A consumable supply chain real-time monitoring method and system based on multi-source data fusion
Patent Information
- Application Number
- CN202611215715.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-12
- Publication Date
- 2026-10-30
AI Technical Summary
这导致各药物之间的消耗并非彼此独立，而是存在因方案配伍关系而产生的用量耦合与时序依赖：当某一药物的实际消耗因病情变化、患者依从性波动或临时调方而出现偏离时，其关联药物的后续需求通常也会随之发生连锁变化，进而引发多品类耗材的同步供需波动
[0047]（1）本发明以药物为节点、以联合用药方案中的共现关系为有向边，构建了能够刻画药物间消耗传导关系的关联子网，并基于历史给药执行数据计算各有向边的协同响应系数和响应时延参数，使某一药物因方案调整或临床因素产生的消耗波动能够沿有向边方向逐级量化传播至关联药物，弥补了现有方法无法预判联合用药连锁需求变化的缺陷。
Smart Images
Figure CN123017398A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of multi-source data monitoring technology, specifically to a method and system for real-time monitoring of the consumable supply chain based on multi-source data fusion. Background Technology
[0002] In clinical drug use, combination therapy is a common treatment approach, especially in areas such as chemotherapy for cancer, anti-infective therapy, and comprehensive management of chronic diseases. Multiple drugs must be used in combination according to a predetermined regimen to achieve the desired therapeutic effect. This results in the consumption of each drug not being independent, but rather exhibiting dosage coupling and time dependence due to the compatibility of the regimens: when the actual consumption of a certain drug deviates due to changes in the patient's condition, fluctuations in patient compliance, or temporary prescription adjustments, the subsequent demand for its related drugs usually changes in a chain reaction, leading to synchronous supply and demand fluctuations in multiple categories of consumables. Meanwhile, the hospital consumables supply chain involves multiple data sources, including medical order systems, hospital information systems (HIS), and inventory management systems. Each system operates independently, with varying event timestamp accuracy and data update frequencies, making it difficult to perceive, in real time, the consumption changes resulting from combination therapy as a whole.
[0003] Existing methods for monitoring and scheduling hospital consumable supply chains mostly set safety stock thresholds for single drug categories or make independent demand forecasts based on historical consumption, failing to incorporate the correlations between drugs arising from combined medication regimens into the analytical framework. When facing combined medication scenarios, these methods typically only passively trigger replenishment or allocation operations after the actual consumption of each drug has deviated, unable to predict in advance changes in demand for related drugs caused by fluctuations in the consumption of a particular drug. Furthermore, because prescriptions, medication administration, and inventory inflows and outflows are recorded in different business systems, existing methods rely heavily on manual experience to retrospectively integrate multi-source data, making it difficult to achieve synchronous correlation analysis of consumption fluctuations and inventory changes under a unified time benchmark.
[0004] Therefore, there is an urgent need for a technical solution that can characterize the transmission relationship between drugs in combination and integrate multi-source heterogeneous data on a unified time axis for real-time analysis, in order to solve the problem of uneven consumption of consumables caused by combination drugs and the resulting imbalance in supply chain distribution, and provide a quantifiable basis for dynamic inventory adjustment. Summary of the Invention
[0005] The purpose of this invention is to provide a real-time monitoring method for the consumable supply chain based on multi-source data fusion, so as to solve the problems mentioned in the background art.
[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:
[0007] A real-time monitoring method for the consumable supply chain based on multi-source data fusion, comprising the following steps:
[0008] Step S1: Extract prescription records for each combination drug regimen within the historical period from the medical order interface. Construct the topological skeleton of the combination drug association subnet with drugs as nodes and co-occurrence relationships in the same combination drug regimen as edges. Calculate the directed weight coefficient and response delay parameter of each directed edge based on the temporal sequence and dosage ratio of drug nodes in each directed edge during the drug administration process.
[0009] Step S2: Construct a unified multi-source data alignment time axis, and map the effective time of the prescription interface, the single drug administration execution time of the hospital information interface, and the inbound and outbound event time of the supply chain inventory interface onto the alignment time axis. Divide the time axis into several synchronous monitoring windows with a fixed window width. Within each monitoring window, collect the consumable characteristic data of each drug. The consumable characteristic data includes the actual consumption amount and actual consumption rate of the drug.
[0010] Step S3: Within the monitoring window, identify the set of source drugs whose actual consumption rate fluctuates, input the fluctuation amplitude and direction of the source drugs into the combined drug association subnet, and calculate the synergistic consumption response prediction of each associated drug by propagating along the directed edge direction; based on the difference between the actual consumption of each drug and the synergistic consumption response prediction, calculate the supply and demand deviation of each drug in combination drug use.
[0011] Step S4: Construct a cross-category quota adjustment matrix. The off-diagonal elements in the cross-category quota adjustment matrix represent the quota transfer amount generated between related drugs due to the combined use relationship. The quota transfer amount is superimposed on the basic planned release amount of each drug in the next monitoring window to generate the corrected dynamic release quota and send it to the supply chain inventory interface.
[0012] Preferably, the topological skeleton of the combined drug use association subnet constructed in step S1 includes:
[0013] Extract all historical prescription records of combination drug regimens from the medical order interface. The prescription records include at least the combination drug regimen identifier code, the drug code of each drug in the combination drug regimen, the single dose of each drug, the dosing frequency of each drug, and the number of days of the combination drug regimen.
[0014] Construct a combined drug use association subnetwork with each drug as a node. ,in For drug node set, It is a set of directed edges;
[0015] For any two medications, if at least [a certain number of medications are listed in the historical prescription record] If they appear together in the same combination therapy regimen, then at the drug node and Establish directed edges between them;
[0016] During drug administration, if drug nodes The average time of drug administration is earlier than the drug delivery time. The corresponding average drug administration time is then used to establish a directed edge. Characterizing drug nodes The corresponding changes in drug consumption will be reflected in the drug node. Corresponding drug conduction direction; if drug node The average time of drug administration is earlier than the drug delivery time. The corresponding average drug administration time is then used to establish a directed edge. .
[0017] Furthermore, step S1 calculates the directed weight coefficients and response delay parameters for each directed edge, including:
[0018] For directed edges Calculate directed edges based on historical dosing data Cooperative response coefficient , as a directed edge Directed weight coefficients:
[0019] Extract all nodes containing drugs Corresponding drugs and drug nodes Record the treatment course of the same combination therapy regimen for the corresponding drugs, and calculate the drug nodes within each treatment course. The corresponding daily consumption and drug delivery points The ratio of the corresponding daily drug consumption, taken as the weighted average of the ratios within each treatment course, is used as the synergistic response coefficient. :
[0020] ;
[0021] in For drug nodes and drug nodes The total number of combined drug treatment courses, and The first Drug nodes in each course of treatment and drug nodes The average daily consumption;
[0022] For directed edges Calculate directed edges Response delay parameters The response delay parameter is taken as the value of the drug node within each treatment course. Corresponding drug single-dose timing and drug delivery time The average time difference between subsequent single doses of the corresponding drug.
[0023] Preferably, in step S2, the window is divided into several synchronous monitoring windows according to a fixed window width, including:
[0024] The earliest of the following three times is used as the zero point of the alignment time axis: the first time the prescription takes effect in the medical order interface, the first time the medication is administered in the hospital information interface, and the first time the inventory is recorded in the supply chain. A unified multi-source data alignment time axis is established.
[0025] Fixed window width Cut into sections along the aligned time axis A synchronized monitoring window will display any number of... Each monitoring window is recorded as ,in .
[0026] Furthermore, step S2 involves compiling consumable characteristic data for each drug, including:
[0027] In the One monitoring window Internally, the system extracts the sequential drug administration records from the monitoring window of the hospital information interface and summarizes them to obtain the drug node. Actual consumption Combined with fixed window width Calculate the actual consumption rate .
[0028] Preferably, step S3 involves progressively calculating the predicted synergistic consumption response of each associated drug along the directed edge direction, including:
[0029] For directed edges Combined with fixed window width , response delay parameter Convert to lag window number ,in It is a rounding function. Characterizing drug nodes Consumption fluctuations propagate to drug nodes The number of monitoring windows that need to be spanned;
[0030] In the One monitoring window Internally, regarding drug nodes Identify drug nodes Actual consumption rate Relative to monitoring window Internal planned consumption rate fluctuation range ,in Values are derived empirically based on historical consumption data from the same period.
[0031] Will satisfy Drug nodes are assigned to the source drug set. ,in The preset fluctuation perception threshold is determined based on empirical values derived from historical consumption data from the same period.
[0032] For drug nodes in the source drug set Along the combined drug use subnet Directed edges in To the primary associated drug node Propagation of collaborative consumption response, first-level associated drug nodes The predicted amount of collaborative consumption response is ,in For an indicator function, if and only if the monitored window number is... With lag window number The difference is greater than or equal to The time value is That is, when The response has now fallen into the current window;
[0033] If the first-level associated drug node Simultaneously serving as another directed edge The starting point points to the secondary related drug node. Then continue along the directed edge Propagation, secondary related drug nodes The predicted amount of collaborative consumption response is obtained by chain-based cumulative recursion using the collaborative consumption response prediction formula until all reachable drug nodes are traversed.
[0034] Furthermore, in step S3, based on the difference between the actual consumption of each drug and the predicted synergistic consumption response, the supply-demand deviation of the combined use of each drug is calculated, including:
[0035] After all transmissions are completed, in the first... One monitoring window Within, for each drug node Combined with actual consumption Computing drug nodes Demand deviation of combined drug use ,in For drug nodes The collaborative consumption response prediction quantity and If drug nodes Not targeted by any source drug and drug node If it does not belong to the source drug set, then , .
[0036] Preferably, step S4 involves constructing a cross-category quota adjustment matrix, including:
[0037] Based on the supply-demand deviation of combined drug use and the drug node set, a system is constructed. Cross-category quota adjustment matrix ,in Represents the set of drug nodes Number of drug nodes included; cross-category quota adjustment matrix off-diagonal elements Representation in directed edges Lower cause drug node With drug nodes The association between combined drug use and the drug node To drug nodes The amount of quota transfer is based on the collaborative response coefficient corresponding to the directed weight coefficient. ,get ,in The preset quota adjustment step size coefficient, In the first One monitoring window Intra-drug nodes Supply and demand deviation of combined drug use; cross-category quota adjustment matrix diagonal elements Characterizing drug nodes The net deviation, based on the first One monitoring window Intra-drug nodes Demand deviation of combined drug use ,get .
[0038] Furthermore, in step S4, the revised dynamic allocation quota is generated and distributed to the supply chain inventory interface, including...
[0039] For drug nodes In the One monitoring window Dynamic allocation quota for ,in Based on the basic plan's deployment volume, For combined drug use related sub-network In and drug nodes There exists a set of neighboring drugs connected by directed edges. Characterization from drug nodes To drug nodes Quota transfer volume, Indicates drug node When acting as a supplier with surplus, the drug nodes corresponding to the shortage neighbors are supplied with the drug. The sum of transferred quotas (making drug nodes) (Corresponding reduction in distribution quotas) Indicates drug node When acting as the party experiencing a supply shortage, the drug nodes corresponding to each neighboring party with a surplus are used for supply. The sum of received quotas (making the drug node) (Corresponding increase in allocation quotas); the two work together to replenish the quotas of drugs in short supply and reduce the quotas of drugs in surplus.
[0040] The revised dynamic allocation quota Issued to the supply chain inventory interface.
[0041] A real-time monitoring system for the consumable supply chain based on multi-source data fusion, comprising:
[0042] The first module is used to extract prescription records of various combined medication regimens within the historical period from the medical order interface. It constructs the topological skeleton of the combined medication association subnet with drugs as nodes and co-occurrence relationships in the same combined medication regimen as edges. Based on the temporal sequence and dosage ratio of drug nodes in each directed edge during the drug administration process, it calculates the directed weight coefficient and response delay parameter of each directed edge.
[0043] The second module is used to construct a unified multi-source data alignment time axis, which maps the effective time of the prescription interface, the single drug administration execution time of the hospital information interface, and the inbound and outbound event time of the supply chain inventory interface to the alignment time axis. It is divided into several synchronous monitoring windows with a fixed window width, and the consumable characteristic data of each drug is statistically analyzed in each monitoring window. The consumable characteristic data includes the actual consumption amount and actual consumption rate of the drug.
[0044] The third module is used to identify the set of source drugs whose actual consumption rate fluctuates within the monitoring window, input the fluctuation amplitude and direction of the source drugs into the combined drug association subnet, and calculate the synergistic consumption response prediction of each associated drug by propagating along the directed edge direction. Based on the difference between the actual consumption of each drug and the synergistic consumption response prediction, the combined drug supply and demand deviation of each drug is calculated.
[0045] The fourth module is used to construct a cross-category quota adjustment matrix. The off-diagonal elements in the cross-category quota adjustment matrix represent the quota transfer amount generated between related drugs due to the combination of drugs. The quota transfer amount is superimposed on the basic planned release amount of each drug in the next monitoring window to generate the corrected dynamic release quota and send it to the supply chain inventory interface.
[0046] The beneficial effects achieved by this invention are:
[0047] (1) This invention uses drugs as nodes and co-occurrence relationships in combination drug regimens as directed edges to construct an associated subnetwork that can characterize the consumption transmission relationship between drugs. Based on historical dosing data, the collaborative response coefficient and response delay parameter of each directed edge are calculated, so that the consumption fluctuation of a certain drug caused by regimen adjustment or clinical factors can be quantitatively propagated to associated drugs along the direction of the directed edge, which makes up for the deficiency of existing methods that cannot predict the chain demand changes of combination drug regimens.
[0048] (2) This invention realizes bidirectional quota transfer between multiple drugs through cross-category quota adjustment matrix. Shortage drugs receive quota supplements from surplus related drugs, and surplus drugs transfer quotas to shortage related drugs. The quantitative results of the propagation of combined drug use are directly converted into dynamic release quotas that can be executed by the supply chain inventory interface, so that the release amount of each drug is adjusted synchronously with the actual consumption of combined drugs, thereby alleviating the supply imbalance caused by uneven consumption rates among drugs. Attached Figure Description
[0049] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.
[0050] Figure 1 This is a schematic diagram illustrating the steps of a real-time monitoring method for the consumable supply chain based on multi-source data fusion according to the present invention. Detailed Implementation
[0051] 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.
[0052] Please see Figure 1 In this first embodiment: a method for real-time monitoring of the consumable supply chain based on multi-source data fusion is provided, the method including the following steps:
[0053] Step S1: Construction of the combined drug use association subnet and calculation of edge parameters;
[0054] First, extract the prescription records of each combination drug regimen within the historical period from the medical order interface; in this embodiment, the historical period is the past 12 months, and the prescription records include at least the combination drug regimen identifier code, the drug code of each drug in the regimen, the single dose of each drug, the dosing frequency of each drug, and the number of days of the regimen treatment course.
[0055] Construct a combined drug use association subnetwork with each drug as a node. ,in For drug node set, Let be a set of directed edges; for any two drugs, if there exists at least one in the historical prescription record... If two drugs appear together in the same combination therapy regimen, a candidate edge is established between them. In this embodiment... Take 5;
[0056] Then, the direction of the edges is determined based on the dosing execution records: the time interval from the effective date of the regimen to the first administration of each drug in all prescriptions containing it is counted, and the average value is taken as the typical dosing time for that drug; if the drug node Typical dosing time is earlier than drug delivery time. The typical dosing time is then established by a directed edge. , indicating drug node Changes in consumption will affect drug nodes. Directional transmission, and conversely, establishment. ;
[0057] For directed edges Calculate its directed weight coefficients Extract all nodes that simultaneously contain drug information. and drug nodes Records of combined drug therapy courses, total One course of treatment; for the first Each treatment course, calculating drug nodes. Daily consumption With drug nodes Daily consumption The ratio of the ratios for all treatment courses is averaged. , average Defined as the collaborative response coefficient, and used as a directed weighting coefficient;
[0058] At the same time, for directed edges Calculate directed edges Response delay parameters For each treatment cycle, the drug nodes At each dosing time, retrieve drug nodes within the treatment course. The time difference between the first administration following this dose and the given time is calculated. The average time difference for all eligible dosing pairs across all treatment courses is then used as the response delay parameter. This parameter is used to determine the consumption fluctuations from the drug node. Spread to drug nodes The required lag time.
[0059] Step S2: Construct a unified timeline, divide monitoring windows, and statistically analyze consumable characteristics;
[0060] A unified multi-source data alignment timeline is established using the earliest of the following three times: the first prescription effective time from the medical order interface, the first medication administration time from the hospital information interface, and the first inbound / outbound record time from the supply chain inventory interface. This timeline is then aligned with a fixed window width. Cut into sections along the aligned time axis The first synchronized monitoring window records the first... Each monitoring window is , All events from the interface are mapped to the corresponding window according to their original timestamps;
[0061] In each monitoring window Within this window, the system extracts the sequential dosing records of each drug from the hospital information interface and summarizes them to obtain the drug node. Actual consumption And calculate the actual consumption rate. At the same time, information on each effective combination drug regimen within the window is obtained from the medical order interface for determining the subsequent planned consumption rate.
[0062] Step S3: Identification of fluctuation sources, propagation of coordinated consumption response, and estimation of supply-demand deviation;
[0063] In the One monitoring window Internally, regarding drug nodes The planned consumption rate is determined based on the currently effective combination therapy regimen and historical data from the same period. (In this embodiment, the average of the actual consumption rate during the same period of the previous four weeks is taken) ), calculate the fluctuation range
[0064] like ,in To preset the fluctuation sensing threshold (which can be 0.2 in this embodiment), the drug node is... Included in the source drug set ;
[0065] For each drug node in the source drug set, the collaborative consumption response is propagated to the first-level associated nodes along the directed edges of the combined drug association subnet, converting the response delay into the number of lag windows. ;
[0066] Then the first-level associated node In the current window The predicted amount of collaborative consumption response is ,in For an indicator function, if and only if A value of 1 indicates that the wave has propagated beyond the response delay and fallen into the current window; if This indicates that the fluctuation has not yet propagated to the current window, and its value is 0.
[0067] If the first-level associated drug node Simultaneously serving as another directed edge The starting point points to the secondary related drug node. Then continue along the directed edge Propagation, at this point the predicted response of the second-level node is ;in This is the set of first-level associated nodes that have generated positive responses; and so on, until all reachable drug nodes are traversed.
[0068] After the entire transmission is completed, for each drug node Calculate the supply-demand deviation of its combined drug use. , If drug nodes If a variable is neither part of the source drug set nor has it received any source drug pointing to it via a directed edge, then let... , ; This indicates that actual consumption exceeds the response volume predicted by fluctuations in the source drug, potentially leading to a supply gap. This indicates that the actual consumption is lower than the predicted response, which may lead to a backlog.
[0069] Step S4: Constructing a cross-category quota adjustment matrix and generating dynamic quota allocation;
[0070] Constructing a system based on the supply-demand deviation of combined drug use Cross-category quota adjustment matrix diagonal elements Characterizing drug nodes Its own net deviation, off-diagonal elements (corresponding directed edges) Characterization of drug-related relationships due to combined drug use To drug nodes The amount of transferred quota, of which This is the preset quota adjustment step size coefficient;
[0071] For the next monitoring window Determine the basic planned dosage for each drug. (Pre-determined by the hospital's procurement department based on historical needs);
[0072] The revised dynamic allocation quota is ,in For combined drug use related sub-network In terms of drug nodes There exists a set of neighboring drugs connected by directed edges;
[0073] The revised dynamic allocation quota The information is sent to the supply chain inventory system through the supply chain inventory interface, serving as the basis for actual replenishment or allocation in the next monitoring window;
[0074] After that, steps S2 to S4 are repeated every time a monitoring window is passed, forming a closed-loop dynamic adjustment, so that the allocation quota of each drug is continuously adjusted according to the fluctuation of the actual consumption of combined drugs, thereby alleviating the problem of supply chain imbalance caused by uneven consumption rates among drugs.
[0075] In this second embodiment: a real-time monitoring system for the consumable supply chain based on multi-source data fusion is provided. This system includes:
[0076] The first module is used to extract prescription records of various combined medication regimens within the historical period from the medical order interface. It constructs the topological skeleton of the combined medication association subnet with drugs as nodes and co-occurrence relationships in the same combined medication regimen as edges. Based on the temporal sequence and dosage ratio of drug nodes in each directed edge during the drug administration process, it calculates the directed weight coefficient and response delay parameter of each directed edge.
[0077] The second module is used to construct a unified multi-source data alignment time axis, which maps the effective time of the prescription interface, the single drug administration execution time of the hospital information interface, and the inbound and outbound event time of the supply chain inventory interface to the alignment time axis. It is divided into several synchronous monitoring windows with a fixed window width, and the consumable characteristic data of each drug is statistically analyzed in each monitoring window. The consumable characteristic data includes the actual consumption amount and actual consumption rate of the drug.
[0078] The third module is used to identify the set of source drugs whose actual consumption rate fluctuates within the monitoring window, input the fluctuation amplitude and direction of the source drugs into the combined drug association subnet, and calculate the synergistic consumption response prediction of each associated drug by propagating along the directed edge direction. Based on the difference between the actual consumption of each drug and the synergistic consumption response prediction, the combined drug supply and demand deviation of each drug is calculated.
[0079] The fourth module is used to construct a cross-category quota adjustment matrix. The off-diagonal elements in the cross-category quota adjustment matrix represent the quota transfer amount generated between related drugs due to the combination of drugs. The quota transfer amount is superimposed on the basic planned release amount of each drug in the next monitoring window to generate the corrected dynamic release quota and send it to the supply chain inventory interface.
[0080] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. The method and specific examples of real-time monitoring of the consumable supply chain based on multi-source data fusion described in Embodiment 1 are also applicable to the real-time monitoring system of the consumable supply chain based on multi-source data fusion in this embodiment. Through the foregoing detailed description of the method of real-time monitoring of the consumable supply chain based on multi-source data fusion, those skilled in the art can clearly understand the real-time monitoring system of the consumable supply chain based on multi-source data fusion in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here. As for the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant details can be found in the method section.
[0081] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0082] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for real-time monitoring of the consumable supply chain based on multi-source data fusion, characterized in that, The method includes the following steps: Step S1: Extract prescription records for each combination drug regimen within the historical period from the medical order interface. Construct the topological skeleton of the combination drug association subnet with drugs as nodes and co-occurrence relationships in the same combination drug regimen as edges. Calculate the directed weight coefficient and response delay parameter of each directed edge based on the temporal sequence and dosage ratio of drug nodes in each directed edge during the drug administration process. Step S2: Construct a unified multi-source data alignment time axis, and map the effective time of the prescription interface, the single drug administration execution time of the hospital information interface, and the inbound and outbound event time of the supply chain inventory interface onto the alignment time axis. Divide the time axis into several synchronous monitoring windows with a fixed window width. Within each monitoring window, collect the consumable characteristic data of each drug. The consumable characteristic data includes the actual consumption amount and actual consumption rate of the drug. Step S3: Within the monitoring window, identify the set of source drugs whose actual consumption rate fluctuates, input the fluctuation amplitude and direction of the source drugs into the combined drug association subnet, and calculate the synergistic consumption response prediction of each associated drug by propagating along the directed edge direction; based on the difference between the actual consumption of each drug and the synergistic consumption response prediction, calculate the supply and demand deviation of each drug in combination drug use. Step S4: Construct a cross-category quota adjustment matrix. The off-diagonal elements in the cross-category quota adjustment matrix represent the quota transfer amount generated between related drugs due to the combined use relationship. The quota transfer amount is superimposed on the basic planned release amount of each drug in the next monitoring window to generate the corrected dynamic release quota and send it to the supply chain inventory interface.
2. The method for real-time monitoring of the consumable supply chain based on multi-source data fusion according to claim 1, characterized in that, Step S1 involves constructing the topological framework of the combined drug use association subnet, including: Extract all historical prescription records of combination drug regimens from the medical order interface. The prescription records include at least the combination drug regimen identifier code, the drug code of each drug in the combination drug regimen, the single dose of each drug, the dosing frequency of each drug, and the number of days of the combination drug regimen. Construct a combined drug use association subnetwork with each drug as a node. ,in For drug node set, It is a set of directed edges; For any two medications, if at least [a certain number of medications are listed in the historical prescription record] If they appear together in the same combination therapy regimen, then at the drug node and Establish directed edges between them; During drug administration, if drug nodes The average time of drug administration is earlier than the drug delivery time. The corresponding average drug administration time is then used to establish a directed edge. Characterizing drug nodes The corresponding changes in drug consumption will be reflected in the drug node. Corresponding drug conduction direction; if drug node The average time of drug administration is earlier than the drug delivery time. The corresponding average drug administration time is then used to establish a directed edge. .
3. The method for real-time monitoring of the consumable supply chain based on multi-source data fusion according to claim 2, characterized in that, Step S1 calculates the directed weight coefficients and response delay parameters for each directed edge, including: For directed edges Calculate directed edges based on historical dosing data Cooperative response coefficient , as a directed edge Directed weight coefficients: Extract all nodes containing drugs Corresponding drugs and drug nodes Record the treatment course of the same combination therapy regimen for the corresponding drugs, and calculate the drug nodes within each treatment course. The corresponding daily consumption and drug delivery points The ratio of the corresponding daily drug consumption, taken as the weighted average of the ratios within each treatment course, is used as the synergistic response coefficient. : ； in For drug nodes and drug nodes The total number of combined drug treatment courses, and The first Drug nodes in each course of treatment and drug nodes The average daily consumption; For directed edges Calculate directed edges Response delay parameters The response delay parameter is taken as the value of the drug node within each treatment course. Corresponding drug single-dose timing and drug delivery time The average time difference between subsequent single doses of the corresponding drug.
4. The method for real-time monitoring of the consumable supply chain based on multi-source data fusion according to claim 1, characterized in that, In step S2, the window is divided into several synchronous monitoring windows according to a fixed window width, including: The earliest of the following three times is used as the zero point of the alignment time axis: the first time the prescription takes effect in the medical order interface, the first time the medication is administered in the hospital information interface, and the first time the inventory is recorded in the supply chain. A unified multi-source data alignment time axis is established. Fixed window width Cut into sections along the aligned time axis A synchronized monitoring window will display any number of... Each monitoring window is recorded as ,in .
5. The method for real-time monitoring of the consumable supply chain based on multi-source data fusion according to claim 4, characterized in that, Step S2 involves compiling consumable characteristic data for each drug, including: In the One monitoring window Internally, the system extracts the sequential drug administration records from the monitoring window of the hospital information interface and summarizes them to obtain the drug node. Actual consumption Combined with fixed window width Calculate the actual consumption rate .
6. The method for real-time monitoring of the consumable supply chain based on multi-source data fusion according to claim 1, characterized in that, Step S3 involves propagating the predicted synergistic consumption response of each associated drug step-by-step along the directed edge direction, including: For directed edges Combined with fixed window width , response delay parameter Convert to lag window number ,in It is a rounding function. Characterizing drug nodes Consumption fluctuations propagate to drug nodes The number of monitoring windows that need to be spanned; In the One monitoring window Internally, regarding drug nodes Identify drug nodes Actual consumption rate Relative to monitoring window Internal planned consumption rate fluctuation range ,in ; Will satisfy Drug nodes are assigned to the source drug set. ,in The preset fluctuation sensing threshold is used; For drug nodes in the source drug set Along the combined drug use subnet Directed edges in To the primary associated drug node Propagation of collaborative consumption response, first-level associated drug nodes The predicted amount of collaborative consumption response is ,in For an indicator function, if and only if the monitored window number is... With lag window number The difference is greater than or equal to The time value is ; If the first-level associated drug node Simultaneously serving as another directed edge The starting point points to the secondary related drug node. Then continue along the directed edge Propagation, secondary related drug nodes The predicted amount of collaborative consumption response is obtained by chain-based cumulative recursion using the collaborative consumption response prediction formula until all reachable drug nodes are traversed.
7. The method for real-time monitoring of the consumable supply chain based on multi-source data fusion according to claim 6, characterized in that, In step S3, based on the difference between the actual consumption of each drug and the predicted synergistic consumption response, the supply-demand deviation of the combined use of each drug is calculated, including: After all transmissions are completed, in the first... One monitoring window Within, for each drug node Combined with actual consumption Computing drug nodes Demand deviation of combined drug use ,in For drug nodes The collaborative consumption response prediction quantity and If drug nodes Not targeted by any source drug and drug node If it does not belong to the source drug set, then , .
8. The method for real-time monitoring of the consumable supply chain based on multi-source data fusion according to claim 1, characterized in that, Step S4 involves constructing a cross-category quota adjustment matrix, including: Based on the supply-demand deviation of combined drug use and the drug node set, a system is constructed. Cross-category quota adjustment matrix ,in Represents the set of drug nodes Number of drug nodes included; cross-category quota adjustment matrix off-diagonal elements Representation in directed edges Lower cause drug node With drug nodes The association between combined drug use and the drug node To drug nodes The amount of quota transfer is based on the collaborative response coefficient corresponding to the directed weight coefficient. ,get ,in The preset quota adjustment step size coefficient, In the first One monitoring window Intra-drug nodes Supply and demand deviation of combined drug use; cross-category quota adjustment matrix diagonal elements Characterizing drug nodes The net deviation, based on the first One monitoring window Intra-drug nodes Demand deviation of combined drug use ,get .
9. A method for real-time monitoring of a consumable supply chain based on multi-source data fusion according to claim 8, characterized in that, In step S4, the revised dynamic allocation quota is generated and distributed to the supply chain inventory interface, including... For drug nodes In the One monitoring window Dynamic allocation quota for ,in Based on the basic plan's deployment volume, For combined drug use related sub-network In terms of drug nodes There exists a set of neighboring drugs connected by directed edges. Characterization from drug nodes To drug nodes Quota transfer volume; The revised dynamic allocation quota Issued to the supply chain inventory interface.
10. A real-time monitoring system for a consumable supply chain based on multi-source data fusion, executing the real-time monitoring method for a consumable supply chain based on multi-source data fusion as described in claim 1, characterized in that, The real-time monitoring system for the consumables supply chain includes: The first module is used to extract prescription records of various combined medication regimens within the historical period from the medical order interface. It constructs the topological skeleton of the combined medication association subnet with drugs as nodes and co-occurrence relationships in the same combined medication regimen as edges. Based on the temporal sequence and dosage ratio of drug nodes in each directed edge during the drug administration process, it calculates the directed weight coefficient and response delay parameter of each directed edge. The second module is used to construct a unified multi-source data alignment time axis, which maps the effective time of the prescription interface, the single drug administration execution time of the hospital information interface, and the inbound and outbound event time of the supply chain inventory interface to the alignment time axis. It is divided into several synchronous monitoring windows with a fixed window width, and the consumable characteristic data of each drug is statistically analyzed in each monitoring window. The consumable characteristic data includes the actual consumption amount and actual consumption rate of the drug. The third module is used to identify the set of source drugs whose actual consumption rate fluctuates within the monitoring window, input the fluctuation amplitude and direction of the source drugs into the combined drug association subnet, and calculate the synergistic consumption response prediction of each associated drug by propagating along the directed edge direction. Based on the difference between the actual consumption of each drug and the synergistic consumption response prediction, the combined drug supply and demand deviation of each drug is calculated. The fourth module is used to construct a cross-category quota adjustment matrix. The off-diagonal elements in the cross-category quota adjustment matrix represent the quota transfer amount generated between related drugs due to the combination of drugs. The quota transfer amount is superimposed on the basic planned release amount of each drug in the next monitoring window to generate the corrected dynamic release quota and send it to the supply chain inventory interface.