A large model-based data management system and method

Abstract

This invention discloses a data management system and method based on a large model, belonging to the field of big data analytics. By setting a data monitoring and management cycle, the system collects information on changes in the inventory of various drugs in hospitals. It combines this with time-series model analysis of historical departmental outbound data to accurately calculate the first outbound prediction value. Furthermore, it compares the average daily dose with the standard daily dose, deeply mining the correlation data of compatible drugs in auxiliary medication scenarios, analyzing the outbound trends of compatible drugs, and accurately grasping the patterns of synergistic consumption. By tracking the changes in the outbound ratio of target drug data and compatible drug data over historical periods and predicting the ratio trend, it dynamically calculates the demand for auxiliary drugs by combining the predicted outbound volume with compatible drug data. This deep deconstruction of inventory change data enables forward-looking prediction of inventory changes, helping hospitals optimize their inventory structure and improve the accuracy and responsiveness of inventory data management.

Description

Technical Field

[0001] This invention relates to the field of big data analytics, specifically a data management system and method based on a large model. Background Technology

[0002] Data management is a core support in the digital age. By relying on technologies such as big data and artificial intelligence to collect, clean, and deeply analyze multi-source data, it can uncover the potential patterns and value of data, providing a scientific basis for enterprise operation optimization and industrial upgrading. In the medical field, it can also provide basic support by managing drug data and medical records. Overall, it plays an irreplaceable and crucial role in promoting efficiency improvement and ensuring accurate decision-making in various fields.

[0003] In the medical field, drug usage data across departments is significantly influenced by factors such as differences in treatment plans, patient conditions, and clinical medication habits, exhibiting marked spatiotemporal dynamics. Some drug data demonstrate multi-faceted roles: when used as a primary treatment, the data shows high-frequency consumption; when used as an adjunct therapy, the frequency of use decreases sharply. In adjunct therapy scenarios, drug data further differentiates into two categories: one corresponding to one-time adjunctive needs with small doses and short cycles, exhibiting instantaneous data; the other due to intermittent use resulting from alternating treatment phases, exhibiting intermittent data. The data for intermittent use, due to its dispersed distribution and weak regularity, makes it difficult for traditional statistical analysis methods to accurately capture inventory consumption trends. This ultimately leads to insufficient accuracy in drug consumption prediction and a lag in data-driven inventory adjustment mechanisms, resulting in an imbalance between drug inventory data and actual clinical needs. Consequently, problems arise such as a single dimension of drug reserve data analysis and insufficient real-time data. Summary of the Invention

[0004] The purpose of this invention is to provide a data management system and method based on a large model to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a data management method based on a large model, the data management method comprising the following steps:

[0006] Step S1: Set a data monitoring and management cycle, obtain the outbound data of various types of drugs in the hospital within the data monitoring and management cycle, select any one type of drug among the various types of drugs as the research object, and record it as the target drug data; analyze the outbound data of the target drug data of each department of the hospital to obtain the first outbound prediction value;

[0007] Step S1-1: By acquiring the hospital's various drug outbound records, departmental requisition records, and inventory change data during the data monitoring and management period, the data is standardized and classified according to drug name, dosage form, and specifications to form a basic drug inventory database. The basic drug inventory database is a set of structured data tables constructed based on the hospital's various drug outbound records, inbound records, and inventory change data, which is used to store and manage drug inbound and outbound information.

[0008] Step S1-2: Analyze and extract the outbound quantity of target drugs in each department of the hospital in each historical data monitoring and management cycle, and obtain the outbound quantity in the next data monitoring and management cycle through time series prediction model;

[0009] The calculation formula for the time series forecasting model is as follows:

[0010]

[0011] In the formula, E i，t+1 E represents the predicted outbound quantity of the target drug for the i-th department in period t+1; i，1 To E i，t TSModel represents the actual outbound quantity of the i-th department from the first data monitoring and management period to the t-th data monitoring and management period; TSModel represents the time series forecasting model.

[0012] Step S1-3: Summarize the outbound quantities predicted by each department of the hospital for the next data monitoring and management cycle, and obtain the total outbound quantity of the target drug data for the next data monitoring and management cycle, which is recorded as the first outbound prediction value.

[0013] By setting data monitoring and management cycles, building a basic database of inventory data, and using time series forecasting models to analyze historical outbound data of departments to obtain the first outbound forecast value of target drug data, it is possible to systematically integrate drug inbound and outbound information, accurately predict drug demand, provide data support for hospitals to rationally arrange drug procurement and inventory management, effectively avoid drug backlog or shortage, and improve drug management efficiency and resource utilization.

[0014] Step S2: Obtain the outbound data of medicines from various departments of the hospital, analyze and extract the record data of the target medicine when it is used as an auxiliary drug in the data monitoring and management cycle, analyze and extract the data of the compatible drugs when the target medicine is used as an auxiliary drug; analyze the outbound data of the compatible drugs in various departments of the hospital to obtain the outbound quantity of each compatible drug in the next data monitoring and management cycle.

[0015] Step S2-1: Determine the usage status of the target drug data. The specific process is as follows: when the actual average daily dose of the target drug data is lower than the standard limit daily dose, it is determined to be an auxiliary drug; when the actual average daily dose of the target drug data is equal to the standard limit daily dose, it is determined to be a primary drug.

[0016] Step S2-2: Obtain the record data of medicines in various departments of the hospital. The record data includes the quantity of medicines issued, the details of the compatibility of medicines issued in the same batch, the issuance time and department; the details of the compatibility of medicines issued include the name and quantity of medicines.

[0017] Step S2-3: Based on the recorded data of the target drug data during the data monitoring and management cycle when it is used as an auxiliary drug, analyze and extract other drugs that are compatible with the target drug data, and record them as the compatible drug data;

[0018] Step S2-4: Based on the drug names of each compatible drug data, predict the outbound quantity of each compatible drug data in the next data monitoring and management cycle by executing step S1-2.

[0019] By determining the usage attributes of target drug data, extracting the data of drugs used in combination with auxiliary drugs from departmental drug record data, and then analyzing the outbound data of drugs used in combination based on the target drug data prediction method, we can accurately grasp the usage relationship and demand trend of auxiliary drugs and their combination drugs. This helps hospitals optimize drug combination management, rationally allocate resources, ensure the synergy and effectiveness of clinical drug use, and improve the level of precision in drug management.

[0020] Step S3: Analyze the changes in the proportion of target drug data as compatibility drug data in each historical data monitoring and management cycle to obtain the proportion of target drug data in the next data monitoring and management cycle, which is recorded as the predicted proportion. Calculate the quantity of target drug data to be used as an auxiliary drug in the next data monitoring and management cycle by using the quantity of compatibility drug data shipped out and the predicted proportion.

[0021] Step S3-1: Extract the outbound quantity of target drug data and compatible drug data in each cycle, select any one type of compatible drug data as the research object, and denot it as target compatible drug data. Calculate the proportion of target drug data in the total outbound quantity of target compatible drug data in each data monitoring and management cycle.

[0022] Step S3-2: By combining the percentage values ​​calculated in multiple historical steps S3-1 with the time series prediction model, the percentage of the target drug data in the next data monitoring and management cycle is obtained and recorded as the predicted percentage.

[0023] Step S3-3: Simultaneously use the time series forecasting model to predict the outbound quantity of the target compatible drugs in the next data monitoring and management cycle;

[0024] Step S3-4: Multiply the predicted outbound quantity of the target compatible drug data by the predicted proportion to obtain the outbound quantity of the target drug data as an auxiliary drug in the next data monitoring and management cycle.

[0025] By extracting historical outbound quantities of target drugs and compatible drugs, calculating the proportion, and using a time series forecasting model to obtain the predicted proportion, and combining the predicted outbound quantities of compatible drugs to calculate the outbound quantities of auxiliary drugs for the target drug data, we can deeply explore the patterns of drug compatibility and use, scientifically predict the demand for auxiliary drugs for the target drug data, facilitate hospitals to plan their procurement plans in advance, optimize inventory structure, and improve the foresight and accuracy of drug management.

[0026] Step S4: Summarize the outbound quantities of the target drug data when it is used as the compatibility drug data to obtain the total outbound quantity of the target drug data when it is used as an auxiliary drug, and record it as the auxiliary total outbound quantity; Simultaneously sum the outbound quantities of the target drug data in each department of the hospital to obtain the total outbound quantity of the target drug data when it is used as the main drug, and record it as the main outbound quantity.

[0027] Step S4-1: Process the data of each compatible drug by executing step S3 to obtain the target drug data as multiple outbound quantities when used as an auxiliary drug;

[0028] Step S4-2: Sum the multiple outbound quantities obtained from the drug compatibility data to obtain the total outbound quantity of the target drug when it is used as an auxiliary drug, and record it as the auxiliary total outbound quantity.

[0029] Step S4-3: Obtain the outbound quantity of target drugs in each hospital department when they are the main drugs used in multiple historical data monitoring and management cycles. Simultaneously execute step S1-2 to obtain the outbound quantity of target drugs in each hospital department when they are the main drugs used in the next data monitoring and management cycle.

[0030] Step S4-4: When the target drug data in each hospital department is used as the main drug, the outbound quantity in the next data monitoring and management cycle is summarized and summed to obtain the total outbound quantity of the target drug data as the main drug, which is recorded as the main outbound quantity.

[0031] By integrating the analysis results of drug compatibility data, the total outbound quantity of target drugs can be obtained. Combined with departmental data and predictive models, the total outbound quantity of primary drugs can be calculated. This allows for a comprehensive and accurate classification of the demand scale of target drugs in different usage scenarios, helping hospitals to clearly understand the drug usage structure. This provides a reliable basis for scientifically formulating procurement strategies and optimizing resource allocation, effectively balancing the supply of primary and auxiliary drugs, and improving the overall efficiency of drug management.

[0032] Step S5: Analyze and calculate the second outbound prediction value of the target drug data based on the total outbound quantity of auxiliary drugs and the total outbound quantity of primary drugs. Combine the first outbound prediction value with the weighted fusion calculation to obtain the final outbound quantity of the target drug data. Monitor and manage the target drug data based on the final outbound quantity.

[0033] Step S5-1: Sum the auxiliary total outbound quantity of the target drug data with the main total outbound quantity to obtain the second outbound prediction value of the target drug data; assign preset weight parameters to the first outbound prediction value and the second outbound prediction value respectively; according to the weight parameters, perform weighted fusion calculation on the first outbound prediction value and the second outbound prediction value to obtain the final outbound quantity of the target drug data.

[0034] The final outbound quantity of the target drug data is calculated using the following formula:

[0035] E all =E1×w1+E2×w2;

[0036] In the formula, E all E1 represents the final outbound quantity of the target drug data; w1 represents the first outbound forecast value of the target drug data in the next data monitoring and management cycle obtained through step S1-2; w1 represents the weighting coefficient of the first outbound forecast value; E2 represents the second outbound forecast value of the target drug data in the next data monitoring and management cycle; w2 represents the weighting coefficient of the second outbound forecast value; satisfying w1+w2=1.

[0037] Step S5-2: Select the drug name of the target drug data as the index, obtain the inventory quantity of the target drug data in the drug inventory basic database within the current data monitoring and management period, and use the inventory quantity of the target drug data and the final outbound quantity for monitoring and management. The specific process is as follows:

[0038] When the inventory quantity exceeds the final outbound quantity, it is determined that the remaining inventory of the target drug data meets medical needs.

[0039] When the inventory quantity does not exceed the final outbound quantity, it is determined that the inventory balance of the target drug data does not meet medical needs, and an inventory data management early warning signal is issued.

[0040] By integrating outbound data from auxiliary and primary scenarios, a second predicted value is obtained. This value is then weighted and fused with a first predicted value based on departmental data to arrive at the final outbound quantity. Combined with inventory data, surplus monitoring is achieved. This allows for multi-dimensional and accurate prediction of drug demand and scientific assessment of inventory status. It avoids deviations caused by a single prediction dimension and responds promptly to inventory shortages through an early warning mechanism. This provides refined decision support for hospital drug procurement and inventory adjustment, ensuring clinical drug needs while improving management efficiency.

[0041] Furthermore, a data management system based on a large model is provided, which includes a basic data processing module, an initial outbound forecasting module, a real-time inventory analysis module, an auxiliary outbound forecasting module, and an inventory monitoring and decision-making module.

[0042] The basic data processing module is used to collect drug data, standardize and classify it to build a database, and extract historical data. The initial outbound prediction module is used to analyze historical data through a time series prediction model to obtain the outbound prediction value for the target drug data in the next data monitoring and management cycle. The real-time inventory analysis module is used to determine the usage of the target drug data, extract the data of compatible drugs, and predict the outbound quantity. The auxiliary outbound prediction module is used to calculate the proportion of the target drug data in the data of compatible drugs and predict the outbound quantity of auxiliary drugs. The inventory monitoring and decision-making module is used to integrate various prediction values ​​to obtain the final outbound quantity, and monitor the inventory and issue early warnings based on the final outbound quantity.

[0043] The output of the basic data processing module is electrically connected to the input of the initial outbound forecast module; the output of the initial outbound forecast module is electrically connected to the input of the real-time inventory analysis module; the output of the real-time inventory analysis module is electrically connected to the input of the auxiliary outbound forecast module; and the output of the auxiliary outbound forecast module is electrically connected to the input of the inventory monitoring and decision module.

[0044] The basic data processing module includes a data integration unit and a historical data extraction unit; the data integration unit is used to integrate drug outbound, inbound, and inventory change data, and construct a basic inventory data database after standardized processing; the historical data extraction unit is used to extract historical target drug outbound data from the basic drug inventory database for each department.

[0045] The initial outbound prediction module includes a department prediction unit and a total quantity summary unit. The department prediction unit is used to analyze the historical outbound data of target drugs in each department and predict the outbound quantity of target drugs in each department in the next data monitoring and management cycle. The total quantity summary unit is used to summarize the predicted outbound quantities of each department and obtain the total outbound quantity of target drugs in the next data monitoring and management cycle.

[0046] The real-time inventory analysis module includes a data-assisted judgment unit and a compatibility data processing unit. The data-assisted judgment unit is used to compare the actual daily dosage of the target drug with the standard average daily dosage to determine the usage status of the target drug. The compatibility data processing unit is used to acquire drug record data, extract auxiliary drug compatibility data, and predict the outbound quantity.

[0047] The auxiliary outbound prediction module includes a percentage calculation unit and an auxiliary quantity prediction unit. The percentage calculation unit is used to calculate the proportion of the target drug data to the total outbound volume of the target compatible drug data in each period. The auxiliary quantity prediction unit is used to obtain the predicted percentage through a time series model, combine it with the compatible drug data to predict the outbound volume, and calculate the outbound quantity of the target drug data when it is used as an auxiliary drug.

[0048] The inventory monitoring and decision-making module includes a predicted value fusion unit and an inventory data management and early warning unit. The predicted value fusion unit is used to calculate the sum of the auxiliary and main outbound quantities, and weighted fusion to obtain the final outbound quantity. The inventory data management and early warning unit is used to compare the inventory quantity with the final outbound quantity, determine the inventory balance, and issue an early warning signal.

[0049] Compared with the prior art, the beneficial effects of the present invention are:

[0050] 1. This invention, by setting a data monitoring and management cycle, collects various medical-related data such as hospital drug outbound records, departmental requisition records, and inventory change data. It then standardizes and stores these data according to drug name, dosage form, and specifications to construct a basic inventory data database. Finally, it combines this data with a time series forecasting model to analyze historical target drug outbound data from each department, accurately calculating the first predicted outbound value for the target drug. This achieves systematic integration and standardized management of medical data, effectively solving the problems of scattered medical data and insufficient forecasting accuracy in existing technologies. It provides a quantitative basis for hospital drug procurement planning and dynamic inventory data management, avoiding drug resource waste and supply shortages.

[0051] 2. This invention establishes a medical data comparison mechanism between the actual daily average dose of the target drug and the standard-limited daily dose, scientifically defining its primary or auxiliary drug attributes. Based on medical record data within the data monitoring and management cycle, it deeply mines the correlation between the target drug data and the combined drug data in the auxiliary drug scenario. By using the same prediction model to analyze the outbound trend of the combined drug data, it fills the gap in the insufficient medical data correlation mining in the existing technology, accurately grasps the synergistic consumption pattern of auxiliary drugs and combined drug data, and provides refined decision support for clinical drug combination data management and medical resource allocation.

[0052] 3. This invention extracts medical data on the outbound quantities of target drugs and compatible drugs within historical data monitoring and management cycles, calculates the proportion of target drugs in the total outbound quantity of compatible drugs, uses a time series prediction model to predict the trend of this proportion in the next cycle, and dynamically calculates the outbound quantity of auxiliary drugs based on the predicted outbound quantity of compatible drugs. This deeply deconstructs the implicit patterns in medical data regarding drug compatibility, solves the problem of delayed prediction of auxiliary drug demand in existing technologies, and achieves forward-looking prediction of auxiliary drug demand driven by medical data. This helps hospitals optimize their inventory data structure and improve the accuracy and response efficiency of medical data management. Attached Figure Description

[0053] Figure 1 This is a flowchart illustrating a data management method based on a large model according to the present invention.

[0054] Figure 2 This is a schematic diagram of the structure of a data management system based on a large model according to the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Example 1: As Figure 1 As shown, the present invention provides a technical solution, a data management method based on a large model, the data management method comprising the following steps:

[0057] Step S1: Set a data monitoring and management cycle, obtain the outbound data of various types of drugs in the hospital within the data monitoring and management cycle, select any one type of drug among the various types of drugs as the research object, and record it as the target drug data; analyze the outbound data of the target drug data of each department of the hospital to obtain the first outbound prediction value;

[0058] Step S1-1: By acquiring the hospital's various drug outbound records, departmental requisition records, and inventory change data during the data monitoring and management period, the data is standardized and classified according to drug name, dosage form, and specifications to form a basic drug inventory database. The basic drug inventory database is a set of structured data tables constructed based on the hospital's various drug outbound records, inbound records, and inventory change data, which is used to store and manage drug inbound and outbound information.

[0059] Step S1-2: Analyze and extract the outbound quantity of target drugs in each department of the hospital in each historical data monitoring and management cycle, and obtain the outbound quantity in the next data monitoring and management cycle through time series prediction model;

[0060] Step S1-3: Summarize the outbound quantities predicted by each department of the hospital for the next data monitoring and management cycle, and obtain the total outbound quantity of the target drug data for the next data monitoring and management cycle, which is recorded as the first outbound prediction value.

[0061] In practical implementation, taking commonly used hospital drug A as an example, the monitoring cycle is set to monthly. The hospital information system collects the drug's outbound records, departmental usage data, and inventory change information from each department within that cycle, and stores them in a standardized database according to drug name, dosage form, and specifications. Time series models, such as the ARIMA model, are used to analyze the periodicity and trend of historical outbound data, and the first forecast value is formed by summarizing departmental data. Data collection must cover the entire cycle. If cross-month data is involved, dosage form and specifications must be distinguished during standardized classification to avoid prediction deviations due to incorrect data classification.

[0062] Step S2: Obtain the outbound data of medicines from various departments of the hospital, analyze and extract the record data of the target medicine when it is used as an auxiliary drug in the data monitoring and management cycle, analyze and extract the data of the compatible drugs when the target medicine is used as an auxiliary drug; analyze the outbound data of the compatible drugs in various departments of the hospital to obtain the outbound quantity of each compatible drug in the next data monitoring and management cycle.

[0063] Step S2-1: Determine the usage status of the target drug data. The specific process is as follows: when the actual average daily dose of the target drug data is lower than the standard limit daily dose, it is determined to be an auxiliary drug; when the actual average daily dose of the target drug data is equal to the standard limit daily dose, it is determined to be a primary drug.

[0064] Step S2-2: Obtain the record data of medicines in various departments of the hospital. The record data includes the quantity of medicines issued, the details of the compatibility of medicines issued in the same batch, the issuance time and department; the details of the compatibility of medicines issued include the name and quantity of medicines.

[0065] Step S2-3: Based on the recorded data of the target drug data during the data monitoring and management cycle when it is used as an auxiliary drug, analyze and extract other drugs that are compatible with the target drug data, and record them as the compatible drug data;

[0066] Step S2-4: Based on the drug names of each compatible drug data, predict the outbound quantity of each compatible drug data in the next data monitoring and management cycle by executing step S1-2.

[0067] In practical implementation, taking drug A as an example, when the actual daily average dosage used by a department is lower than its standard daily dosage limit, it is determined to be an auxiliary drug. By extracting the compatibility details from the same batch of outbound records of the department, data on drugs A that were outbound at the same time as a certain drug are mined. Based on the dosage threshold, the drug attributes are defined, and the compatibility relationship is extracted through correlation analysis of the same batch of outbound data. Then, the same time series model is used to predict the outbound volume of the compatibility drug data. The standard daily dosage limit is set with reference to industry specifications, and the outbound conditions of the same batch are limited when extracting compatibility data to avoid the mixing of drug data from different periods.

[0068] Step S3: Analyze the changes in the proportion of target drug data as compatibility drug data in each historical data monitoring and management cycle to obtain the proportion of target drug data in the next data monitoring and management cycle, which is recorded as the predicted proportion. Calculate the quantity of target drug data to be used as an auxiliary drug in the next data monitoring and management cycle by using the quantity of compatibility drug data shipped out and the predicted proportion.

[0069] Step S3-1: Extract the outbound quantity of target drug data and compatible drug data in each cycle, select any one type of compatible drug data as the research object, and denot it as target compatible drug data. Calculate the proportion of target drug data in the total outbound quantity of target compatible drug data in each data monitoring and management cycle.

[0070] Step S3-2: By combining the percentage values ​​calculated in multiple historical steps S3-1 with the time series prediction model, the percentage of the target drug data in the next data monitoring and management cycle is obtained and recorded as the predicted percentage.

[0071] Step S3-3: Simultaneously use the time series forecasting model to predict the outbound quantity of the target compatible drugs in the next data monitoring and management cycle;

[0072] Step S3-4: Multiply the predicted outbound quantity of the target compatible drug data by the predicted proportion to obtain the outbound quantity of the target drug data as an auxiliary drug in the next data monitoring and management cycle.

[0073] In practice, the outbound percentage of drug A and its combination drug B in historical data monitoring and management cycles is analyzed. For example, the percentages for the past six months are 15%, 18%, and 12%, respectively. The percentage for the next data monitoring and management cycle is predicted using a time series model, and the outbound quantity of combination drug B is predicted simultaneously. Multiplying the two results gives the outbound quantity of a certain drug as an auxiliary drug. By utilizing the time-series pattern of drug combination usage percentages, the demand for the target drug is inferred from the usage of combination drug data. Drugs with stable clinical compatibility relationships are selected as the analysis objects to avoid the impact of percentage fluctuations caused by temporary combinations on prediction accuracy.

[0074] Step S4: Summarize the outbound quantities of the target drug data when it is used as the compatibility drug data to obtain the total outbound quantity of the target drug data when it is used as an auxiliary drug, and record it as the auxiliary total outbound quantity; Simultaneously sum the outbound quantities of the target drug data in each department of the hospital to obtain the total outbound quantity of the target drug data when it is used as the main drug, and record it as the main outbound quantity.

[0075] Step S4-1: Process the data of each compatible drug by executing step S3 to obtain the target drug data as multiple outbound quantities when used as an auxiliary drug;

[0076] Step S4-2: Sum the multiple outbound quantities obtained from the drug compatibility data to obtain the total outbound quantity of the target drug when it is used as an auxiliary drug, and record it as the auxiliary total outbound quantity.

[0077] Step S4-3: Obtain the outbound quantity of target drugs in each hospital department when they are the main drugs used in multiple historical data monitoring and management cycles. Simultaneously execute step S1-2 to obtain the outbound quantity of target drugs in each hospital department when they are the main drugs used in the next data monitoring and management cycle.

[0078] Step S4-4: When the target drug data in each hospital department is used as the main drug, the outbound quantity in the next data monitoring and management cycle is summarized and summed to obtain the total outbound quantity of the target drug data as the main drug, which is recorded as the main outbound quantity.

[0079] In practice, the auxiliary outbound quantities calculated from the data of drug A and all compatible drugs are summarized, and the outbound quantities of drug A as the main drug in each department are also summarized. By classifying and summarizing, the demand structure of drugs in main and auxiliary scenarios is clarified, providing data support for differentiated supply strategies.

[0080] Step S5: Analyze and calculate the second outbound prediction value of the target drug data based on the total outbound quantity of auxiliary drugs and the total outbound quantity of primary drugs. Combine the first outbound prediction value with the weighted fusion calculation to obtain the final outbound quantity of the target drug data. Monitor and manage the target drug data based on the final outbound quantity.

[0081] Step S5-1: Sum the auxiliary total outbound quantity of the target drug data with the main total outbound quantity to obtain the second outbound prediction value of the target drug data; assign preset weight parameters to the first outbound prediction value and the second outbound prediction value respectively; according to the weight parameters, perform weighted fusion calculation on the first outbound prediction value and the second outbound prediction value to obtain the final outbound quantity of the target drug data.

[0082] Step S5-2: Select the drug name of the target drug data as the index, obtain the inventory quantity of the target drug data in the drug inventory basic database within the current data monitoring and management period, and use the inventory quantity of the target drug data and the final outbound quantity for monitoring and management. The specific process is as follows:

[0083] When the inventory quantity exceeds the final outbound quantity, it is determined that the remaining inventory of the target drug data meets medical needs.

[0084] When the inventory quantity does not exceed the final outbound quantity, it is determined that the inventory balance of the target drug data does not meet medical needs, and an inventory data management early warning signal is issued.

[0085] In practical implementation, weight parameters are set for the first predicted value based on departmental data and the second predicted value based on the fusion of auxiliary and primary scenarios. The weighted calculation yields the final outbound predicted value, which is then compared with the current inventory data to trigger an early warning. The error of a single model is reduced by weighting the predicted values ​​from multiple dimensions, and dynamic monitoring is achieved by combining real-time inventory. The weight parameters need to be iteratively optimized regularly based on historical prediction accuracy, and the inventory data needs to be synchronized with the hospital's HIS system in real time to ensure the timeliness of the early warning mechanism.

[0086] Example 2, as Figure 2 As shown, the present invention provides a data management system based on a large model, which includes a basic data processing module, an initial outbound forecasting module, a real-time inventory analysis module, an auxiliary outbound forecasting module, and an inventory monitoring and decision-making module.

[0087] The basic data processing module is used to collect drug data, standardize and classify it to build a database, and extract historical data. The initial outbound prediction module is used to analyze historical data through a time series prediction model to obtain the outbound prediction value for the target drug data in the next data monitoring and management cycle. The real-time inventory analysis module is used to determine the usage of the target drug data, extract the data of compatible drugs, and predict the outbound quantity. The auxiliary outbound prediction module is used to calculate the proportion of the target drug data in the data of compatible drugs and predict the outbound quantity of auxiliary drugs. The inventory monitoring and decision-making module is used to integrate various prediction values ​​to obtain the final outbound quantity, and monitor the inventory and issue early warnings based on the final outbound quantity.

[0088] The output of the basic data processing module is electrically connected to the input of the initial outbound forecast module; the output of the initial outbound forecast module is electrically connected to the input of the real-time inventory analysis module; the output of the real-time inventory analysis module is electrically connected to the input of the auxiliary outbound forecast module; and the output of the auxiliary outbound forecast module is electrically connected to the input of the inventory monitoring and decision module.

[0089] The basic data processing module includes a data integration unit and a historical data extraction unit; the data integration unit is used to integrate drug outbound, inbound, and inventory change data, and construct a basic inventory data database after standardized processing; the historical data extraction unit is used to extract historical target drug outbound data from the basic drug inventory database for each department.

[0090] The initial outbound prediction module includes a department prediction unit and a total quantity summary unit. The department prediction unit is used to analyze the historical outbound data of target drugs in each department and predict the outbound quantity of target drugs in each department in the next data monitoring and management cycle. The total quantity summary unit is used to summarize the predicted outbound quantities of each department and obtain the total outbound quantity of target drugs in the next data monitoring and management cycle.

[0091] The real-time inventory analysis module includes a data-assisted judgment unit and a compatibility data processing unit. The data-assisted judgment unit is used to compare the actual daily dosage of the target drug with the standard average daily dosage to determine the usage status of the target drug. The compatibility data processing unit is used to acquire drug record data, extract auxiliary drug compatibility data, and predict the outbound quantity.

[0092] The auxiliary outbound prediction module includes a percentage calculation unit and an auxiliary quantity prediction unit. The percentage calculation unit is used to calculate the proportion of the target drug data to the total outbound volume of the target compatible drug data in each period. The auxiliary quantity prediction unit is used to obtain the predicted percentage through a time series model, combine it with the compatible drug data to predict the outbound volume, and calculate the outbound quantity of the target drug data when it is used as an auxiliary drug.

[0093] The inventory monitoring and decision-making module includes a predicted value fusion unit and an inventory data management and early warning unit. The predicted value fusion unit is used to calculate the sum of the auxiliary and main outbound quantities, and weighted fusion to obtain the final outbound quantity. The inventory data management and early warning unit is used to compare the inventory quantity with the final outbound quantity, determine the inventory balance, and issue an early warning signal.

[0094] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A data management method based on a large model, characterized in that: The data management method includes the following steps: Step S1: Set a data monitoring and management cycle, obtain the outbound data of various types of drugs in the hospital within the data monitoring and management cycle, select any one type of drug among the various types of drugs as the research object, and record it as the target drug data; analyze the outbound data of the target drug data of each department of the hospital to obtain the first outbound prediction value; Step S2: Obtain the outbound data of medicines from various departments of the hospital, analyze and extract the record data of the target medicine when it is used as an auxiliary drug in the data monitoring and management cycle, analyze and extract the data of the compatible drugs when the target medicine is used as an auxiliary drug; analyze the outbound data of the compatible drugs in various departments of the hospital to obtain the outbound quantity of each compatible drug in the next data monitoring and management cycle. Step S3: Analyze the changes in the proportion of target drug data as compatibility drug data in each historical data monitoring and management cycle to obtain the proportion of target drug data in the next data monitoring and management cycle, which is recorded as the predicted proportion. Calculate the quantity of target drug data to be used as an auxiliary drug in the next data monitoring and management cycle by using the quantity of compatibility drug data shipped out and the predicted proportion. Step S4: Summarize the outbound quantities of the target drug data when it is used as the compatibility drug data to obtain the total outbound quantity of the target drug data when it is used as an auxiliary drug, and record it as the auxiliary total outbound quantity; Simultaneously sum the outbound quantities of the target drug data in each department of the hospital to obtain the total outbound quantity of the target drug data when it is used as the main drug, and record it as the main outbound quantity. Step S5: Analyze and calculate the second outbound prediction value of the target drug data based on the total outbound quantity of auxiliary drugs and the total outbound quantity of primary drugs. Combine the first outbound prediction value with the weighted fusion calculation to obtain the final outbound quantity of the target drug data. Monitor and manage the target drug data based on the final outbound quantity. The specific steps of step S5 are as follows: Step S5-1: Sum the auxiliary total outbound quantity of the target drug data with the main total outbound quantity to obtain the second outbound prediction value of the target drug data; assign preset weight parameters to the first outbound prediction value and the second outbound prediction value respectively; according to the weight parameters, perform weighted fusion calculation on the first outbound prediction value and the second outbound prediction value to obtain the final outbound quantity of the target drug data. Step S5-2: Select the drug name of the target drug data as the index, obtain the inventory quantity of the target drug data in the drug inventory basic database within the current data monitoring and management period, and use the inventory quantity of the target drug data and the final outbound quantity for monitoring and management. The specific process is as follows: When the inventory quantity exceeds the final outbound quantity, it is determined that the remaining inventory of the target drug data meets medical needs. When the inventory quantity does not exceed the final outbound quantity, it is determined that the remaining inventory of the target drug data does not meet medical needs, and an inventory data management early warning signal is issued.

2. The data management method based on a large model according to claim 1, characterized in that: The specific steps of step S1 are as follows: Step S1-1: By acquiring the hospital's various drug outbound records, departmental requisition records, and inventory change data during the data monitoring and management period, the data is standardized and classified according to drug name, dosage form, and specifications to form a basic drug inventory database. The basic drug inventory database is a set of structured data tables constructed based on the hospital's various drug outbound records, inbound records, and inventory change data, which is used to store and manage drug inbound and outbound information. Step S1-2: Analyze and extract the outbound quantity of target drugs in each department of the hospital in each historical data monitoring and management cycle, and obtain the outbound quantity in the next data monitoring and management cycle through time series prediction model; Steps S1-3: Summarize the outbound quantities predicted by each department of the hospital for the next data monitoring and management cycle, and obtain the total outbound quantity of the target drug data for the next data monitoring and management cycle, which is recorded as the first outbound prediction value.

3. The data management method based on a large model according to claim 2, characterized in that: The specific steps of step S2 are as follows: Step S2-1: Determine the usage status of the target drug data. The specific process is as follows: when the actual average daily dose of the target drug data is lower than the standard limit daily dose, it is determined to be an auxiliary drug; when the actual average daily dose of the target drug data is equal to the standard limit daily dose, it is determined to be a primary drug. Step S2-2: Obtain the record data of medicines in various departments of the hospital. The record data includes the quantity of medicines issued, the details of the compatibility of medicines issued in the same batch, the issuance time and department; the details of the compatibility of medicines issued include the name and quantity of medicines. Step S2-3: Based on the recorded data of the target drug data during the data monitoring and management cycle when it is used as an auxiliary drug, analyze and extract other drugs that are compatible with the target drug data, and record them as the compatible drug data; Step S2-4: Based on the drug names of each compatible drug data, predict the outbound quantity of each compatible drug data in the next data monitoring and management cycle by executing step S1-2.

4. The data management method based on a large model according to claim 3, characterized in that: The specific steps of step S3 are as follows: Step S3-1: Extract the outbound quantity of target drug data and compatible drug data in each cycle, select any one type of compatible drug data as the research object, and denot it as target compatible drug data. Calculate the proportion of target drug data in the total outbound quantity of target compatible drug data in each data monitoring and management cycle. Step S3-2: By combining the percentage values ​​calculated in multiple historical steps S3-1 using the time series prediction model, the percentage of the target drug data in the next data monitoring and management cycle is obtained and recorded as the predicted percentage.

5. The data management method based on a large model according to claim 4, characterized in that: Step S3 also includes: Step S3-3: Simultaneously use the time series forecasting model to predict the outbound quantity of the target compatible drugs in the next data monitoring and management cycle; Step S3-4: Multiply the predicted outbound quantity of the target compatible drug data by the predicted proportion to obtain the outbound quantity of the target drug data as an auxiliary drug in the next data monitoring and management cycle.

6. The data management method based on a large model according to claim 5, characterized in that: The specific steps of step S4 are as follows: Step S4-1: Process the data of each compatible drug by executing step S3 to obtain the target drug data as multiple outbound quantities when used as an auxiliary drug; Step S4-2: Sum the multiple outbound quantities obtained from the drug compatibility data to obtain the total outbound quantity of the target drug when it is used as an auxiliary drug, and record it as the auxiliary total outbound quantity. Step S4-3: Obtain the outbound quantity of target drugs in each hospital department when they are the main drugs used in multiple historical data monitoring and management cycles. Simultaneously execute step S1-2 to obtain the outbound quantity of target drugs in each hospital department when they are the main drugs used in the next data monitoring and management cycle. Step S4-4: When the target drug data in each hospital department is used as the main drug, the outbound quantity in the next data monitoring and management cycle is summarized and summed to obtain the total outbound quantity of the target drug data as the main drug, which is recorded as the main outbound quantity.

7. A data management system based on a large model, applied to the data management method based on a large model as described in any one of claims 1-6, characterized in that: The data management system includes a basic data processing module, an initial outbound forecasting module, a real-time inventory analysis module, an auxiliary outbound forecasting module, and an inventory monitoring and decision-making module. The basic data processing module is used to collect drug data, standardize and classify it to build a database, and extract historical data; the initial outbound prediction module is used to analyze historical data through a time series prediction model to obtain the outbound prediction value of the target drug data in the next data monitoring and management cycle; the real-time inventory analysis module is used to determine the usage of the target drug data, extract the compatible drug data, and predict the outbound quantity; the auxiliary outbound prediction module is used to calculate the proportion of the target drug data in the compatible drug data and predict the outbound quantity of auxiliary drugs. The inventory monitoring and decision-making module is used to integrate various forecast values ​​to calculate the final outbound quantity, and to monitor and issue warnings based on the final outbound quantity; The output of the basic data processing module is electrically connected to the input of the initial outbound prediction module; The output of the initial outbound forecast module is electrically connected to the input of the real-time inventory analysis module; the output of the real-time inventory analysis module is electrically connected to the input of the auxiliary outbound forecast module; and the output of the auxiliary outbound forecast module is electrically connected to the input of the inventory monitoring and decision module.

8. A data management system based on a large model according to claim 7, characterized in that: The basic data processing module includes a data integration unit and a historical data extraction unit; the data integration unit is used to integrate drug outbound, inbound, and inventory change data, and construct a basic inventory data database after standardized processing; the historical data extraction unit is used to extract historical target drug outbound data from the basic drug inventory database for each department. The initial outbound prediction module includes a department prediction unit and a total quantity summary unit. The department prediction unit is used to analyze the historical outbound data of target drugs in each department and predict the outbound quantity of target drugs in each department in the next data monitoring and management cycle. The total quantity summary unit is used to summarize the predicted outbound quantities of each department and obtain the total outbound quantity of target drugs in the next data monitoring and management cycle. The real-time inventory analysis module includes a data-assisted judgment unit and a compatibility data processing unit. The data-assisted judgment unit is used to compare the actual daily dosage of the target drug with the standard daily average dosage to determine the usage status of the target drug. The compatibility data processing unit is used to acquire drug record data, extract auxiliary drug compatibility data, and predict the quantity to be shipped out.

9. A data management system based on a large model according to claim 7, characterized in that: The auxiliary outbound prediction module includes a percentage calculation unit and an auxiliary quantity prediction unit. The percentage calculation unit is used to calculate the proportion of the target drug data to the total outbound volume of the target compatible drug data in each period. The auxiliary quantity prediction unit is used to obtain the predicted percentage through a time series model, combine it with the compatible drug data to predict the outbound volume, and calculate the outbound quantity of the target drug data when it is used as an auxiliary drug. The inventory monitoring and decision-making module includes a forecast value fusion unit and an inventory data management and early warning unit; The predicted value fusion unit is used to calculate the sum of the auxiliary and main outbound quantities, and then weighted and fused to obtain the final outbound quantity. The inventory data management early warning unit is used to compare the inventory quantity with the final outbound quantity, determine the remaining inventory, and issue an early warning signal.

Images