital area; the branch hospital director's authority can only view the data and ranking of indicators of the affiliated hospital area, all departments in the affiliated hospital area, and specific departments in the affiliated hospital area; the department director's authority is limited to viewing the data of the configured departments in the affiliated hospital area and the ranking of doctors in the affiliated department. The multi-dimensional screening module supports flexible screening of data according to dimensions such as hospital area, department, and time. For example, it can be selected to view the operation data of the "Oncology Department" in the "First Quarter of 2023" in the "Cancer Hospital" area. The indicator card display module intuitively displays the key indicator values and the year-on-year and month-on-month growth conditions.

[0136] The hospital director decision-making system, which is used to display the hospital operation status in the form of a cockpit. The hospital director decision-making system includes an overall overview module, an operation management module, a medical quality module, a medication analysis module, a medical expense module, an information data module, a cancer type screening module, and a full-screen display module. Among them, the overall overview module displays the key indicators of the overall hospital operation status, the operation management module focuses on the analysis of indicators related to hospital operation, the medical quality module displays the indicators related to medical quality control, the medication analysis module provides the analysis of drug use and expenses, the medical expense module analyzes the composition and change trend of medical expenses, the information data module displays the indicators related to informatization construction and data quality, the cancer type screening module supports screening data by cancer type for specialized analysis, and the full-screen display module supports the full-screen switching function in the meeting display mode.

[0137] Among them, the hospital director decision-making system includes:

[0138] The overall overview theme module, which is used to display the core indicator data of the whole hospital, and the overall overview theme module supports screening by hospital area and cancer type;

[0139] The operation management theme module is used to display hospital operation - related indicators, including outpatient volume data, inpatient volume data, operation volume data, revenue data, and expenditure data;

[0140] The medical quality theme module is used to display medical quality - related indicators, including average length of stay data, surgical complication rate data, hospital infection rate data, and antibacterial drug utilization rate data;

[0141] The medication analysis theme module is used to display the analysis of drug use, including drug cost composition data, drug - to - total - cost ratio data, antibiotic use data, and large - equipment use data;

[0142] The medical expense theme module is used to display the composition and changes of medical expenses, including average outpatient expense data, average inpatient expense data, medical insurance expense data, and out - of - pocket expense data;

[0143] The information data theme module is used to display the hospital informatization construction and data quality situation, including system online rate, data integrity rate, data accuracy rate, and system utilization rate data.

[0144] Specifically, as Figure 3 shown, the hospital president's decision - making system visually displays the hospital operation status in the form of a cockpit, including theme modules such as overall overview, operation management, medical quality, medication analysis, medical expenses, and information data. The overall overview module displays key indicators of the hospital's overall operation status, such as outpatient volume, inpatient volume, operation volume, revenue, and expenditure. The operation management module focuses on hospital operation - related indicators, such as the month - on - month growth rate of outpatient visits, the year - on - year growth rate of discharged patients, and the bed utilization rate. The medical quality module displays medical quality control - related indicators, such as average length of stay, surgical complication rate, hospital infection rate, and antibacterial drug utilization rate. The medication analysis module provides drug use and cost analysis, such as drug - to - total - cost ratio, antibiotic use intensity, and large - equipment utilization rate. The medical expense module analyzes the composition and change trend of medical expenses, such as average outpatient expense, average inpatient expense, medical insurance expense ratio, and out - of - pocket expense ratio. The information data module displays informatization construction and data quality - related indicators, such as system online rate, data integrity rate, data accuracy rate, and system utilization rate. The cancer type screening module supports data screening by cancer type for specialized analysis. For example, it can be used to view the diagnosis and treatment data of specific cancer types such as "lung cancer", "gastric cancer", and "colorectal cancer".

[0145] The hierarchical diagnosis and treatment platform is connected to the big data governance and information platform and the operation management platform, and is used to establish multi - campus patient index associations and hierarchical diagnosis and treatment collaborations.

[0146] Among them, the hierarchical diagnosis and treatment platform includes:

[0147] A patient master index system, which is used to establish a unique identifier for patients within a multi-site hospital medical system. The patient master index system is configured with a patient query module, a patient 360 - view module, a suspected patient management module, an operation history record module, and a weight configuration module. Among them, the patient query module supports querying patient information in multiple ways through the master index number, name, and ID number. The patient 360 - view module displays the complete diagnosis and treatment information of patients in different hospitals. The suspected patient management module automatically identifies different patient records that may be the same person based on a weight algorithm and supports manual review and merging. The operation history record module records all index operations to ensure data traceability. The weight configuration module supports configuring the weight parameters of the patient matching algorithm.

[0148] Specifically, the hierarchical diagnosis and treatment platform is the application layer of the system. It is associated with the standardized data processed by the big data governance and information platform and the patient master index to achieve collaborative hierarchical diagnosis and treatment services. The patient master index system establishes a unique identifier for patients within a multi-site hospital medical system, conforming to the IHE - PIX integration specification and the HL7V3 standard. The system supports querying patient information in multiple ways through the patient query module. The patient 360 - view module displays the complete diagnosis and treatment information of patients in different hospitals. The suspected patient management module automatically identifies different patient records that may be the same person based on a weight algorithm. For example, when the similarity of the names of two patient records is high, the genders are the same, the dates of birth are close, but there are minor differences in the ID numbers, the system will mark these two records as suspected of being the same patient and support manual review and merging.

[0149] Among them, the weight configuration module of the patient master index system includes the following functional units:

[0150] A field weight configuration unit, which is used to set the weight values of each field in the patient matching process. The configurable fields include name, gender, date of birth, ID number, address, and contact phone number;

[0151] A weight range configuration unit, which is used to set the weight range for suspected patient recommendation. Patient records with a weight calculation score within this range are recommended to the suspected patient list;

[0152] A matching algorithm configuration unit, which is used to set the matching algorithms between fields, including exact matching algorithm, fuzzy matching algorithm, edit distance algorithm, and phonetic similarity algorithm;

[0153] A weight test unit, which is used to test the matching effect of the current weight configuration and generate a test report;

[0154] A weight optimization unit, which is used to automatically adjust weight parameters according to historical matching data and optimize the matching accuracy rate.

[0155] Specifically, the weight configuration module supports configuring the weight parameters of the patient matching algorithm. The field weight configuration unit is used to set the weight values of each field in the patient matching process. For example, the weight of the ID number is 0.5, the weight of the name is 0.3, the weight of the gender is 0.1, and the weight of the date of birth is 0.1. The weight range configuration unit is used to set the weight range for the recommendation of suspected patients. For example, if the set range is 0.7 - 0.9, then the patient records with a weight calculation score between 0.7 and 0.9 will be recommended to the list of suspected patients. The matching algorithm configuration unit is used to set the matching algorithm between fields. For example, the name field can adopt the edit distance algorithm (calculating the minimum number of editing steps between two strings) or the pronunciation similarity algorithm (calculating the similarity based on pinyin or pronunciation rules). The weight test unit is used to test the matching effect of the current weight configuration and generate a test report of indicators such as accuracy rate, recall rate, and F1 value. The weight optimization unit automatically adjusts the weight parameters according to historical matching data to improve the matching accuracy rate.

[0156] A remote consultation module, which is used to support cross-hospital expert remote consultation based on the information associated with the patient master index. The remote consultation module includes a consultation application sub-module, an expert scheduling sub-module, a consultation arrangement sub-module, a consultation execution sub-module, and a consultation record sub-module. Among them, the consultation application sub-module supports the initiating hospital to fill in patient information and consultation requirements. The expert scheduling sub-module manages the outpatient schedules of experts in each hospital area. The consultation arrangement sub-module coordinates the consultation time and participants. The consultation execution sub-module supports real-time audio and video interaction and medical image sharing. The consultation record sub-module records the consultation process and conclusions.

[0157] A two-way referral module, which is used to support the orderly referral of patients between medical institutions at different levels. The two-way referral module has a referral application sub-module, a referral review sub-module, a referral reservation sub-module, a referral record sub-module, and a follow-up management sub-module. Among them, the referral application sub-module supports the initiation of upward and downward referral applications and the filling in of basic information. The referral review sub-module reviews the referral application and assigns the receiving department and doctor. The referral reservation sub-module arranges the appointment time for the patient to visit the transferred hospital. The referral record sub-module records all the information of the patient's referral process. The follow-up management sub-module supports the follow-up plan and execution for the referred patients.

[0158] A telemedicine module, which is used to provide cross-campus tele-diagnosis and treatment services. The telemedicine module is configured with a remote outpatient sub-module, a remote ward-round sub-module, a remote pathology sub-module, and a remote imaging sub-module. Among them, the remote outpatient sub-module supports remote outpatient services via video, the remote ward-round sub-module supports senior hospital experts to conduct remote ward-rounds on inpatients in lower-level hospitals, the remote pathology sub-module supports remote diagnosis of pathological sections, and the remote imaging sub-module supports remote diagnosis of medical images.

[0159] Specifically, the remote consultation module supports cross-campus expert remote consultations based on the information associated with the patient master index. The consultation application sub-module supports the initiating hospital to fill in patient information and consultation requirements. The expert scheduling sub-module manages the outpatient schedules of experts in each campus. The consultation arrangement sub-module coordinates the consultation time and participants. The consultation execution sub-module supports real-time audio and video interaction and medical image sharing. The consultation record sub-module records the consultation process and conclusions. The two-way referral module supports the orderly referral of patients between different-level medical institutions, including sub-modules such as referral application, referral review, referral appointment, referral record, and follow-up management. The telemedicine module provides cross-campus tele-diagnosis and treatment services, including sub-modules such as remote outpatient, remote ward-round, remote pathology, and remote imaging.

[0160] In the entire multi-center intelligent specialized hospital collaborative management platform, data flow and processing follow strict sequences and logics. First, the campus end of the data platform and the data collection and aggregation system are responsible for collecting raw data from the business systems of each campus and aggregating it to the central platform. Subsequently, the big data quality detection system conducts quality detection on the aggregated data and screens out the data that meets the quality requirements. Then, the big data standardization system and the big data post-structuring system respectively perform standardization processing on structured data and unstructured data to form medical data assets with unified standards. These standardized data are transmitted to the operation management platform to support the indicator calculation of the operation management decision-making system and the cockpit visualization of the dean decision-making system. At the same time, the standardized data is also used in the patient master index system to establish a unique patient identifier and support hierarchical diagnosis and treatment services such as remote consultation, two-way referral, and telemedicine.

[0161] In a specific embodiment, for example, a large specialized medical group has 7 campuses, including a general hospital and 6 branch hospitals, and each campus has an independent information system. After implementing the multi-center intelligent hospital collaborative management platform, the ETL nodes are first deployed at each campus on the campus side of the data platform to collect data from systems such as HIS (Hospital Information System), EMR (Electronic Medical Record System), LIS (Laboratory Information System), and PACS (Picture Archiving and Communication System) in each campus. The data collection and aggregation system then aggregates the data from each campus to the central platform through a secure data transmission channel. The big data quality detection system performs quality detection tasks on the aggregated data. For example, for the patient basic information table, it executes the "ID number format check" rule and finds that 126 records in Campus A do not conform to the 18-digit standard format. The system automatically marks these records as quality problems and generates a quality report. The big data standardization system standardizes the diagnostic codes in each campus, mapping the ICD-9 codes used in Campus A and the custom codes used in Campus B to the ICD-10 coding system. The big data post-structuring system structures the text data such as admission records and discharge summaries in each campus, extracts key information such as the chief complaint, current medical history, past medical history, and diagnosis conclusion, etc., to form structured clinical data.

[0162] After data governance, the standardized data is transmitted to the operation management platform. The operation management decision-making system calculates various management indicators based on this data. For example, it calculates indicators such as the number of outpatient visits, inpatient admissions, surgical volume, income, and expenditure in the first quarter of 2023 in each campus, and generates year-on-year and month-on-month analyses. After the general dean logs in to the system, he can view the indicator overview of all campuses and finds that the outpatient volume in Campus B has increased by 15% year-on-year, while the outpatient volume in Campus C has decreased by 8% year-on-year. By clicking on the indicator card, he can view the outpatient volume ranking of each campus and analyze the reasons for the change in outpatient volume. The dean decision-making system displays these indicators in the form of a cockpit. The general dean can switch different theme modules to deeply understand the situation in aspects such as medical quality, medication analysis, and medical expenses. For example, in the medical quality module, it is found that the average length of stay in Campus D is significantly higher than that of other campuses. Further analysis reveals that this is due to the relatively high proportion of elderly patients in this campus and the longer postoperative recovery time.

[0163] Meanwhile, the standardized data is also used in the hierarchical diagnosis and treatment platform. Based on this data, the patient master index system matches and correlates the patient information in each hospital area to establish a unique patient identifier. The system discovers that patient "Zhang San" has medical records in both Hospital Area A and Hospital Area E, but there is a one-digit difference in the ID number. After weight calculation, the matching score for this patient is 0.85, and the system recommends the patient to the list of suspected patients. After the administrator reviews and confirms, the system merges the two records to establish a unified patient index. Based on this index, when a doctor in Hospital Area E examines this patient, they can view the complete medical information of the patient in Hospital Area A through the patient 360-view module, including previous medical records, test reports, and imaging materials, etc. When the doctor in Hospital Area E determines that the patient requires a higher level of specialist treatment, they initiate a consultation application to the experts in Hospital Area A through the remote consultation module. The experts in Hospital Area A conduct a three-party consultation with the doctor and the patient in Hospital Area E through the video consultation system to jointly discuss the condition and treatment plan. After determining that the patient needs to be transferred, the doctor in Hospital Area E initiates an upward transfer application through the two-way transfer module. After Hospital Area A receives it, it arranges an expert outpatient appointment time for the patient. After the patient is transferred to Hospital Area A, the system automatically correlates the medical information of the patient in the two hospital areas to form a complete medical record.

[0164] The device embodiments described above are merely illustrative. 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 to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0165] A multi-center intelligent specialized hospital collaborative platform provided by the present invention, in terms of technical architecture, breaks through the limitations of the "chimney-style" construction mode of traditional hospital information systems and constructs a three-layer architecture system with high integration and strong scalability. The big data governance and information platform serves as the basic layer, and through ETL technology and standardized models, it realizes the unified processing of heterogeneous data; the operation management platform serves as the middle layer, converting standardized data into the basis for management decisions; the hierarchical diagnosis and treatment platform serves as the application layer, and based on the patient master index mechanism, it realizes business collaboration. This clear and tightly connected architecture design solves the technical problems of multi-system integration; in terms of data processing capabilities, the present invention realizes comprehensive coverage from structured to unstructured data. Especially in the processing of unstructured medical texts, through innovative post-structuring technology, traditional clinical documents, pathological reports, image descriptions, etc. that are difficult to utilize are transformed into computable data assets, greatly expanding the depth and breadth of medical big data. This processing ability enables the present invention to mine clinical knowledge that traditional systems cannot obtain and provides a data basis for precision medicine; in terms of permission management design, the present invention adopts a role-based multi-level permission system, which precisely matches the hierarchical characteristics of hospital management. Different from simple function point authorization, this permission system organically combines data dimensions, index granularity, and management responsibilities to ensure that managers at all levels obtain appropriate decision-making information while ensuring data security and patient privacy. This refined permission control mechanism adapts to the complex organizational structure and management needs of medical institutions; in terms of the patient master index, the present invention innovatively introduces a configurable weight algorithm and matching technology, improving the accuracy and efficiency of patient identity recognition. Different from traditional hard matching rules, this technology can handle data inconsistency problems in real-world scenarios, such as name spelling differences and ID number entry errors, ensuring high-accuracy patient information association in complex environments and providing technical support for medical continuity; in terms of clinical application value, the hierarchical diagnosis and treatment platform of the present invention realizes the leap from information sharing to business collaboration. Through functional modules such as remote consultation, two-way referral, and telemedicine, it not only breaks down information barriers, but more importantly, reshapes the medical service process, turning hierarchical diagnosis and treatment from a policy concept into an operable technical practice, effectively promoting the rational allocation of medical resources and the improvement of utilization efficiency.

[0166] The present invention is comprehensively innovative in terms of technical implementation and application value. Through systematic, standardized, and intelligent technical means, it effectively solves the key problems in multi-center hospital management and provides strong technical support for the integrated management and collaborative services of medical institutions.

[0167] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A multi-center smart specialist hospital collaboration platform, characterized by: include: Big data governance and information platform, which is used to collect, aggregate, quality test, standardize and structure heterogeneous data from multiple hospital areas, and to reserve standardized data interfaces and index structures for disease-specific research databases; An operations management platform, which is connected to the big data governance and information platform and is used for performing indicator calculation and decision visualization based on standardized data; The hierarchical diagnosis and treatment platform is connected with the big data governance and information platform and the operation management platform, and is used to establish multi-hospital patient index association and hierarchical diagnosis and treatment collaboration.

2. A multi-center smart specialist hospital collaboration platform according to claim 1, characterized in that: The big data governance and information platform includes: The data platform campus end is used to execute the ETL process based on the J2EE framework. The data platform campus end has a data resource management module, a metadata management module, a data service module, a conversion task module, a job flow module, a scheduling plan module and an execution monitoring module. The data resource management module supports the connection configuration of various relational databases, flat files and special data sources. The metadata management module supports the import of table structures and lineage analysis. The data service module provides WebService interface publishing and receiving functions. The conversion task module supports variable setting, data extraction, SQL execution, stored procedure call and external program execution functions. The job flow module supports visual workflow design. The scheduling plan module supports automatic scheduling based on time and events. A data collection and aggregation system, which is used to centrally aggregate data from heterogeneous networks and heterogeneous data sources in each hospital area to a central data platform. The data collection and aggregation system has a collection system status monitoring module, an interface management module, a data dictionary management module, a hospital archive management module, a data period management module and a data monitoring module, wherein the collection system status monitoring module tracks the connection status of the data collection system in each hospital area, the interface management module monitors the call status of the platform's open interfaces, the data dictionary management module maintains the field definition of the data reported by the hospital area, the hospital archive management module records the archive information of the data accessed by each hospital, and the data period management module performs partition management on the data according to the time period; A big data quality detection system, the big data quality detection system is used to perform quality control detection on the aggregated data, the big data quality detection system is configured with a quality control rule management module, a quality control task management module, a quality control result management module, a quality control plan management module and a plan monitoring module, wherein the quality control rule management module supports multiple quality control rules of data relevance, integrity, uniqueness, standardization, rationality, consistency and continuity, the quality control task management module configures the execution parameters and scheduling plan of the quality control task, the quality control result management module records and displays the execution results of the quality control task, and the quality control plan management module combines multiple quality control tasks into a plan and sets a reference standard; A big data standardization system, the big data standardization system is used to perform standardization processing on data that has passed quality inspection, the big data standardization system includes a metadata management module, a data lineage management module, a data security authority management module, a master data management module and a standardized model application module, wherein the metadata management module manages metadata of all data sources in a unified manner, the data lineage management module tracks dependencies during data processing, the data security authority management module controls data access based on roles, and the master data management module maintains dictionaries and mapping relationships of key master data; A big data post-structuring system, wherein the big data post-structuring system is used to perform structured processing on unstructured medical data, and the big data post-structuring system is configured with a data component management module, a task scheduling management module, an offline development management module and a structured model application module, wherein the data component management module provides a preprocessing model for information extraction, the task scheduling management module configures the execution plan and dependency of structured tasks, the offline development management module supports remote calls, HiveSQL, preprocessing and data extraction task types, and the structured model application module performs structured processing on medical texts; A special disease data indexing module, which is used to create special disease related indexes for standardized data. The special disease data indexing module includes a form indexing unit, a field mapping unit and a data association unit, wherein the form indexing unit establishes an index structure for special disease related forms, the field mapping unit provides mapping definitions between standardized fields and special disease research fields, and the data association unit provides association paths between multi-table data.

3. A multi-center smart specialist hospital collaboration platform according to claim 2, characterized in that: The conversion task modules of the data platform campus end include: A variable setting task component, which is used to call a SELECT statement in a data connection to assign a value to a variable, and the variable can be referenced in subsequent conversion tasks; A data extraction task component, which is used to complete data collection and conversion between two data sources or within the same data source. The data extraction task component includes a basic information setting unit, a source table query unit, a target table update unit, and a field mapping unit. The source table query unit supports multiple source table selections and writes SELECT query statements. The target table update unit supports multiple target table update methods, including not deleting, clearing the target table, deleting existing data, deleting the current batch of data, executing custom statements, inserting new data, and adding and updating data. The field mapping unit sets the target table fields corresponding to the source table fields. An SQL execution task component, which is used to directly execute a batch of SQL statements or SQL code blocks. The SQL execution task component supports DDL statements and DML statements, as well as SQL code blocks including variable declaration, judgment, loop and exception handling; A stored procedure task component, which is used to call a database stored procedure to complete data processing, and which supports configuring input parameters, output parameters, and return value variables of a stored procedure; An external command task component, which is used to execute programs and commands under the operating system command line, and supports configuration of execution scripts, successful return values, and return value variable assignment; A data analysis task component, which is used to perform data quality exploration on a target table or SQL statement, and supports analysis types including numeric value analysis, value matching check, character value analysis, date value analysis, Boolean value analysis, duplicate value check, expression matching, reference integrity check, and value distribution analysis; A disease-specific data conversion component is used to convert the data structure of standardized data according to the needs of disease-specific research. The disease-specific data conversion component supports disease-specific library model mapping, multi-level data aggregation and fine-grained data extraction.

4. A multi-center smart specialist hospital collaboration platform according to claim 2, characterized in that: The rule types of the quality control rule management module of the big data quality detection system include: Data association rules, where the data association rules are used to check association relationships between fields across tables; A data integrity rule, wherein the data integrity rule is used to check whether a field value is empty; Data uniqueness rule, wherein the data uniqueness rule is used to check whether a field value is repeated; Data normalization rules, the data normalization rules are used to check whether the field value conforms to a predetermined format, the data normalization rules include a date format check sub-rule, a numeric format check sub-rule, a text format check sub-rule and a value range dictionary check sub-rule; Data rationality rules, which are used to check whether field values ​​comply with medical logic; Data consistency rules, which are used to check the consistency between fields in a table or between fields in different tables; Data continuity rules, which are used to check the continuity of time series data.

5. The multi-center smart specialist hospital collaboration platform according to claim 2, characterized in that: The preprocessing model of the data component management module of the big data post-structuring system includes: A date format conversion model, wherein the date format conversion model is used to convert date strings of various formats into a standard date format; An identity card number digit conversion model, wherein the identity card number digit conversion model is used to unify the number digit format of the identity card number; An electronic medical record document extraction model, the electronic medical record document extraction model is used to extract specific document content from electronic medical records, and the electronic medical record document extraction model supports configuration of start character, end character and ignore character parameters; An electronic medical record chapter extraction model, the electronic medical record chapter extraction model is used to extract the content of a specified chapter from an electronic medical record, and the electronic medical record chapter extraction model supports configuration of expression, start character, and end character parameters; A key cleaning model, wherein the key cleaning model is used to clean key fields in the data; Removing a Chinese character model, wherein the Chinese character removal model is used to remove Chinese characters in a text; An ID card number information extraction model, which is used to extract date of birth and gender information from the ID card number; A full-width to half-width model, wherein the full-width to half-width model is used to convert full-width characters into half-width characters; A sensitive word replacement model, which is used to desensitize key patient information; The encryption and decryption model group includes a CBC encryption model, a CBC decryption model, an MD5 encryption model, a BASE64 encryption model and a BASE64 decryption model, and is used to encrypt and decrypt key patient information.

6. The multi-center smart specialist hospital collaboration platform according to claim 1, characterized in that: The operation management platform includes: An operation management decision system, which is used to perform multi-angle indicator analysis on the standardized data. The operation management decision system has a multi-level authority management module, a multi-dimensional screening module, an indicator card display module, a chart analysis module, a ranking analysis module, a detailed data query module and a department and doctor performance module, wherein the multi-level authority management module sets hierarchical permissions based on the roles of the president, the branch president and the department director, the multi-dimensional screening module supports flexible data screening by the dimensions of hospital area, department and time, the indicator card display module intuitively displays the key indicator values ​​and year-on-year and month-on-month growth, the chart analysis module displays the indicator trend and distribution in a visual way of line charts and bar charts, the ranking analysis module supports indicator ranking comparisons between hospitals, departments and doctors, the detailed data query module provides detailed data query and export functions related to indicators, and the department and doctor performance module automatically calculates department and doctor performance indicators based on business data; The hospital director's decision system is used to display the hospital's operating status in the form of a cockpit. The hospital director's decision system includes an overall overview module, an operation management module, a medical quality module, a medication analysis module, a medical expense module, an information data module, a cancer screening module and a full-screen display module. The overall overview module displays key indicators of the hospital's overall operating status, the operation management module focuses on the analysis of hospital operation-related indicators, the medical quality module displays indicators related to medical quality control, the medication analysis module provides drug use and cost analysis, the medical expense module analyzes the composition and changing trends of medical expenses, the information data module displays indicators related to information construction and data quality, the cancer screening module supports filtering data by cancer type for specialist analysis, and the full-screen display module supports the full-screen switching function of the conference display mode.

7. A multi-center smart specialist hospital collaboration platform according to claim 6, characterized in that: The authority design of the multi-level authority management module of the operation management decision system includes: The chief dean's authority supports viewing the indicator data and rankings of all hospital districts, all departments of any hospital district, and specific departments of any hospital district; The branch dean's authority supports viewing the indicator data and ranking of the affiliated hospital district, all departments of the affiliated hospital district, and specific departments of the affiliated hospital district; The authority of the department director is used to view the indicator data of the configured departments of the hospital area to which he belongs and the ranking of the doctors in the department.

8. The multi-center smart specialist hospital collaboration platform according to claim 6, characterized in that: The dean decision-making system includes: An overall overview theme module, which is used to display the core indicator data of the entire hospital and supports screening by hospital area and cancer type; Operation management theme module, which is used to display hospital operation related indicators, including outpatient volume data, inpatient volume data, surgical volume data, income data and expenditure data; A medical quality theme module, which is used to display indicators related to medical quality, including data on average length of stay, surgical complication rates, hospital infection rates, and antibiotic use rates; A medication analysis theme module, which is used to display drug usage analysis, including drug cost composition data, drug ratio data, antibiotic usage data, and large equipment usage data; A medical expense theme module, which is used to display the composition and changes of medical expenses, including outpatient average cost data, hospitalization average cost data, medical insurance cost data, and self-paid cost data; The information data theme module is used to display the hospital's information construction and data quality status, including system online rate, data completeness rate, data accuracy rate and system utilization rate data.

9. The multi-center smart specialist hospital collaboration platform according to claim 1, characterized in that: The hierarchical diagnosis and treatment platform includes: Patient master index system, the patient master index system is used to establish a unique identifier for patients in a multi-area medical system of a hospital. The patient master index system is configured with a patient query module, a patient 360-degree view module, a suspected patient management module, an operation history module and a weight configuration module, wherein the patient query module supports multiple ways of querying patient information by primary index number, name and ID number, the patient 360-degree view module displays the complete diagnosis and treatment information of patients in different hospitals, the suspected patient management module automatically identifies different patient records that may be of the same person based on a weight algorithm and supports manual review and merging, the operation history module records all index operations to ensure data traceability, and the weight configuration module supports configuration of weight parameters of the patient matching algorithm; A remote consultation module, which is used to support cross-hospital expert remote consultation based on the information associated with the patient's main index. The remote consultation module includes a consultation application submodule, an expert scheduling submodule, a consultation arrangement submodule, a consultation execution submodule and a consultation record submodule, wherein the consultation application submodule supports the initiating hospital to fill in patient information and consultation requirements, the expert scheduling submodule manages the consultation time of experts in each hospital, the consultation arrangement submodule coordinates the consultation time and participants, the consultation execution submodule supports real-time audio and video interaction and medical image sharing, and the consultation record submodule records the consultation process and conclusions; A two-way referral module, which is used to support the orderly referral of patients between medical institutions of different levels. The two-way referral module has a referral application submodule, a referral review submodule, a referral appointment submodule, a referral record submodule and a follow-up management submodule, wherein the referral application submodule supports the initiation of upward and downward transfer applications and the filling of basic information, the referral review submodule reviews the referral application and assigns the receiving department and doctor, the referral appointment submodule arranges the patient's referral time to the hospital, the referral record submodule records the patient's referral process information, and the follow-up management submodule supports the follow-up plan and execution of patients after referral; A telemedicine module is used to provide remote diagnosis and treatment services across hospital campuses. The telemedicine module is configured with a remote outpatient submodule, a remote ward round module, a remote pathology submodule and a remote imaging submodule. The remote outpatient submodule supports remote outpatient services via video, the remote ward round module supports experts from higher-level hospitals to conduct remote ward rounds on inpatients in lower-level hospitals, the remote pathology submodule supports remote diagnosis of pathological sections, and the remote imaging submodule supports remote diagnosis of medical images.

10. A multi-center smart specialist hospital collaboration platform according to claim 9, characterized in that: The weight configuration module of the patient master index system includes the following functional units: A field weight configuration unit, which is used to set the weight value of each field in the patient matching process. The configurable fields include name, gender, date of birth, ID number, address and contact number; A weight range configuration unit, the weight range configuration unit is used to set a weight range for the recommendation of suspected patients, and patient records with weight calculation scores within this range are recommended to the suspected patient list; A matching algorithm configuration unit, the matching algorithm configuration unit is used to set the matching algorithm between fields, including an exact matching algorithm, a fuzzy matching algorithm, an edit distance algorithm, and a pronunciation similarity algorithm; A weight testing unit, which is used to test the matching effect of the current weight configuration and generate a test report; The weight optimization unit is used to automatically adjust the weight parameters according to the historical matching data to optimize the matching accuracy.

Images