Medical business quality control monitoring and early warning system and method based on artificial intelligence

Abstract

The invention discloses a medical business quality control monitoring and early warning system and method based on artificial intelligence, and the system comprises an obtaining module which is used for obtaining medical execution data uploaded by a medical worker through a medical terminal; the medical execution data is data generated when the medical staff executes medical services; the first determination module is used for performing data processing on the medical execution data and determining the data type of the medical execution data; and the monitoring and early warning module is used for executing a corresponding quality control monitoring and early warning strategy based on the data type. The invention aims to solve the problems of low efficiency, poor accuracy, early warning lag and the like in the traditional medical service quality control monitoring process, and realizes the intelligence and precision of medical service quality control by automatically acquiring medical execution data, intelligently determining a data type and executing a corresponding quality control monitoring and early warning strategy based on the data type. The quality control efficiency and accuracy are improved, and the medical risk is reduced.

Description

Technical Field

[0001] The present invention relates to the field of monitoring and early warning technologies, and in particular to an artificial intelligence-based medical service quality control monitoring and early warning system and method. Background Art

[0002] Medical quality is the lifeline of medical institutions. Medical service quality control monitoring and early warning are key links in ensuring medical quality and preventing medical risks. With the rapid development of medical informatization, the amount of data generated by medical services has exploded, covering a wide range of types, including electronic medical records, laboratory test reports, nursing records, and surgical information. However, most medical institutions currently rely primarily on manual review and empirical judgment for medical service quality control monitoring. For example, quality control personnel need to manually review large amounts of medical records to verify information such as the execution of medical orders and medication dosages. This method is not only labor-intensive and time-consuming, but also susceptible to subjective factors and carries a high risk of missed detections and misjudgments.

[0003] At the same time, due to the complexity and diversity of medical data, different types of data have different quality control standards and monitoring priorities. However, traditional quality control and monitoring methods often lack effective classification and targeted processing of medical data, and are unable to achieve accurate quality control of different data types. For example, if unified quality control standards and methods are used for medication data and nursing operation data, it will be difficult to detect specific problems such as unreasonable medication dosage and non-standard nursing operations. In addition, traditional early warning mechanisms are usually based on simple threshold settings and lack in-depth analysis and intelligent judgment of medical data. They are unable to timely and accurately detect potential medical risks and issue early warnings, resulting in the difficulty of timely resolution of medical quality issues and posing a threat to patient health and safety.

[0004] Existing technologies present the following technical challenges: Medical personnel generate a large volume of diverse data during the course of medical services. Traditional manual quality control methods struggle to comprehensively and promptly process and analyze this massive volume of data, making it easy to miss potential medical quality issues. Furthermore, due to a lack of accurate identification and classification of data types, effective quality control monitoring and early warning strategies cannot be developed for different types of data. This results in the inability to promptly identify and address medical risks, impacting both medical quality and patient safety. Summary of the Invention

[0005] The present invention aims to at least partially address one of the technical problems in the aforementioned technologies. To this end, the present invention aims to propose an artificial intelligence-based medical service quality control monitoring and early warning system and method, which aims to address the problems of low efficiency, poor accuracy, and delayed early warning in traditional medical service quality control monitoring. By automatically acquiring medical execution data, intelligently determining the data type, and executing corresponding quality control monitoring and early warning strategies based on the data type, the present invention achieves intelligent and precise medical service quality control, improves quality control efficiency and accuracy, and reduces medical risks.

[0006] To achieve the above objectives, the present invention proposes an artificial intelligence-based medical service quality control monitoring and early warning system, including:

[0007] An acquisition module is used to acquire medical execution data uploaded by medical personnel through a medical terminal; the medical execution data is data generated by the medical personnel when performing medical services;

[0008] A first determining module is used to process the medical execution data and determine the data type of the medical execution data;

[0009] The monitoring and early warning module is used to execute corresponding quality control monitoring and early warning strategies based on data types.

[0010] According to some embodiments of the present invention, the first determining module includes:

[0011] The parsing module is used to parse and process the medical execution data and determine the attribute information;

[0012] The preprocessing module is used to introduce a dynamic time window mechanism to the medical execution data, set the initial window size, perform data standardization and feature extraction on the data within the window, and determine the feature vector;

[0013] The judgment module is used to judge the data type of the medical execution data according to the attribute information and the characteristic vector; the data type includes single-point execution data, single business execution data and composite business execution data.

[0014] According to some embodiments of the present invention, the parsing module includes:

[0015] The second determining module is used to perform time series decomposition on the medical execution data, determine sub-data of each time series, establish a first association relationship, and obtain time attribute information;

[0016] The third determination module is used to perform spatial sequence decomposition on the medical execution data, determine the sub-data of each spatial sequence, and establish a second association relationship to obtain spatial attribute information;

[0017] The fourth determining module is used to determine attribute information according to the temporal attribute information and the spatial attribute information.

[0018] According to some embodiments of the present invention, the determination module determines the data type of the medical execution data based on the attribute information and the feature vector, including:

[0019] The medical execution data whose time attribute information is a single time point, whose spatial attribute information is a single spatial position, and whose feature vector is consistent with the preset feature vector is regarded as single-point execution data;

[0020] The medical execution data whose time attribute information is less than a preset time point, whose spatial attribute information is less than a preset number of spatial positions, and whose feature vector has a difference with a preset feature vector within a preset range is regarded as a single business execution data;

[0021] Medical execution data whose time attribute information is greater than or equal to a preset time point, whose spatial attribute information is greater than or equal to a preset number of spatial positions, and whose feature vector and the difference between the preset feature vector and the preset feature vector are not within a preset range are regarded as composite business execution data.

[0022] According to some embodiments of the present invention, the monitoring and early warning module includes:

[0023] The first execution module is used to perform single-point behavior quality control monitoring and early warning strategies on single-point execution data;

[0024] The second execution module is used to execute a single business behavior quality control monitoring and early warning strategy for a single business execution data; the single business behavior quality control monitoring and early warning strategy includes a single point behavior quality control monitoring and early warning strategy and a first business node interactive behavior quality control monitoring and early warning strategy;

[0025] The third execution module is used to execute composite business behavior quality control monitoring and early warning strategies on composite business execution data; the composite business behavior quality control monitoring and early warning strategies include single-point behavior quality control monitoring and early warning strategies, first business node interactive behavior quality control monitoring and early warning strategies, and second business node interactive behavior quality control monitoring and early warning strategies.

[0026] According to some embodiments of the present invention, the first execution module is configured to:

[0027] Send a rule request for single-point behavior quality control monitoring and early warning strategy to the monitoring and early warning rule configuration platform. The monitoring and early warning rule configuration platform verifies the legitimacy of the rule request and returns the single-point rule configuration file.

[0028] Parse the single-point rule configuration file to determine the programming language information; select the appropriate target processing engine from the candidate processing engines based on the programming language information;

[0029] The target processing engine reads the single-point rule configuration file, determines the single-point user ID and the execution content corresponding to the single-point user based on the reading result and the single-point execution data, and generates a candidate recommendation rule set for the single-point user;

[0030] Generate single-point behavior quality control monitoring and early warning strategies based on the candidate recommendation rule set, and execute single-point behavior quality control monitoring and early warning strategies on single-point execution data.

[0031] According to some embodiments of the present invention, the second execution module determines the quality control monitoring and early warning strategy for the interaction behavior of the first service node, including:

[0032] Determine the data node list of internal nodes included in the execution data of a single business, and generate the logical link corresponding to the single business;

[0033] Traverse the logical link, for each internal node, calculate the distance to the adjacent node and determine the weight coefficient;

[0034] Determine the rules for each internal node;

[0035] Fusing the rules based on the order of interaction of the first service node and the weight coefficient to determine the fusion rule;

[0036] The fusion rules are configured on all physical ports of the logical link to generate a quality control monitoring and early warning strategy for the interaction behavior of the first service node.

[0037] According to some embodiments of the present invention, the third execution module generates a second service node interactive behavior quality control monitoring and early warning strategy, including:

[0038] Set up the intermediate interaction area for composite business;

[0039] The interaction data of each business is displayed in the middle interaction area; the interaction data includes parameter setting options for node types and parameter setting options for clock deviations between clock signals with timing relationships;

[0040] Model the interaction data based on the Gaussian mixture function and extract the interaction feature parameters;

[0041] The similarity matrix is ​​calculated based on the interaction feature parameters, and the spectral clustering method is used to divide the intermediate interaction area into several sub-areas.

[0042] Based on the hierarchical clustering method, several sub-regions are merged and the interactive data chain is obtained based on the merging results;

[0043] Based on the interaction sequence of the interactive data chain, the rules corresponding to each interactive data chain are fused, and based on the fusion result, the second business node interactive behavior quality control monitoring and early warning strategy is generated.

[0044] According to some embodiments of the present invention, the further comprising: a fifth determination module, configured to receive a status code returned by executing a corresponding quality control monitoring and early warning strategy based on the data type, and determine an execution result based on the status code.

[0045] According to some embodiments of the present invention, the monitoring and early warning method of the medical service quality control monitoring and early warning system based on artificial intelligence as described above includes:

[0046] Obtaining medical execution data uploaded by medical personnel through medical terminals; the medical execution data is data generated by the medical personnel performing medical services;

[0047] Processing medical execution data to determine the data type of the medical execution data;

[0048] Execute corresponding quality control monitoring and early warning strategies based on data types.

[0049] The present invention proposes an artificial intelligence-based medical service quality control monitoring and early warning system and method, which is committed to solving the problems of low efficiency, poor accuracy, and delayed early warning in the traditional medical service quality control monitoring process. By automatically acquiring medical execution data, intelligently determining the data type, and executing corresponding quality control monitoring and early warning strategies based on the data type, it realizes the intelligent and precise quality control of medical services, improves quality control efficiency and accuracy, and reduces medical risks.

[0050] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0051] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0053] Figure 1 is a block diagram of a medical service quality control monitoring and early warning system based on artificial intelligence according to one embodiment of the present invention;

[0054] Figure 2 is a block diagram of a first determination module according to one embodiment of the present invention;

[0055] Figure 3 The present invention is a flowchart of an artificial intelligence-based medical service quality control monitoring and early warning method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0057] like Figure 1 As shown, the embodiment of the present invention proposes an artificial intelligence-based medical service quality control monitoring and early warning system, including:

[0058] An acquisition module is used to acquire medical execution data uploaded by medical personnel through a medical terminal; the medical execution data is data generated by the medical personnel when performing medical services;

[0059] A first determining module is used to process the medical execution data and determine the data type of the medical execution data;

[0060] The monitoring and early warning module is used to execute corresponding quality control monitoring and early warning strategies based on data types.

[0061] The working principle of the above technical solution is as follows: the acquisition module is responsible for real-time or batch collection of medical execution data uploaded by medical staff through medical terminals (such as electronic medical record systems, nursing workstations, testing and inspection equipment, etc.). Medical execution data includes: diagnosis and treatment data, such as medical order execution records, surgical records, medication dosage and time, examination report results, etc.; nursing data, such as nursing operation time, vital signs monitoring records, patient admission and discharge assessments, etc.; management data, such as medical staff scheduling, consumables usage records, department operation indicators, etc. The first determination module is used to determine the data type of medical execution data; the monitoring and early warning module executes corresponding quality control monitoring and early warning strategies based on the data type.

[0062] The beneficial effects of the above technical solution are: by automatically acquiring medical execution data, intelligently determining data types, and executing corresponding quality control monitoring and early warning strategies based on data types, intelligent and precise quality control of medical services can be achieved, quality control efficiency and accuracy can be improved, and medical risks can be reduced.

[0063] like Figure 2 As shown, according to some embodiments of the present invention, the first determining module includes:

[0064] The parsing module is used to parse and process the medical execution data and determine the attribute information;

[0065] The preprocessing module is used to introduce a dynamic time window mechanism to the medical execution data, set the initial window size, perform data standardization and feature extraction on the data within the window, and determine the feature vector;

[0066] The judgment module is used to judge the data type of the medical execution data according to the attribute information and the characteristic vector; the data type includes single-point execution data, single business execution data and composite business execution data.

[0067] The working principle of the above technical solution is as follows: The parsing module parses the acquired medical execution data and extracts its attribute information. The preprocessing module introduces a dynamic time window mechanism that flexibly adjusts the window size based on data characteristics and analysis requirements. After the initial window size is set, the data within the window is normalized to eliminate dimensional differences. Key features, including statistical features (such as mean, variance, maximum, and minimum values) and temporal features (such as rate of change and trend), are extracted and combined into a feature vector. The feature vector represents the characteristics corresponding to the medical staff's operational behavior. The judgment module determines the data type of the medical execution data based on the attribute information and feature vector. Single-point execution data refers to a single data point collected at a single point in time and at a single spatial location through a single operation by the medical staff (such as a single heart rate measurement). Single service execution data refers to a collection of data related to a specific service process, collected over a period of time and across multiple departments and locations, based on multiple operations performed by the medical staff (such as a physical examination data set). Composite service execution data includes multiple single service execution data points. A data set involving multiple business processes (such as all medical records during hospitalization) collected in multiple time periods, multiple departments, and based on more operations performed by medical staff.

[0068] Beneficial effects of the above technical solution: The first determination module accurately determines the data type of the medical execution data through the collaborative work of the three sub-modules of analysis, preprocessing and judgment.

[0069] According to some embodiments of the present invention, the parsing module includes:

[0070] The second determining module is used to perform time series decomposition on the medical execution data, determine sub-data of each time series, establish a first association relationship, and obtain time attribute information;

[0071] The third determination module is used to perform spatial sequence decomposition on the medical execution data, determine the sub-data of each spatial sequence, and establish a second association relationship to obtain spatial attribute information;

[0072] The fourth determining module is used to determine attribute information according to the temporal attribute information and the spatial attribute information.

[0073] The working principle of the above technical solution is as follows: The second determination module divides the medical execution data into multiple time windows based on sliding window technology. Each window contains data from a specific time period. Time series analysis techniques (such as moving average and exponential smoothing) are used to identify trends and periodicity in the data. Correlations between data in different time windows are calculated, and similar time series patterns are clustered together to identify common characteristics. Relationships between different time series sub-data are determined to understand the data's changing patterns in the temporal dimension. Temporal attribute information includes time window divisions, trend analysis results, and periodic characteristics. The third determination module divides the medical execution data into multiple spatial regions (such as data from different hospital departments or different equipment). Spatial statistical techniques (such as spatial autocorrelation analysis) are used to identify the spatial distribution characteristics of the data. Correlations between data in different spatial regions are calculated, and similar spatial data patterns are clustered together to identify common characteristics. Relationships between different spatial series sub-data are determined to understand the data's distribution patterns in the spatial dimension. Spatial attribute information includes spatial region divisions, spatial distribution characteristics, and spatial correlations. The fourth determination module analyzes the interactions between the data in the temporal and spatial dimensions, combining temporal and spatial characteristics to extract more comprehensive attribute information. Attribute information includes temporal attribute information, spatial attribute information, and the results of spatiotemporal interaction.

[0074] The beneficial effects of the above technical solution are as follows: the parsing module can comprehensively analyze the medical execution data from two dimensions, time and space, so as to accurately determine the attribute information.

[0075] According to some embodiments of the present invention, the determination module determines the data type of the medical execution data based on the attribute information and the feature vector, including:

[0076] The medical execution data whose time attribute information is a single time point, whose spatial attribute information is a single spatial position, and whose feature vector is consistent with the preset feature vector is regarded as single-point execution data;

[0077] The medical execution data whose time attribute information is less than a preset time point, whose spatial attribute information is less than a preset number of spatial positions, and whose feature vector has a difference with a preset feature vector within a preset range is regarded as a single business execution data;

[0078] Medical execution data whose time attribute information is greater than or equal to a preset time point, whose spatial attribute information is greater than or equal to a preset number of spatial positions, and whose feature vector and the difference between the preset feature vector and the preset feature vector are not within a preset range are regarded as composite business execution data.

[0079] The working principle of the above technical solution is as follows: For single-point execution data, the check is whether it is a single time point (such as a specific moment), whether it is a single spatial location (such as a specific department or equipment), and the similarity between the current data feature vector and the preset feature vector is calculated (such as using cosine similarity or Euclidean distance) to determine whether they are consistent. If the above conditions are met, it is determined to be single-point execution data. For single business execution data, the check is whether the time span is less than a preset time point (such as a few hours or a day), whether the number of spatial locations involved is less than a preset number (such as involving less than 3 departments or equipment), and the difference between the current data feature vector and the preset feature vector is calculated. If the difference is within the preset range, it is determined to be single business execution data. For composite business execution data, the check is whether the time span is greater than or equal to the preset time point, whether the number of spatial locations involved is greater than or equal to the preset number (such as involving 3 or more departments or equipment), and the difference between the current data feature vector and the preset feature vector is calculated. If the difference is not within the preset range, it is determined to be composite business execution data.

[0080] The beneficial effects of the above technical solution are: the judgment module can accurately classify medical execution data, providing a basis for subsequent data analysis and processing.

[0081] According to some embodiments of the present invention, the monitoring and early warning module includes:

[0082] The first execution module is used to perform single-point behavior quality control monitoring and early warning strategies on single-point execution data;

[0083] The second execution module is used to execute a single business behavior quality control monitoring and early warning strategy for a single business execution data; the single business behavior quality control monitoring and early warning strategy includes a single point behavior quality control monitoring and early warning strategy and a first business node interactive behavior quality control monitoring and early warning strategy;

[0084] The third execution module is used to execute composite business behavior quality control monitoring and early warning strategies on composite business execution data; the composite business behavior quality control monitoring and early warning strategies include single-point behavior quality control monitoring and early warning strategies, first business node interactive behavior quality control monitoring and early warning strategies, and second business node interactive behavior quality control monitoring and early warning strategies.

[0085] The working principle of the above technical solution is as follows: The first execution module is used to implement single-point behavior quality control monitoring and early warning strategies on single-point execution data, ensuring that medical execution data collected at a single time point and a single spatial location meets preset quality standards. When single-point behavior quality control monitoring detects data anomalies, an early warning is triggered. Relevant personnel are notified through sound, light, text message, email, etc. The second execution module implements single-service behavior quality control monitoring and early warning strategies on single-service execution data, including single-point behavior quality control monitoring and early warning strategies and first-service node interaction behavior quality control monitoring and early warning strategies. This includes single-point behavior quality control testing, similar to the first execution module, but targeting single-point data within a single service execution data. First-service node interaction behavior quality control monitoring ensures that interactions between different service nodes (such as different departments or different equipment) during a single service execution comply with preset processes and standards. First-service nodes are nodes between single services. Interactions between service nodes are verified to follow predefined processes and to check whether interactions between service nodes are completed within a reasonable timeframe. Anomalies in single-point behavior and first-service node interaction behavior trigger corresponding early warnings. The warning content should include specific information about the anomaly, possible impacts, and recommended handling measures. The third execution module executes composite business behavior quality control monitoring and early warning strategies on composite business execution data, including single-point behavior quality control monitoring and early warning strategies, first business node interactive behavior quality control monitoring and early warning strategies, and second business node interactive behavior quality control monitoring and early warning strategies. Single-point behavior quality control monitoring and first business node interactive behavior quality control monitoring are similar to the second execution model. The second business node is a node in the composite business. The second business node interactive behavior quality control monitoring ensures that during the composite business execution process, the interactive behaviors with multiple businesses comply with preset standards and protocols. Corresponding warnings are triggered for anomalies in single-point behavior, interactive behavior of the first business node, and interactive behavior of the second business node.

[0086] Beneficial effects of the above technical solution: The monitoring and early warning module can execute corresponding quality control monitoring and early warning strategies for different types of data, such as single point, single business and complex business, to ensure the quality and safety of medical execution.

[0087] According to some embodiments of the present invention, the first execution module is configured to:

[0088] Send a rule request for single-point behavior quality control monitoring and early warning strategy to the monitoring and early warning rule configuration platform. The monitoring and early warning rule configuration platform verifies the legitimacy of the rule request and returns the single-point rule configuration file.

[0089] Parse the single-point rule configuration file to determine the programming language information; select the appropriate target processing engine from the candidate processing engines based on the programming language information;

[0090] The target processing engine reads the single-point rule configuration file, determines the single-point user ID and the execution content corresponding to the single-point user based on the reading result and the single-point execution data, and generates a candidate recommendation rule set for the single-point user;

[0091] Generate single-point behavior quality control monitoring and early warning strategies based on the candidate recommendation rule set, and execute single-point behavior quality control monitoring and early warning strategies on single-point execution data.

[0092] The above technical solution works as follows: A rule request is sent to the rule configuration platform via a RESTful API or other communication protocol. The rule request includes the request type (for single-point behavior quality control) and data type identifier. Authentication is performed using methods such as API keys and OAuth tokens, and the integrity and validity of the request parameters are checked. This ensures the legitimacy of the request and prevents unauthorized access. After verification, the rule configuration platform returns a rule configuration file for the single-point behavior quality control monitoring and early warning strategy. The configuration file is parsed using tools such as JSON parsers and XML parsers. Key information such as programming language information and rule logic is extracted. A list of candidate processing engines is maintained, each supporting one or more programming languages. Based on the parsed programming language information, an appropriate engine is selected from the candidate engines. The rule configuration file is loaded into the memory of the target processing engine, where the rule logic is parsed and prepared for execution. User identification information is extracted from the single-point execution data, and the specific data items and thresholds to be monitored are determined based on the rule configuration file. The user data is matched against the rules in the rule library to filter out applicable rules. Based on the matching results, a set of candidate recommended rules is generated. Monitor single-point execution data in real time to check whether it complies with the requirements of the rule set. When the data is abnormal, trigger the early warning mechanism and notify relevant personnel.

[0093] The beneficial effects of the above technical solution are: the first execution module can effectively implement single-point behavior quality control monitoring and early warning strategies to ensure the quality and security of medical execution data.

[0094] According to some embodiments of the present invention, the second execution module determines the quality control monitoring and early warning strategy for the interaction behavior of the first service node, including:

[0095] Determine the data node list of internal nodes included in the execution data of a single business, and generate the logical link corresponding to the single business;

[0096] Traverse the logical link, for each internal node, calculate the distance to the adjacent node and determine the weight coefficient;

[0097] Determine the rules for each internal node;

[0098] Fusing the rules based on the order of interaction of the first service node and the weight coefficient to determine the fusion rule;

[0099] The fusion rules are configured on all physical ports of the logical link to generate a quality control monitoring and early warning strategy for the interaction behavior of the first service node.

[0100] The working principle of the above technical solution is as follows: Information on all internal nodes is extracted from business execution data to form a data node list. Node information includes node identification, node type, and node function. Based on the business execution process, the interaction sequence and relationships between nodes are determined. Using graph theory or network topology analysis methods, a logical link diagram is generated that represents the node interaction sequence. Based on the logical relationships between nodes, the distances between nodes and adjacent nodes are calculated, which can be retrieved by querying a preset logical relationship-distance database. An initial weight coefficient is set for the first node in the logical link. The weight coefficients for the second to last nodes are then determined based on the ratios of the distances between the node and its adjacent nodes. Quality control rules applicable to each node are defined based on the node type and function. The node interaction sequence and weight coefficients are analyzed to determine the priority and order for rule fusion. Rules from different nodes are fused using methods such as a rule engine or decision table. Fusion rules are generated that reflect the overall business logic and node interaction characteristics. The fusion rules should be able to guide quality control monitoring and early warning of the interaction behavior of the entire first business node. The fusion rules are configured on the physical ports involved in the logical link to ensure that the rules are correctly executed when data passes through these ports. Based on the configured rules, a quality control monitoring and early warning strategy for the interaction behavior of the first business node is generated.

[0101] Beneficial effects of the above technical solution: the second execution module can effectively determine the quality control monitoring and early warning strategy of the interactive behavior of the first business node, ensuring that the interactive behavior of the internal nodes during the business execution process meets the preset standards and requirements.

[0102] According to some embodiments of the present invention, the third execution module generates a second service node interactive behavior quality control monitoring and early warning strategy, including:

[0103] Set up the intermediate interaction area for composite business;

[0104] The interaction data of each business is displayed in the middle interaction area; the interaction data includes parameter setting options for node types and parameter setting options for clock deviations between clock signals with timing relationships;

[0105] Model the interaction data based on the Gaussian mixture function and extract the interaction feature parameters;

[0106] The similarity matrix is ​​calculated based on the interaction feature parameters, and the spectral clustering method is used to divide the intermediate interaction area into several sub-areas.

[0107] Based on the hierarchical clustering method, several sub-regions are merged and the interactive data chain is obtained based on the merging results;

[0108] Based on the interaction sequence of the interactive data chain, the rules corresponding to each interactive data chain are fused, and based on the fusion result, the second business node interactive behavior quality control monitoring and early warning strategy is generated.

[0109] The working principle of the above technical solution is as follows: A logical area is set up in the system specifically for storing and displaying data interacting with the second service node. Access rights to the intermediate interaction area are configured to ensure that only authorized personnel can access and modify data. Data update frequency and retention policies are set to ensure data timeliness and accuracy. Interaction data is displayed in the intermediate interaction area using visualization tools (such as charts and tables). The displayed data should include key information such as node type, parameter settings, clock signal, and deviation. A parameter setting interface for node types is provided, allowing users to adjust parameters as needed. A parameter setting option for clock signal deviation is provided to calibrate and synchronize clocks between different nodes. A Gaussian mixture model (GMM) is used to model the interaction data to capture its distribution characteristics. Model training determines the parameters of the Gaussian mixture function (such as mean, variance, and mixing coefficient). Key parameters that reflect interaction behavior characteristics, such as the center location and dispersion of the data distribution, are extracted from the trained Gaussian mixture model. The extracted interaction feature parameters are used to calculate the similarity between different interaction data. Similarity can be calculated using methods such as Euclidean distance and cosine similarity. Based on the similarity matrix, a spectral clustering algorithm is used to partition the intermediate interaction area. Spectral clustering can effectively process nonlinearly separable data, resulting in more reasonable subregion divisions. Each subregion should contain data with similar interaction characteristics. A hierarchical clustering algorithm is used to merge subregions. Hierarchical clustering can gradually merge subregions into larger regions based on their similarity, forming a hierarchical structure. Based on the hierarchical clustering results, an appropriate merging threshold is selected to merge similar subregions into larger regions. The merged regions should reflect the overall characteristics and trends of the interaction behavior. Based on the merged regions, an interaction data chain is generated. The interaction data chain should clearly display the interaction sequence and relationships between different business nodes, facilitating a better understanding and analysis of interaction behavior. The interaction sequence and relationships of each interaction data chain are analyzed to identify key points requiring monitoring and early warning. The rules for different interaction data chains are integrated to form a unified set of quality control monitoring and early warning rules. Based on this integrated rule set, a quality control monitoring and early warning strategy for the interaction behavior of the second business node is generated.

[0110] Beneficial effects of the above technical solution: The third execution module can effectively generate quality control monitoring and early warning strategies for the interactive behaviors of the second business nodes, ensuring that the interactive behaviors between various businesses during the execution of the composite business comply with preset standards and protocols.

[0111] According to some embodiments of the present invention, the further comprising: a fifth determination module, configured to receive a status code returned by executing a corresponding quality control monitoring and early warning strategy based on the data type, and determine an execution result based on the status code.

[0112] The working principle of the above technical solution is: through a preset interface or callback mechanism, status codes are received from various execution modules (such as the first execution module, the second execution module, the third execution module, etc.). The status code can be a number, a string or a specific identifier, which is used to represent different states of policy execution (such as success, failure, partial success, etc.). A mapping relationship between the status code and the specific execution result is established. For example, the status code "200" may indicate successful execution, and the status code "500" may indicate failed execution. According to the mapping relationship, the received status code is parsed to determine the corresponding execution result. According to the parsed result of the status code, the execution status of the quality control monitoring and early warning strategy is clarified.

[0113] Beneficial effects of the above technical solution: The fifth determination module can effectively receive, parse and process the status code after the quality control monitoring and early warning strategy is executed, which facilitates the stable operation of the system.

[0114] like Figure 3 As shown, according to some embodiments of the present invention, the monitoring and early warning method of the medical service quality control monitoring and early warning system based on artificial intelligence as described above includes steps S1-S3:

[0115] S1. Obtaining medical execution data uploaded by medical personnel through a medical terminal; the medical execution data is data generated by the medical personnel performing medical services;

[0116] S2. Process the medical execution data to determine the data type of the medical execution data;

[0117] S3. Execute corresponding quality control monitoring and early warning strategies based on data types.

[0118] The beneficial effects of the above technical solution are: by automatically acquiring medical execution data, intelligently determining data types, and executing corresponding quality control monitoring and early warning strategies based on data types, intelligent and precise quality control of medical services can be achieved, quality control efficiency and accuracy can be improved, and medical risks can be reduced.

[0119] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A medical service quality control monitoring and early warning system based on artificial intelligence, characterized by: include: The acquisition module is used to obtain the medical execution data uploaded by medical staff through the medical terminal; The medical execution data is the data generated by the medical staff when performing medical services; A first determining module is used to process the medical execution data and determine the data type of the medical execution data; The monitoring and early warning module is used to execute corresponding quality control monitoring and early warning strategies based on data types.

2. The artificial intelligence-based medical service quality control monitoring and early warning system according to claim 1, characterized in that: The first determination module includes: The parsing module is used to parse and process the medical execution data and determine the attribute information; The preprocessing module is used to introduce a dynamic time window mechanism to the medical execution data, set the initial window size, perform data standardization and feature extraction on the data within the window, and determine the feature vector; The judgment module is used to judge the data type of the medical execution data according to the attribute information and the characteristic vector; the data type includes single-point execution data, single business execution data and composite business execution data.

3. The artificial intelligence-based medical service quality control monitoring and early warning system according to claim 2, characterized in that: The parsing module includes: The second determining module is used to perform time series decomposition on the medical execution data, determine sub-data of each time series, establish a first association relationship, and obtain time attribute information; The third determination module is used to perform spatial sequence decomposition on the medical execution data, determine the sub-data of each spatial sequence, and establish a second association relationship to obtain spatial attribute information; The fourth determining module is used to determine attribute information according to the temporal attribute information and the spatial attribute information.

4. The artificial intelligence-based medical service quality control monitoring and early warning system according to claim 3, characterized in that: The determination module determines the data type of the medical execution data based on the attribute information and the feature vector, including: The medical execution data whose time attribute information is a single time point, whose spatial attribute information is a single spatial position, and whose feature vector is consistent with the preset feature vector is regarded as single-point execution data; The medical execution data whose time attribute information is less than a preset time point, whose spatial attribute information is less than a preset number of spatial positions, and whose feature vector has a difference with a preset feature vector within a preset range is regarded as a single business execution data; Medical execution data whose time attribute information is greater than or equal to a preset time point, whose spatial attribute information is greater than or equal to a preset number of spatial positions, and whose feature vector and the difference between the preset feature vector and the preset feature vector are not within a preset range are regarded as composite business execution data.

5. The artificial intelligence-based medical service quality control monitoring and early warning system according to claim 2, characterized in that: The monitoring and early warning module includes: The first execution module is used to perform single-point behavior quality control monitoring and early warning strategies on single-point execution data; The second execution module is used to execute a single business behavior quality control monitoring and early warning strategy for a single business execution data; the single business behavior quality control monitoring and early warning strategy includes a single point behavior quality control monitoring and early warning strategy and a first business node interactive behavior quality control monitoring and early warning strategy; The third execution module is used to execute composite business behavior quality control monitoring and early warning strategies on composite business execution data; the composite business behavior quality control monitoring and early warning strategies include single-point behavior quality control monitoring and early warning strategies, first business node interactive behavior quality control monitoring and early warning strategies, and second business node interactive behavior quality control monitoring and early warning strategies.

6. The artificial intelligence-based medical service quality control monitoring and early warning system according to claim 5, characterized in that: The first execution module is configured to: Send a rule request for single-point behavior quality control monitoring and early warning strategy to the monitoring and early warning rule configuration platform. The monitoring and early warning rule configuration platform verifies the legitimacy of the rule request and returns the single-point rule configuration file. Parse the single-point rule configuration file to determine the programming language information; Selecting an adapted target processing engine from candidate processing engines according to programming language information; The target processing engine reads the single-point rule configuration file, determines the single-point user ID and the execution content corresponding to the single-point user based on the reading result and the single-point execution data, and generates a candidate recommendation rule set for the single-point user; Generate single-point behavior quality control monitoring and early warning strategies based on the candidate recommendation rule set, and execute single-point behavior quality control monitoring and early warning strategies on single-point execution data.

7. The artificial intelligence-based medical service quality control monitoring and early warning system according to claim 6, characterized in that: The second execution module determines a quality control monitoring and early warning strategy for the interactive behavior of the first service node, including: Determine the data node list of internal nodes included in the execution data of a single business, and generate the logical link corresponding to the single business; Traverse the logical link, for each internal node, calculate the distance to the adjacent node and determine the weight coefficient; Determine the rules for each internal node; Fusing the rules based on the order of interaction of the first service node and the weight coefficient to determine the fusion rule; The fusion rules are configured on all physical ports of the logical link to generate a quality control monitoring and early warning strategy for the interaction behavior of the first service node.

8. The artificial intelligence-based medical service quality control monitoring and early warning system according to claim 7, characterized in that: The third execution module generates a second service node interactive behavior quality control monitoring and early warning strategy, including: Set up the intermediate interaction area for composite business; The interaction data of each business is displayed in the middle interaction area; the interaction data includes parameter setting options for node types and parameter setting options for clock deviations between clock signals with timing relationships; Model the interaction data based on the Gaussian mixture function and extract the interaction feature parameters; The similarity matrix is ​​calculated based on the interaction feature parameters, and the spectral clustering method is used to divide the intermediate interaction area into several sub-areas. Based on the hierarchical clustering method, several sub-regions are merged and the interactive data chain is obtained based on the merging results; Based on the interaction sequence of the interactive data chain, the rules corresponding to each interactive data chain are fused, and based on the fusion result, the second business node interactive behavior quality control monitoring and early warning strategy is generated.

9. The artificial intelligence-based medical service quality control monitoring and early warning system according to claim 1, characterized in that: Also includes: The fifth determination module is used to receive a status code returned by executing a corresponding quality control monitoring and early warning strategy based on the data type, and determine the execution result based on the status code.

10. The monitoring and early warning method of the medical service quality control monitoring and early warning system based on artificial intelligence according to any one of claims 1 to 9, characterized in that: include: Obtain medical execution data uploaded by medical staff through medical terminals; The medical execution data is the data generated by the medical staff when performing medical services; Processing medical execution data to determine the data type of the medical execution data; Execute corresponding quality control monitoring and early warning strategies based on data types.

Images