Medical product full-chain adverse reaction data monitoring integration and early warning platform

Abstract

The medical product full-chain adverse reaction data monitoring, integration and early warning platform proposed in the present application relates to the field of data processing technology. The full chain of the product is determined by retrieving the product data source and traversing to obtain multiple product nodes; the full chain of the product is collected in real time based on the node and then a collection cycle is formulated; data monitoring and sensing of the product nodes are performed according to the cycle; the monitoring data is identified and integrated according to the product data source to determine the adverse reaction data set and perform risk assessment, generate early warning information, and finally release the early warning information through the early warning platform, which solves the technical problems of incomplete data integration, delayed monitoring and inaccurate risk assessment in the monitoring of adverse reaction data of medical products, achieves the goal of improving the coverage, timeliness and accuracy of monitoring, and thus effectively improves the technical effect of preventing and responding to adverse reactions of medical products.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a full-chain adverse reaction data monitoring integration and early warning platform for medical products. Background Art

[0002] In the modern medical industry, with the increasing variety and frequency of use of medical products, the safety supervision of these products has become increasingly important, especially the monitoring and management of adverse reactions of medical products, which is a key link in ensuring public health and safety. Traditional medical product adverse reaction monitoring systems usually rely on post-event reporting and decentralized data management methods, which leads to delayed and incomplete data collection. In addition, existing technologies often lack comprehensive monitoring of the entire chain of medical products, from production, distribution to final consumption, making early warning and response to potential risks not timely or effective, which in turn leads to information islands and threatens public health and safety.

[0003] In summary, the existing technology for monitoring adverse reaction data of medical products has technical problems such as incomplete data integration, delayed monitoring and inaccurate risk assessment. Summary of the invention

[0004] This application provides a full-chain adverse reaction data monitoring integration and early warning platform for medical products, which is used to solve the technical problems in the existing technology of incomplete data integration, delayed monitoring and inaccurate risk assessment in the monitoring of adverse reaction data of medical products.

[0005] This application provides a full-chain adverse reaction data monitoring integration and early warning platform for medical products, which includes:

[0006] A node information acquisition module is used to retrieve multiple product data sources to determine the entire chain of medical products, traverse the entire chain of medical products, and determine multiple product nodes; a real-time acquisition module is used to perform real-time acquisition of the entire chain of medical products based on the multiple product nodes, and formulate a data acquisition cycle according to the real-time acquisition results; a monitoring and sensing module is used to perform data monitoring and sensing on the multiple product nodes according to the data acquisition cycle to generate multiple product monitoring data; a data identification module is used to perform data identification and integration on the multiple product monitoring data according to the multiple product data sources to determine an adverse reaction data set; a risk assessment module is used to perform risk assessment on the adverse reaction data set, generate early warning information based on the assessment results, and publish the early warning information to the entire chain through an early warning platform.

[0007] The medical product full-chain adverse reaction data monitoring integration and early warning platform proposed in this application is intended to determine the full chain of medical products by retrieving multiple product data sources, traverse the full chain of medical products, and determine multiple product nodes; based on the multiple product nodes, real-time collection of the full chain of medical products is performed, and a data collection cycle is formulated according to the real-time collection results; data monitoring and sensing of the multiple product nodes are performed according to the data collection cycle to generate multiple product monitoring data; data identification and integration of the multiple product monitoring data are performed according to the multiple product data sources to determine the adverse reaction data set; risk assessment is performed on the adverse reaction data set, and early warning information is generated according to the assessment results. The early warning information is released to the entire chain through the early warning platform, which solves the technical problems of incomplete data integration, delayed monitoring, and inaccurate risk assessment in the prior art in the monitoring of adverse reaction data of medical products. By realizing full-chain monitoring, real-time data collection adjustment, and integrated data processing and risk assessment, the coverage, timeliness and accuracy of monitoring are improved, thereby effectively improving the technical effect of preventing and responding to adverse reactions of medical products. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A schematic diagram of the structure of the medical product full-chain adverse reaction data monitoring integration and early warning platform provided in the embodiment of the present application.

[0009] Figure 2 A flow chart of the data identification module in the medical product full-chain adverse reaction data monitoring integration and early warning platform provided in the embodiment of the present application.

[0010] Explanation of the accompanying drawings: node information acquisition module 11, real-time collection module 12, monitoring sensor module 13, data identification module 14, risk assessment module 15. DETAILED DESCRIPTION

[0011] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0012] The present application embodiment provides a medical product full-chain adverse reaction data monitoring integration and early warning platform, such as Figure 1 As shown, the platform includes:

[0013] The node information acquisition module 11 is used to retrieve multiple product data sources to determine the entire chain of medical products, and traverse the entire chain of medical products to determine multiple product nodes.

[0014] Specifically, the main function of the node information acquisition module is to retrieve and integrate information from multiple product data sources to determine the entire chain of medical products. By carefully traversing the entire chain, the module can accurately determine the various product nodes involved. First, the data source is retrieved, which involves extracting information from multiple independent but related data sources, including production batch records, distribution logs, user feedback systems, etc. Each data source provides data from different perspectives for the life cycle of medical products. Through efficient data extraction technology, the node information acquisition module can build a comprehensive map of the medical product chain. Next, the entire medical product chain is traversed using advanced algorithms. During this process, the module uses graph traversal techniques such as depth-first search (DFS) or breadth-first search (BFS) to ensure that each node is systematically visited and recorded. The application of this method ensures that every key node is covered without omission, from raw material supply to the sale of the final product. Finally, the data collected during the traversal process is analyzed to identify key product nodes. For example, if the raw material records of a batch show an abnormal trend in the data analysis, the production and distribution nodes related to the batch will be marked as key nodes and require further monitoring. In this way, the node information acquisition module not only enhances the ability to see through the medical product chain, but also provides data support for risk assessment and the formulation of preventive measures.

[0015] In summary, the node information acquisition module significantly improves the efficiency and effectiveness of the medical safety monitoring platform by systematically retrieving, traversing and determining multiple nodes in the entire chain of medical products, providing a solid data foundation for timely identification and response to potential adverse reactions.

[0016] The real-time collection module 12 is used to collect data of the entire chain of medical products in real time based on the multiple product nodes, and formulate a data collection cycle according to the real-time collection results.

[0017] Optionally, the main responsibility of the real-time acquisition module is to collect real-time data from each product node in the entire chain and dynamically adjust the data collection cycle based on the collected data results. This module ensures real-time data updating and optimization of monitoring efficiency through advanced data acquisition technology and intelligent analysis algorithms.

[0018] During the implementation process, real-time data collection is first performed on each product node that has been determined by the node information acquisition module. This includes collecting data from multiple touchpoints such as sensors on the production line, logistics tracking systems, and consumer feedback. Each node is set with specific data monitoring requirements. For example, in the production stage, key monitoring parameters may include production rate, raw material usage, and quality control data; while in the distribution stage, the focus may be on product circulation speed and storage conditions. The determination of data monitoring priorities depends on the results of preliminary risk assessment, where high-risk nodes or data indicators such as abnormal product return rates or unqualified test results will be marked as key monitoring objects. In addition, the frequency of data occurrence is also a key factor in formulating the collection cycle. For example, if the data of a product node shows abnormal high volatility, such as a sudden increase in the production defect rate, the real-time collection module will increase the data collection frequency of the node in order to more closely monitor possible quality decline trends. The formulation of the data collection cycle is based on the continuous analysis of real-time collected data. The module uses algorithms to automatically analyze data change trends and patterns to determine the optimal data collection cycle for each node. This cycle may be adjusted according to different stages of the product life cycle or changes in market feedback. For example, the early stages of a new product introduction may require more frequent data updates to quickly capture any teething problems or market acceptance issues.

[0019] The real-time acquisition module effectively supports the continuous safety monitoring of the entire chain of medical products through precise real-time data monitoring and intelligent adjustment of data acquisition cycles. This not only improves the response speed to potential adverse reactions, but also optimizes resource allocation, ensuring that data collection activities are both efficient and cost-effective, thereby ensuring the safety of medical products while maintaining the health and sustainable development of the entire medical ecosystem.

[0020] The monitoring sensing module 13 is used to perform data monitoring sensing on the multiple product nodes according to the data collection cycle to generate multiple product monitoring data.

[0021] Exemplarily, the monitoring sensor module performs detailed data monitoring on multiple product nodes in the whole chain according to the set data collection cycle. This module uses high-precision sensors and advanced data collection technology to ensure that real-time and accurate monitoring data is obtained from each key node, where the contact refers to the data collection source of multiple parties. First, the monitoring sensor module sets and deploys the sensor network according to the data collection cycle obtained from the real-time acquisition module. The sensor deployment involves installing sensor equipment at key nodes of the whole chain, such as installing temperature and humidity sensors on the production line, or installing location and environmental monitors in the logistics center. These sensors automatically collect data according to a preset frequency, such as recording once an hour or providing instant feedback based on environmental changes. Then, the data monitoring phase begins. In this phase, the monitoring sensor module collects data on product quality, environmental conditions, operating procedures, etc. through its networked sensors. In particular, at high-risk nodes, such as quality control points of raw materials or product testing stages, monitoring will be more intensive to ensure that any signs that may cause adverse reactions are captured in a timely manner. In addition, the monitoring sensor module also includes data verification and preprocessing functions to ensure the quality and reliability of monitoring data. Data preprocessing includes steps such as denoising, normalization, and outlier detection, aiming to improve the accuracy of data analysis. For example, if the temperature recorded by the sensor suddenly jumps to an unreasonably high value, the system will automatically detect this anomaly and may trigger re-collection. Finally, the monitoring sensor module compiles all valid monitoring data into product monitoring data reports and transmits them to the data recognition module for further analysis and processing. These reports include not only the original data, but also the pre-processed data summary, providing a solid data foundation for subsequent risk assessment and early warning generation.

[0022] Through this systematic and efficient monitoring strategy, the monitoring sensor module significantly improves the entire platform's ability to monitor adverse reactions of medical products throughout the entire chain, ensuring that any potential safety issues can be responded to quickly and accurately. The design of this module demonstrates the ability to achieve efficient data monitoring in complex systems and is one of the key technologies to ensure the quality of medical products and patient safety.

[0023] The data identification module 14 is used to identify and integrate the multiple product monitoring data according to the multiple product data sources to determine the adverse reaction data set.

[0024] Furthermore, the data identification module is the core analysis unit of the full-chain adverse reaction data monitoring integration and early warning platform for medical products. It is responsible for integrating the collected multiple product monitoring data with multiple product data sources, and then accurately identifying and determining the adverse reaction data set. Through data analysis technology and machine learning algorithms, this module ensures that key information can be identified from large amounts of data, thereby improving the accuracy and timeliness of adverse reaction early warnings.

[0025] First, the data integration process begins by connecting the data collected by the monitoring sensor module with other existing product data sources, which may include historical quality control records, customer service feedback, and other relevant production and sales data. By using data integration techniques such as data fusion and assimilation, the data identification module is able to create a comprehensive data view to provide a complete data context for deeper analysis. Next, feature extraction is performed. In this stage, the module uses various data mining techniques, such as cluster analysis and principal component analysis, to extract key features from the integrated data that help identify adverse reactions. For example, by analyzing the relationship between product return rates and production batches, production lines or time periods that may have quality problems can be identified. The module then performs anomaly detection, using algorithms such as isolation forests or neural networks to identify data points that deviate from normal patterns. This step is critical because it is directly related to whether potential adverse reaction risks can be discovered in a timely manner. For example, if a product node shows temperature fluctuations that are significantly different from historical data, this may indicate equipment failure or operating errors, which require further investigation. Ultimately, the process of determining the adverse reaction dataset involves integrating all identified abnormal features and data points into a dedicated dataset that will be used to generate early warning information and conduct risk assessment. This dataset not only includes direct abnormal indicators, but also integrates relevant information extracted from the data source, such as the scope of impact, potential severity, etc.

[0026] The data identification module provides the platform with a highly accurate and real-time updated adverse reaction monitoring foundation, greatly improving the efficiency and response speed of medical product safety monitoring. The module ensures that medical institutions can respond quickly and effectively to potential risks.

[0027] The risk assessment module 15 is used to perform risk assessment on the adverse reaction data set, generate warning information according to the assessment results, and publish the warning information to the entire chain through the warning platform.

[0028] Specifically, the risk assessment module is responsible for conducting in-depth risk assessments on the adverse reaction data set obtained from the data identification module, and generating targeted early warning information based on this, which is ultimately widely released to the entire medical product chain through the early warning platform, thereby achieving efficient risk management and control.

[0029] First, when the risk assessment process is initiated, the module analyzes the adverse reaction data set using statistical analysis, machine learning algorithms, and rule engines. During this process, the module considers various risk factors, such as product quality fluctuations, the severity and frequency of adverse events. For example, if a particular batch of products shows continuous quality problems, the batch will be assessed as a high risk level. Further, risk classification is a key step in this stage, which involves classifying the identified risks into preset risk levels, such as low risk, medium risk, and high risk, based on the potential impact and probability of occurrence of the risk. For example, a small-scale deviation that occurs occasionally during the production process may be classified as low risk, while a serious adverse event that is widely reported will be classified as high risk. Next, the generation of warning information involves converting the risk assessment results into specific warning notices, which detail the nature of the adverse reaction, the associated risk level, the recommended response measures, and any necessary product recall information. During this information generation process, the risk assessment module will apply text generation technology to ensure the accuracy and understandability of the information. Finally, the release of early warning information will be carried out through an integrated early warning platform, which can ensure that all relevant parties, including manufacturers, distributors, medical institutions and end users, can receive early warning information in real time. The dissemination mechanisms used by this platform may include automatic email systems, SMS alerts and dedicated online early warning dashboards, so that all parties can take quick measures to prevent or mitigate the impact of adverse reactions.

[0030] The risk assessment module not only enhances the accuracy and timeliness of medical product safety monitoring, but also provides an efficient and reliable risk management and emergency response platform for all stakeholders in the medical industry chain. The design of this module greatly enhances the safety assurance capabilities of the entire medical industry.

[0031] Furthermore, multiple product data sources are retrieved to determine the entire chain of medical products, and multiple product nodes are determined based on the entire chain of medical products. The node information acquisition module 11 is also used to:

[0032] Data is extracted based on the multiple product data sources to obtain a product data set.

[0033] A graph data structure is drawn according to the multiple product data sources, and the product data set is added to the graph data structure to construct a full-chain model.

[0034] The full chain model is traversed and marked using a graph traversal algorithm to determine the multiple product nodes.

[0035] Specifically, data can be extracted from multiple product data sources, a graph data structure can be constructed, and a graph traversal algorithm can be applied to accurately mark and identify each product node, thereby achieving an in-depth understanding and monitoring of the entire product chain. First, the data extraction step involves extracting key data from various data sources, which may include but are not limited to production records, quality control reports, logistics information, and consumer feedback. Through advanced data extraction techniques, the platform is able to obtain detailed data about medical products from production to consumption, including specific information such as timestamps and operation results, which are key data for evaluating product quality and safety. Next, the graph data structure construction process begins, in which all extracted product data sets are integrated and mapped into a graph data structure. Graph data structures provide an efficient way to represent and handle complex relationships and dependencies, especially in a full-chain environment. For example, production batches may be nodes, and logistics paths can be represented as edges between nodes, which can intuitively show the flow process of products from manufacturing to end users. Finally, the application of graph traversal algorithms is a key step in determining product nodes. Using graph traversal algorithms such as depth-first search (DFS) or breadth-first search (BFS), the platform can systematically access every node in the graph, thereby marking key nodes and possible risk points in the entire chain. During the traversal process, the data of each node, such as timestamps and operation results, are recorded and analyzed in detail to identify patterns or anomalies that may lead to adverse reactions.

[0036] Through this approach, the medical product full-chain adverse reaction data monitoring integration and early warning platform can provide a comprehensive, data-driven view for monitoring and warning of medical product safety risks. This integration and intelligent processing not only improves the efficiency of risk management, but also ensures the safety and reliability of medical products in the market, ultimately protecting the health and safety of consumers.

[0037] Furthermore, data monitoring and sensing are performed on the multiple product nodes according to the data collection cycle to generate multiple product monitoring data. The monitoring sensing module 13 is also used to:

[0038] The multiple product nodes are traversed to match with multiple sensor devices, and a sensor network is deployed for the entire chain according to the matching results.

[0039] The plurality of product nodes are monitored and analyzed through the sensor network to determine a plurality of data monitoring features.

[0040] A plurality of data tables are defined according to the plurality of data monitoring features to construct a sensor monitoring database.

[0041] The multiple product nodes are monitored and sensed in real time according to the data collection cycle in combination with the sensor network to generate multiple real-time monitoring data.

[0042] The sensor monitoring database is stored and updated based on the multiple real-time monitoring data to obtain a sensor monitoring optimization database.

[0043] The plurality of product monitoring data are extracted according to the sensor monitoring optimization database.

[0044] Optionally, the product nodes of the entire chain are monitored in real time through advanced data monitoring sensor systems to ensure data accuracy and timely response capabilities. The platform effectively builds and optimizes the sensor monitoring database to achieve continuous monitoring of medical product safety.

[0045] First, the platform matches the sensor equipment with the product node. This step is completed by traversing each product node on the entire chain and matching it with the corresponding sensor equipment. This matching ensures that the monitoring requirements of each node are correctly met according to its characteristics, such as temperature-sensitive or vibration-sensitive nodes are equipped with corresponding sensors. Next, in the process of laying out the sensor network, a sensor network covering the entire chain is constructed based on the matching results. This network not only connects all key product nodes, but also ensures the continuity and integrity of the data, making the monitoring process both efficient and comprehensive. The platform further monitors and analyzes each node through the deployed sensor network to determine the data monitoring characteristics. These characteristics may include but are not limited to physical or chemical parameters such as temperature, humidity, and pressure of the product. These are key indicators for determining whether the product is in normal operation. Then, multiple data tables are defined based on the determined data monitoring characteristics, and a sensor monitoring database is constructed accordingly. These data tables not only systematically record the real-time monitoring data of each node, but also include timestamps and geographic location information, making data analysis more accurate and traceable. Next, according to the set data collection cycle, the platform monitors each node in real time through the sensor network and generates real-time monitoring data. These real-time data are continuously uploaded to the sensor monitoring database to keep the data up to date. Finally, the platform continuously updates the sensor monitoring database based on the real-time monitoring data, thereby optimizing the database performance and data quality. The optimized database will serve as the sensor monitoring optimization database, which supports the extraction of monitoring data from the database, which is the basis for risk assessment and generating early warning information.

[0046] Through the above steps, the platform ensures that every link of medical products from production to users is under monitoring, greatly improving the speed and accuracy of response to potential adverse reactions. The design of this system demonstrates comprehensive consideration of the safety monitoring of the entire chain of medical products.

[0047] Furthermore, a plurality of data tables are defined according to the plurality of data monitoring features to construct a sensor monitoring database, and the monitoring sensor module 13 is further used for:

[0048] Feature classification is performed based on the multiple data monitoring features to obtain multiple feature classes.

[0049] The characteristic fields of the plurality of characteristic classes are extracted and combined with the plurality of common fields to define a plurality of data tables, and the plurality of data tables correspond to the characteristic fields of the plurality of characteristic classes.

[0050] Perform correlation analysis on the multiple data tables to determine multiple correlation coefficients.

[0051] The plurality of data tables are associated according to the plurality of correlation coefficients to construct the sensor monitoring database.

[0052] Exemplarily, the platform performs feature classification based on multiple data monitoring features. This process involves systematically classifying various types of data collected from the sensor network according to their properties, such as physical features such as temperature, humidity, and pressure, as well as more complex chemical and biological features. Each type of feature not only defines the specific content of the monitoring, but also determines the priority of data processing and analysis. Further, the platform extracts the characteristic fields of the feature class and combines them with multiple common fields to define multiple data tables. These common fields include timestamps (recording the specific collection time of the data), product IDs (identifying the specific product nodes being monitored), and monitoring point IDs (identifying the specific monitoring locations or sensors). Such a design not only improves the logic of data organization, but also facilitates subsequent data query and analysis. Subsequently, the platform performs association analysis on these data tables to determine multiple association coefficients between tables. The purpose of association analysis is to understand the relationship between different data sets, such as how the same product ID is represented in different data tables, which helps to reveal the performance of the product in the entire chain and possible problem points. The determination of the association coefficient is based on the logical and actual relationship between the data, and reasonable table relationships are designed, such as one-to-one, one-to-many, or many-to-many relationships. Finally, the sensor monitoring database is constructed based on the determined correlation coefficients. This step involves systematically connecting all data tables according to their correlation to form a comprehensive and interconnected database architecture, which not only improves the efficiency of data query, but also makes cross-table comprehensive analysis possible, thereby enabling more comprehensive monitoring and evaluation of the safety status of medical products.

[0053] Through the above steps, complex data from the entire chain can be efficiently processed and analyzed, providing solid data support for the safety monitoring of medical products. This highly structured and systematic data management method not only improves the accuracy of data processing, but also enhances the platform's response capabilities to potential risks.

[0054] Furthermore, based on the multiple real-time monitoring data, the sensor monitoring database is stored and updated to obtain a sensor monitoring optimization database, and the monitoring sensor module 13 is also used for:

[0055] The sensor monitoring database is traversed according to the data collection cycle to perform data timeliness evaluation and generate a data timeliness result, wherein the data timeliness result is a plurality of valid data and / or a plurality of invalid data.

[0056] When the data aging result is only the multiple valid data, a storage adding instruction is generated.

[0057] The sensor monitoring database is updated by traversing the plurality of real-time monitoring data through the storage adding instruction to obtain the sensor monitoring optimization database.

[0058] Furthermore, the platform regularly updates its sensor monitoring database to ensure that the collected data always reflects the latest monitoring conditions, thereby improving the efficiency and accuracy of medical product safety monitoring.

[0059] First, the platform traverses the sensor monitoring database according to the predetermined data collection cycle to evaluate the timeliness of the data. During this process, the system automatically checks the data in the database to determine whether they are still valid or have expired. Valid data refers to data that still meets the current monitoring requirements and is consistent with the real-time situation, while invalid data may no longer be accurate or relevant due to a long time or changes in conditions. Next, if the evaluation results show that all data are valid, the system will generate a storage addition instruction, which is used to guide the database management system to integrate new real-time monitoring data during the traversal process. Specifically, this includes adding new information in the real-time monitoring data (such as new monitoring point data or updated parameter values) to the database to ensure the integrity and update of the database content. After that, the platform will execute the storage addition instruction to update the data in the sensor monitoring database through these instructions. This update process involves comparing the collected real-time monitoring data with the existing data in the database, screening out new data that has not yet been recorded in the database, and integrating this data into the database. It not only includes the addition of data, but may also involve the modification or replacement of old data to reflect the latest monitoring results. Finally, the platform obtained a sensor monitoring optimization database, which provides a more accurate and timely data view, allowing the platform to timely and accurately evaluate the performance of medical products in the entire chain, and promptly detect and respond to possible adverse reactions. This system design not only ensures the accuracy and practicality of the data, but also greatly improves the response speed and efficiency of the medical product monitoring system. In this way, the platform can effectively manage and utilize a large amount of data, providing solid technical support for medical safety.

[0060] Furthermore, the monitoring sensor module 13 is also used for:

[0061] When the data aging result is only the multiple invalid data, or the multiple valid data and the multiple invalid data, a data cleaning instruction is generated.

[0062] The multiple failure data are identified for impact analysis, and are sorted in descending order according to the impact analysis results to generate a data impact sequence.

[0063] The plurality of failure data are sorted in ascending order according to the failure time sequence to generate a data failure sequence.

[0064] The data impact sequence and the data failure sequence are integrated to formulate a cleanup task.

[0065] The data cleaning instruction is executed on the sensor monitoring database based on the cleaning task in combination with the multiple real-time monitoring data to obtain the sensor monitoring optimization database.

[0066] Optionally, when the platform identifies that there is invalid data in the data timeliness evaluation results, it will automatically generate data cleaning instructions. The generation of these instructions is based on real-time monitoring of the data in the current database to ensure that all processing is performed on the latest data status. Then, the platform will perform an impact analysis on these invalid data. During this process, the system analyzes the impact of each invalid data on the overall database and data analysis results, and sorts them in descending order according to the size of the impact to form a data impact sequence. In this sequence, the data with the least impact is ranked first, which means that the removal of these data has the least impact on the overall performance of the system. At the same time, the invalid data will also be sorted in ascending order according to their expiration time to generate a data failure sequence. This sequence is arranged in the order of data failure, so that the data with the longest expiration time is ranked first and is considered for cleaning first. After that, the platform will integrate the data impact sequence and the data failure sequence to formulate the final data cleaning task. This step is critical because it takes into account the dual factors of the degree of data failure and the impact on the system, ensuring the efficiency and rationality of the cleaning work. Finally, according to the established cleaning tasks, the platform will execute data cleaning instructions and update the sensor monitoring database in combination with real-time monitoring data, which not only involves deleting or updating invalid data, but also includes necessary data migration and archiving work to maintain the accuracy and integrity of the database. Through these steps, the platform effectively obtains an optimized sensor monitoring database that more accurately reflects the current monitoring situation and the safety status of medical products.

[0067] Through this comprehensive data management strategy, the platform is able to maintain the cutting-edge and efficiency of its data processing capabilities, ensuring the platform's high performance and reliability in medical safety monitoring. These steps are key to ensuring data quality and system stability.

[0068] Furthermore, if Figure 2 As shown, the multiple product monitoring data are identified and integrated according to the multiple product data sources to determine the adverse reaction data set. The data identification module 14 is also used to:

[0069] The plurality of product monitoring data are cleansed to obtain a plurality of product monitoring standard data.

[0070] Anomalies are identified based on the multiple product monitoring standard data in combination with the multiple product data sources to determine multiple abnormal data.

[0071] The plurality of product data sources are traversed to perform mapping analysis on the plurality of abnormal data, and the analysis results are verified, and when the verification passes, a plurality of abnormal features are extracted.

[0072] An association rule mining algorithm is used to combine the multiple abnormal features to perform data mining on the multiple product monitoring standard data to construct a data association network.

[0073] The multiple abnormal data are integrated according to the data association network to obtain the adverse reaction data set.

[0074] Exemplarily, the platform first performs data cleaning, which is the basis of data processing and aims to remove incomplete, erroneous or irrelevant data from multiple product monitoring data to obtain standardized product monitoring data, including normalizing data formats, processing missing values, filtering out noise, etc., to ensure the data quality and consistency of subsequent steps. Then, the platform performs anomaly identification based on the cleaned data. By analyzing and comparing multiple product monitoring standard data with product data sources, it can identify abnormal data that does not conform to the normal pattern. These anomalies may indicate potential adverse reactions or other important health risks. Subsequently, the identified abnormal data is mapped and analyzed. This process involves associating the abnormal data with relevant records from different data sources (such as patient medication records and adverse reaction records) to determine the logical relationship and consistency between the data. Furthermore, the analysis results are verified, including checking whether the logical relationship between the data is reasonable and whether there are contradictions. The verification of this step ensures the accuracy of the data analysis, eliminates accidental errors or data contradictions, and extracts multiple abnormal features when the verification passes. Afterwards, the platform further mines abnormal features and product monitoring data using association rule mining algorithms. This process builds a data association network that reveals the potential connections between different drugs and between drugs and adverse reactions, which is the key to understanding and preventing adverse reactions. Finally, based on the constructed data association network, the platform integrates all relevant abnormal data to form a comprehensive adverse reaction data set, which provides valuable information for medical safety monitoring and supports medical institutions and regulatory agencies to take preventive or response measures.

[0075] Through the above steps, the platform not only improves the accuracy and efficiency of data processing, but also ensures the scientificity and systematicness of medical product safety monitoring. The application of these technologies has greatly enhanced the application value and effect of the platform in actual medical environments, providing strong technical support for protecting public health.

[0076] Furthermore, the risk assessment is performed on the adverse reaction data set, warning information is generated according to the assessment result, and the warning information is released to the whole chain through the warning platform. The risk assessment module 15 is also used for:

[0077] The product is compared with the adverse reaction data set and the adverse reaction ratio data is calculated.

[0078] Determine whether the adverse reaction ratio data is within the expected threshold range.

[0079] If the adverse reaction ratio data is not within the expected threshold interval, an imbalance label is generated.

[0080] The adverse reaction data set is risk assessed according to the imbalance label, and a preset risk level is matched according to the assessment result to generate an adverse reaction risk level.

[0081] The release scope is determined based on the adverse reaction risk level, and the warning information is generated.

[0082] The warning information is synchronized to the warning platform, and the warning information is released to the entire chain according to the release scope through the warning platform.

[0083] Specifically, the platform calculates the adverse reaction ratio data, that is, compares various medical products with the corresponding adverse reaction data sets to calculate the adverse reaction ratio of each product. This ratio is determined by comparing the frequency of use of the same product within a specific period of time with the number of adverse events reported, aiming to quantify the safety performance of each product. Next, the platform needs to determine whether the adverse reaction ratio data is within the expected threshold range. This evaluation standard is based on the pre-set safety standard. If the adverse reaction ratio of a product exceeds this threshold, it is considered that there is a potential risk problem; if the adverse reaction ratio does not exceed the threshold, the product is considered to be within an acceptable safety range. If the adverse reaction ratio is found to exceed the threshold, the platform will generate an imbalance label. This imbalance label is a warning mark to indicate that there is a proportional imbalance between a product and its adverse reactions, which may point to a risk point that is not fully recognized. Subsequently, risk assessment is performed based on the imbalance label. At this stage, the adverse reaction data set marked with the imbalance label is deeply analyzed to match the preset risk level, which is set based on factors such as the severity of the adverse reaction, the frequency of occurrence and its potential impact. Once the risk level is determined, the platform generates early warning information based on the adverse reaction risk level. These early warning information details the name of the relevant drug, the specific description of the adverse reaction, the risk level, and the specific measures recommended for each level of risk. Finally, based on the risk assessment results and the importance of the early warning information, the platform determines the release scope and releases the early warning information through the early warning platform. The release scope may include medical institutions, drug manufacturers, drug regulatory authorities, and the public, etc., to ensure that all relevant parties can receive important safety information in a timely manner and take appropriate preventive or response measures.

[0084] Through the above steps, the platform can effectively monitor, evaluate and report possible medical product safety risks, greatly enhancing the timeliness and effectiveness of medical safety supervision and providing strong technical support for protecting public health.

[0085] Through the technical solutions of the above-mentioned embodiments, the medical product full-chain adverse reaction data monitoring integration and early warning platform provided by the present application solves the technical problems of incomplete data integration, delayed monitoring and inaccurate risk assessment in the prior art in the medical product adverse reaction data monitoring. By realizing full-chain monitoring, real-time data collection and adjustment, and integrated data processing and risk assessment, the coverage, timeliness and accuracy of monitoring are improved, thereby effectively improving the technical effect of improving the prevention and response capabilities of adverse reactions of medical products.

[0086] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A full-chain adverse reaction data monitoring integration and early warning platform for medical products, characterized by: The platform includes: A node information acquisition module, used to retrieve multiple product data sources to determine the entire chain of medical products, and traverse the entire chain of medical products to determine multiple product nodes; A real-time collection module, used for collecting data of the entire chain of medical products in real time based on the multiple product nodes, and formulating a data collection cycle according to the real-time collection results; A monitoring and sensing module, used for performing data monitoring and sensing on the multiple product nodes according to the data collection cycle to generate multiple product monitoring data; A data identification module, used to identify and integrate the plurality of product monitoring data according to the plurality of product data sources to determine an adverse reaction data set; A risk assessment module, used to perform risk assessment on the adverse reaction data set, generate early warning information according to the assessment results, and publish the early warning information to the entire chain through the early warning platform; Wherein, the monitoring sensor module is also used for: Traversing the plurality of product nodes and matching them with the plurality of sensor devices, and deploying a sensor network for the entire chain according to the matching results; Monitoring and analyzing the plurality of product nodes through the sensor network to determine a plurality of data monitoring features; Defining a plurality of data tables according to the plurality of data monitoring features, and constructing a sensor monitoring database; Performing real-time monitoring and sensing of the multiple product nodes according to the data collection cycle in combination with the sensor network to generate multiple real-time monitoring data; Based on the multiple real-time monitoring data, the sensor monitoring database is stored and updated to obtain a sensor monitoring optimization database; Extracting the plurality of product monitoring data according to the sensor monitoring optimization database; wherein, the sensor monitoring database is traversed according to the data collection cycle to perform data timeliness evaluation and generate a data timeliness result, wherein the data timeliness result is a plurality of valid data and / or a plurality of invalid data; When the data aging result is only the plurality of valid data, a storage adding instruction is generated; The sensor monitoring database is updated by traversing the plurality of real-time monitoring data through the storage adding instruction to obtain the sensor monitoring optimization database; Wherein, when the data aging result is only the multiple invalid data, or the multiple valid data and the multiple invalid data, a data cleaning instruction is generated; Identify the multiple failure data to perform impact analysis, sort the data in descending order according to the impact analysis results, and generate a data impact sequence, in which the data with the least impact is arranged in front; Sorting the multiple failure data in ascending order according to the failure time sequence to generate a data failure sequence, wherein the sequence is arranged according to the time sequence of data failure so that the data with the longest failure time is arranged in front; Integrate the data impact sequence with the data failure sequence to formulate a cleanup task; The data cleaning instruction is executed on the sensor monitoring database based on the cleaning task in combination with the multiple real-time monitoring data to obtain the sensor monitoring optimization database.

2. The medical product full-chain adverse reaction data monitoring integration and early warning platform as claimed in claim 1, characterized in that: The node information acquisition module is also used for: Extracting data based on the multiple product data sources to obtain a product data set; Draw a graph data structure according to the multiple product data sources, and add the product data set to the graph data structure to build a full-chain model; The full chain model is traversed and marked using a graph traversal algorithm to determine the multiple product nodes.

3. The medical product full-chain adverse reaction data monitoring integration and early warning platform as claimed in claim 1, characterized in that: The monitoring sensor module is also used for: Performing feature classification based on the multiple data monitoring features to obtain multiple feature classes; Extracting the characteristic fields of the plurality of characteristic classes and combining them with the plurality of common fields to define a plurality of data tables, wherein the plurality of data tables correspond to the characteristic fields of the plurality of characteristic classes; Performing correlation analysis on the multiple data tables to determine multiple correlation coefficients; The plurality of data tables are associated according to the plurality of correlation coefficients to construct the sensor monitoring database.

4. The medical product full-chain adverse reaction data monitoring integration and early warning platform as claimed in claim 1, characterized in that: The data identification module is also used for: Performing data cleaning on the plurality of product monitoring data to obtain a plurality of product monitoring standard data; Performing abnormality identification based on the plurality of product monitoring standard data in combination with the plurality of product data sources to determine a plurality of abnormal data; Traversing the multiple product data sources to perform mapping analysis on the multiple abnormal data, verifying the analysis results, and extracting multiple abnormal features when the verification passes; Using an association rule mining algorithm in combination with the multiple abnormal features to perform data mining on the multiple product monitoring standard data, and construct a data association network; The multiple abnormal data are integrated according to the data association network to obtain the adverse reaction data set.

5. The medical product full-chain adverse reaction data monitoring integration and early warning platform as claimed in claim 1, characterized in that: The risk assessment module is also used to: Compare the product with the adverse reaction data set and calculate adverse reaction ratio data; Determining whether the adverse reaction ratio data is within an expected threshold range; If the adverse reaction ratio data is not within the expected threshold range, an imbalance label is generated; Performing risk assessment on the adverse reaction data set according to the imbalance label, matching a preset risk level according to the assessment result, and generating an adverse reaction risk level; Determine the release scope based on the adverse reaction risk level and generate the warning information; The warning information is synchronized to the warning platform, and the warning information is released to the entire chain according to the release scope through the warning platform.

Images