An AI-driven smart healthcare product circulation data traceability system and method

Abstract

This invention relates to the field of medical product regulatory technology, specifically to an AI-driven intelligent medical product circulation data traceability system and method. Based on the target drug supply chain, a circulation map is obtained, detailing the inflow and outflow of each batch of target drugs between nodes. After determining the circulation path of each node based on the circulation map, the changes and fluctuations in the circulation records of the target drug between nodes are analyzed to derive the cross-regional circulation performance, enabling the sensitive identification of cross-regional circulation anomalies. Further analysis of the differences in circulation records between adjacent nodes yields the drug return performance and hierarchical circulation dominance, accurately identifying return phenomena and hierarchical circulation characteristics. A comprehensive consideration of these three factors is used to obtain the drug circulation complexity, comprehensively and accurately assessing changes in the circulation path. Finally, abnormal paths are screened and warnings are issued, providing an alert before problematic drugs cause serious consequences. This invention achieves precise traceability and supervision through in-depth analysis of drug circulation data.

Description

Technical Field

[0001] This invention relates to the field of medical product regulatory technology, specifically to an AI-driven smart medical product circulation data traceability system and method. Background Technology

[0002] Drug traceability aims to ensure the safety and authenticity of drugs during production, transportation, sales, and use. Simultaneously, drug traceability helps improve drug regulatory capabilities, quickly identify and handle problematic drugs, ensure drug quality meets standards, assist companies in fulfilling relevant regulatory requirements, and promote the standardized development of the industry.

[0003] Current technologies for drug traceability analysis primarily rely on simple data records, such as static inflow-outflow differences, to determine if problems exist in the distribution process. However, in actual drug distribution, multiple information transfers and handovers are involved, along with phenomena such as cross-regional distribution and drug return. Therefore, drug distribution at each node exhibits complex and fluctuating characteristics. If the dynamic changes in drug quantity during distribution are ignored, the credibility of drug supervision is easily reduced in cross-regional and multi-layered drug distribution scenarios, and the accuracy of early warning capabilities will be significantly compromised. Summary of the Invention

[0004] To address the complex and fluctuating characteristics of drug circulation at various stages, including cross-regional distribution and drug return, which arise from the multiple information transfers and handovers involved in the actual drug distribution process, and the potential for significant reductions in the reliability of drug supervision and the accuracy of early warning capabilities in cross-regional and multi-layered drug distribution scenarios if the dynamic changes in drug quantity during circulation are ignored, this invention aims to provide an AI-driven smart medical product circulation data traceability system and method. The specific technical solution adopted is as follows:

[0005] An AI-driven method for tracing the circulation data of smart healthcare products, comprising:

[0006] Based on the supply chain of the target drug, a distribution map is obtained, and the distribution records of the target drug in each batch are obtained between various nodes on the distribution map; wherein, the distribution records include inflow and outflow.

[0007] Obtain all circulation paths corresponding to each node in the circulation diagram; on each circulation path of each node, analyze the change characteristics and fluctuation characteristics of the target drug's circulation records between nodes to obtain the cross-regional circulation performance of the target drug on each circulation path of each node.

[0008] On each flow path at each node, the differences in the flow records of the target drug between adjacent nodes are analyzed to obtain the drug reflux performance and hierarchical flow dominance of the target drug on each flow path at each node; the drug reflux performance and hierarchical flow dominance are fused based on the off-site flow performance to obtain the drug flow complexity on each flow path at each node.

[0009] In the flow diagram, abnormal flow paths are screened and warnings are issued based on the drug flow complexity corresponding to the flow path of each node.

[0010] Furthermore, the method for obtaining the performance of cross-regional circulation includes:

[0011] For any node in the flow diagram, on each flow path of that node, the local flowability of the target drug in each batch at that node is determined based on the quantitative relationship between the total outflow and total inflow of the target drug at that node.

[0012] On each flow path of this node, the off-site flow rate of the target drug at this node is determined based on the quantitative relationship between the total inflow and total outflow of the target drug under each batch in other nodes.

[0013] On each distribution path of this node, the standard deviation of the local circulation of the target drug in all batches at this node is negatively correlated and normalized, and the result is used as the local circulation stability of the target drug at this node.

[0014] On each flow path of this node, analyze the changing characteristics of the flow records of different batches of target drugs between nodes to determine the outflow trend of the target drugs at this node;

[0015] Based on the local circulation stability and outflow trend of the target drug on each circulation path at this node, the off-site circulation performance of the target drug on each circulation path at this node is determined, wherein the off-site circulation performance is positively correlated with the outflow trend and negatively correlated with the local circulation stability.

[0016] Furthermore, the method for obtaining the local circulation degree includes:

[0017] The ratio of the total outflow to the total inflow of the target drug in each batch at this node is used as the local circulation of the target drug in each batch at this node. If there is no inflow, the local circulation is a preset value.

[0018] Furthermore, the method for obtaining the cross-regional flow rate includes:

[0019] Choose any node in the flow diagram as the core node. On each flow path of the core node, take the other nodes besides the core node as hierarchical nodes. Take any batch of the target drug as the target batch.

[0020] The sum of the local circulation of the target drug in all hierarchical nodes under the target batch is used as the off-site circulation of the target drug in the core node under the target batch.

[0021] Furthermore, the method for obtaining the outflow trend degree includes:

[0022] On each flow path of this node, the out-of-town flow of the target drug in each batch at this node is arranged in the order of the batches to obtain a sorting sequence. In the sorting sequence, the first-order difference sequence of the out-of-town flow is calculated. The sum of all values ​​in the first-order difference sequence is normalized and used as the outflow trend of the target drug at this node.

[0023] Furthermore, the method for obtaining the circulation path includes:

[0024] In the flow diagram, for any given node, each node directly connected to that node is considered a directly affected node, and each node directly connected to each of the directly affected nodes is considered an indirectly affected node, thus obtaining a flow path for that node.

[0025] Furthermore, the analysis of differences in the flow records of the target drug between adjacent nodes along each flow path at each node yields the drug reflux performance and hierarchical flow dominance of the target drug along each flow path at each node, including:

[0026] In the flow diagram, any node is selected as the node to be tested. The ratio of the inflow of the target drug in each flow path of the node to be tested to the total outflow of the target drug in each flow path of the node to be tested is used as the direct outflow performance of the target drug in each flow path of the node to be tested.

[0027] The ratio of the inflow of the target drug from all batches to the indirectly affected nodes along each flow path of the tested node to the total outflow of the target drug to the directly affected nodes along each path of the tested node is used as the indirect outflow performance of the target drug along each flow path of the tested node.

[0028] For each flow path of the node under test, the ratio of the inflow of the target drug from the directly affected node to the node under test to the inflow from the node under test to the directly affected node is used as the drug backflow performance of the target drug in each flow path of the node under test.

[0029] The normalized value of the difference between the indirect outflow performance and the direct outflow performance of the target drug in each circulation path of the node under test is used as the hierarchical circulation dominance of the target drug in each circulation path of the node under test.

[0030] Furthermore, the method for obtaining the complexity of drug distribution includes:

[0031] The normalized value of the cross-regional circulation performance of the target drug at each node and each circulation path is used as the cross-regional circulation weight of the target drug at each node and each circulation path.

[0032] In the flow diagram, for each flow path of the target drug at each node, the hierarchical flow dominance is weighted using the weight of cross-regional flow, and the drug return performance is weighted using the difference between the constant 1 and the weight of cross-regional flow. The normalized value of the weighted sum is used as the drug flow complexity of the target drug at each flow path of each node.

[0033] Furthermore, in the flow graph, the process of filtering abnormal flow paths and issuing early warnings based on the drug flow complexity corresponding to the flow path of each node includes:

[0034] In the flow diagram, the drug flow complexity of each flow path at each node is compared with a preset complexity threshold. If it is greater than or equal to the preset complexity threshold, the flow path is marked as an abnormal flow path.

[0035] In the flow diagram, the sum of the drug flow complexity of all abnormal flow paths is normalized and used as the warning level value.

[0036] If the warning level value is greater than the preset warning threshold, a warning needs to be issued.

[0037] An AI-driven smart medical product circulation data traceability system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the AI-driven smart medical product circulation data traceability method.

[0038] The present invention has the following beneficial effects:

[0039] By acquiring a distribution map based on the target drug's supply chain and meticulously recording the inflow and outflow of each batch of the target drug at each node, a clear picture of the complete path and distribution records from production to use can be presented. Based on the distribution map, the corresponding distribution path for each node can be determined. Then, by analyzing the changing and fluctuating characteristics of the target drug's distribution records between nodes along each distribution path, the cross-regional distribution performance can be obtained. This allows for the keen identification of anomalies in the cross-regional distribution of drugs, enabling early detection of these potential safety risks and facilitating timely intervention by regulatory authorities to prevent problematic drugs from entering the market. Furthermore, by analyzing the differences in the distribution records of the target drug between adjacent nodes along each distribution path at each node, the drug return performance and hierarchical distribution dominance can be obtained. This allows for the precise identification of drug return phenomena and significant characteristics of the target drug's distribution between hierarchical nodes. By comprehensively considering the cross-regional distribution performance, drug return performance, and hierarchical distribution dominance, the drug distribution complexity can be obtained, enabling a more comprehensive and accurate assessment of the complex changes in drugs along each distribution path. Finally, abnormal distribution paths can be screened and warnings issued based on the drug distribution complexity, allowing alerts to be issued before problematic drugs cause serious consequences. In summary, through in-depth analysis of drug distribution data, this invention enables regulatory authorities to understand the characteristics and risk levels of drug distribution at different nodes and stages, thereby allocating regulatory resources in a targeted manner and achieving precise regulation. Attached Figure Description

[0040] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 A flowchart illustrating an AI-driven smart medical product circulation data traceability method provided in one embodiment of the present invention;

[0042] Figure 2 A flowchart illustrating a method for obtaining the performance of cross-regional circulation, as provided in an embodiment of the present invention;

[0043] Figure 3 This is a system block diagram of an AI-driven smart medical product circulation data traceability system provided in one embodiment of the present invention;

[0044] Figure 4 This is a schematic diagram of the system structure of an AI-driven smart medical product circulation data traceability system provided in one embodiment of the present invention. Detailed Implementation

[0045] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of an AI-driven smart medical product circulation data traceability system and method proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0047] The following description, in conjunction with the accompanying drawings, details the specific scheme of an AI-driven smart medical product circulation data traceability system and method provided by the present invention.

[0048] Please see Figure 1 The diagram illustrates a flowchart of an AI-driven smart medical product circulation data traceability method according to an embodiment of the present invention, which includes the following steps:

[0049] Step S1: Obtain a distribution map based on the supply chain of the target drug, and obtain the distribution records of the target drug in each batch between various nodes on the distribution map; wherein, the distribution records include the inflow and outflow.

[0050] Drug traceability is generally analyzed for key drugs, especially those with higher risks and greater complexity, such as biopharmaceuticals, vaccines, and chemically synthesized drugs. The production, distribution, and sales of these drugs require stricter supervision. On the other hand, the traceability requirements for some low-risk, routine drugs may be more lenient. For example, some common over-the-counter drugs may not require the same strict tracking throughout the entire process as high-risk drugs.

[0051] Therefore, this study focuses on key drugs (i.e., specially managed drugs, biological products, high-risk injectables, high-value drugs, and drugs for children / pregnant women), while some common drugs (such as over-the-counter drugs (e.g., cold medicines, vitamins) and low-risk oral solid dosage forms) are not the focus of this study.

[0052] Drug distribution refers to the supply chain process from drug manufacturers to end consumers (such as hospitals and pharmacies). Therefore, the supply chain of the target drug (each key drug) can be extracted by integrating the company's internal ERP system, logistics information, and sales data (data within the past six months). Based on the supply chain, a distribution map and distribution records for each batch of the target drug can be constructed. Each node in the distribution map specifically refers to each sales region or distributor of the drug, and the distribution map should be a two-way diagram, with arrows indicating the flow of the target drug. The distribution records include inflow and outflow. For example, if the batch number is AB2024xxxxxxx, and the outflow from node A to node B is 1000 boxes, then the inflow from node A to node B is also 1000 boxes.

[0053] It should be noted that the acquisition of data such as the circulation records of the target drug was authorized by the relevant users, and the process did not violate relevant laws and regulations, nor did it go against public order and good morals.

[0054] Step S2: Obtain all circulation paths corresponding to each node in the circulation diagram; on each circulation path of each node, analyze the change characteristics and fluctuation characteristics of the target drug's circulation records between nodes to obtain the cross-regional circulation performance of the target drug on each circulation path of each node.

[0055] The drug supply chain involves multiple links, and the circulation paths between different nodes are intricate. Therefore, it is necessary to analyze in detail all circulation paths corresponding to each node in order to fully understand the circulation of drugs on different paths. Thus, we first obtain all circulation paths corresponding to each node in the circulation diagram.

[0056] Preferably, in one embodiment of the present invention, the method for obtaining the circulation path includes:

[0057] In the flow diagram, for any node, each node directly connected to that node is considered a node that directly influences that node, meaning that there is a direct business relationship or logistics connection with that node; then each node directly connected to each of the nodes that directly influences that node is considered a node that indirectly influences that node, indicating that there is an indirect business relationship or logistics connection with that node, thus obtaining a flow path for that node.

[0058] It should be noted that in this embodiment of the present invention, each flow path of each node has three layers: the node itself, the directly affected nodes, and the indirectly affected nodes.

[0059] In the drug distribution process, drug manufacturers typically authorize specific sales territories to designated distributors or retailers to ensure that drugs can only circulate within the authorized areas, thereby avoiding interference from cross-regional sales or distribution. Therefore, drug distribution should conform to the drug's inflow volume, meaning that the outflow of drugs within the authorized area should be lower than or equal to the inflow volume of that area. This ensures a stable distribution process, guarantees that drug supply does not exceed demand, and avoids excessive stockpiling or insufficient supply. Moreover, within the authorized sales area, the distributed drugs should not enter other unauthorized areas, ensuring that drugs enter the market only through specific distributors and channels, and preventing them from flowing into other areas through unauthorized channels. Therefore, in this step, the changing characteristics and fluctuation characteristics of the target drug's distribution records between nodes can be analyzed on each distribution path at each node to obtain the cross-regional distribution performance of the target drug on each distribution path at each node. This process transforms the abstract drug distribution situation into quantifiable indicators, which can more intuitively reflect the distribution situation of the target drug on the distribution path.

[0060] Preferably, in one embodiment of the present invention, the method for obtaining the performance of cross-regional circulation includes:

[0061] Please see Figure 2 The diagram illustrates a method flowchart for obtaining the performance of cross-regional circulation in one embodiment of the present invention. The method includes the following steps:

[0062] Step S201: For any node in the flow diagram, on each flow path of that node, based on the quantitative relationship between the total outflow and total inflow of the target drug under each batch at that node, determine the local flowability of the target drug under each batch at that node.

[0063] The ratio of the total outflow to the total inflow of the target drug in each batch at this node is used as the local circulation degree of the target drug in each batch at this node. The local circulation degree reflects the local circulation activity of the target drug in each batch. The larger the value, the higher the local circulation activity. If there is no inflow, the local circulation degree is a preset value. In this embodiment of the invention, the preset value is 0.

[0064] Step S202: On each flow path of this node, determine the off-site flow rate of the target drug at this node based on the quantitative relationship between the total inflow and total outflow of the target drug under each batch in other nodes.

[0065] Choose any node in the flow diagram as the core node. On each flow path of the core node, take the other nodes besides the core node as hierarchical nodes. Take any batch of the target drug as the target batch.

[0066] By analyzing the circulation records of the target drug between hierarchical nodes, the cross-regional circulation characteristics of the target drug on each circulation path of the core node can be more clearly quantified.

[0067] The sum of the local circulation of the target drug in all hierarchical nodes under the target batch is taken as the cross-regional circulation of the target drug in the core node under the target batch. The larger the cross-regional circulation, the more active the circulation of the target drug between the lower-level nodes of the core node.

[0068] Step S203: On each distribution path of this node, analyze the fluctuation characteristics of the local distribution of the target drug under all batches at this node, and determine the local distribution stability of the target drug at this node.

[0069] If multiple batches of drugs exhibit a long-term stable distribution trend in the local market, it usually means that the demand for drugs in the local market is relatively stable, which may limit the trend of drugs flowing out of the local area.

[0070] Therefore, for each distribution path at this node, the standard deviation of the local circulation of the target drug at that node for all batches is calculated. A larger standard deviation indicates more significant fluctuations and lower stability. Therefore, the standard deviation is negatively correlated and normalized to correct the logical relationship, obtaining the local circulation stability of the target drug at that node. A higher local circulation stability indicates that the nodes in the distribution graph for the target drug are relatively fixed, and the activity of cross-regional distribution is lower. This negative correlation mapping and normalization can be performed using the formula... ,in, Let x represent an exponential function with the natural constant e as the base, and let x represent the independent variable.

[0071] Step S204: On each flow path of this node, analyze the change characteristics of the flow records of different batches of target drugs between nodes, and determine the outflow trend of the target drugs at this node.

[0072] If drugs begin to show signs of outflow during the distribution process, that is, if the circulation of drugs to other places is relatively active, then it is necessary to consider whether it has a high trend of outflow to other places.

[0073] Therefore, on each distribution path at this node, the out-of-town flow rate of the target drug in each batch at this node is arranged according to the order of the batches, resulting in a sorted sequence. Then, the first-order difference sequence of out-of-town flow rate is calculated in the sorted sequence. Each value in the first-order difference sequence reflects the change characteristics of the out-of-town flow rate of the target drug in two adjacent batches. The larger the positive value, the more it indicates that the out-of-town flow rate of the target drug is continuously increasing in adjacent batches, and the more obvious the outflow trend is. Therefore, the sum of all values ​​in the first-order difference sequence is calculated, and the normalized value is used as the outflow trend degree of the target drug at this node. The larger the outflow trend degree, the more obvious the out-of-town flow trend of the target drug is considered to be on the distribution path of this node, and the higher the out-of-town activity will be. Since the values ​​in the first-order difference sequence may be positive or negative, the normalization here can be... function.

[0074] Step S205: On each circulation path of the node, combine the local circulation stability and outflow trend of the target drug at the node to obtain the off-site circulation performance of the target drug on each circulation path of the node.

[0075] The local circulation stability and outflow trend of the target drug at each node and along each circulation path reflect the circulation status of the target drug at each node from different perspectives. Therefore, combining the two can more comprehensively and accurately assess the overall circulation performance characteristics of the target drug at each circulation path of the node. Based on the logic in the aforementioned steps, it can be seen that the greater the local circulation stability of the target drug at the node, the lower the activity of the target drug's cross-regional circulation; the greater the outflow trend of the target drug at the node, the higher the activity of the target drug's cross-regional circulation along the node's circulation path. Therefore, the cross-regional circulation performance of the target drug should be positively correlated with the outflow trend of the node and negatively correlated with the local circulation stability of the node. Therefore, in this embodiment of the present invention, the ratio of the outflow trend to the local circulation stability of the target drug at each node and along each circulation path is used as the cross-regional circulation performance of the target drug at each node and along each circulation path. The greater the cross-regional circulation performance, the more significant the cross-regional circulation of the target drug at the node's circulation path, which also indicates that it has more complex and more outward-expanding circulation characteristics. It should be noted that in this embodiment of the present invention, the aforementioned logical relationship is implemented by a ratio. However, if there is a case where the denominator is 0, a preset parameter of 0.01 will be added to the denominator to prevent the denominator from being 0. The specific value can also be adjusted according to the implementation scenario, and is not limited here.

[0076] Step S3: On each flow path of each node, analyze the differences in the flow records of the target drug between adjacent nodes to obtain the drug reflux performance and hierarchical flow dominance of the target drug on each flow path of each node; based on the off-site flow performance, fuse the drug reflux performance and hierarchical flow dominance to obtain the drug flow complexity on each flow path of each node.

[0077] For each distribution path of the target drug at each node, a more comprehensive assessment of the distribution complexity of the target drug along that path is needed. In the actual distribution process, due to changes in market demand, inventory adjustments, transportation errors, etc., drugs may flow back from downstream nodes to upstream nodes. This backflow increases the complexity and uncertainty of the distribution path. Simultaneously, the distribution path of the target drug has a hierarchical structure, with nodes at different levels playing different roles and functions. When a hierarchical node occupies a dominant position in the distribution process, it also increases the complexity of the distribution path. Therefore, for each distribution path at each node, the differences in the distribution records of the target drug between adjacent nodes are analyzed to calculate the drug backflow performance and hierarchical distribution dominance of the target drug at each node and each distribution path, serving as the main reference indicators for evaluating the complexity of drug distribution.

[0078] Preferably, in one embodiment of the present invention, the method for obtaining drug reflux performance and hierarchical flow dominance includes:

[0079] In the flow diagram, any node is selected as the test node. The ratio of the inflow of the target drug from all batches to all directly affected nodes along each flow path of the test node to the total outflow of the target drug along all flow paths of the test node is used as the direct outflow performance of the target drug along each flow path of the test node. The greater the direct outflow performance, the more obvious the direct outflow of the target drug along that flow path of the test node.

[0080] Since each node may exist on multiple distribution paths simultaneously, the ratio of the inflow of the target drug from all batches to the indirectly affected nodes on each distribution path of the node under test to the total outflow of the target drug to the directly affected nodes on each path of the node under test is used as the indirect outflow performance of the target drug on each distribution path of the node under test. The larger this value is, the greater the proportion of the indirect outflow performance of the drug on that distribution path in the total outflow performance of the indirectly affected nodes, and the more obvious the indirect flow characteristics of the target drug on that distribution path of the node under test.

[0081] On each flow path of the node under test, the ratio of the inflow of target drugs from all batches from the directly affected node to the node under test to the inflow from the node under test to the directly affected node is used as the drug backflow performance of the target drugs on each flow path of the node under test. The higher the backflow performance, the more target drugs accumulate and flow back after flowing from the node under test to the directly affected node, thus increasing the complexity of the flow. It should be noted that if there is no inflow of target drugs at the node under test (i.e., no inflow from the directly affected node to the node under test), the drug backflow performance is a preset value, which is 0 in this embodiment of the invention.

[0082] The normalized value of the difference between the indirect and direct outflow performance of the target drug along each distribution path at the test node is used as the hierarchical distribution dominance of the target drug along each distribution path at the test node. A higher hierarchical distribution dominance indicates that the target drug's distribution activity is more significant at the distal end of the distribution path compared to the proximal end, indicating a greater tendency towards multi-level distribution. In this case, potential, difficult-to-detect abnormal drug distribution problems are more likely to occur, increasing the difficulty of monitoring drug safety. It should be noted that since the difference can be positive or negative, the normalization method used here can be... function.

[0083] Drug reflux performance primarily reflects whether and to what extent the target drug refluxes along the distribution path, demonstrating a reverse flow characteristic of the drug along that path. Hierarchical distribution dominance measures the difference in the dominant forces of indirect flow to indirectly influenced nodes versus direct flow to directly influenced nodes along the distribution path, reflecting the propagation and diffusion patterns of the drug between different levels of the distribution network. Interregional distribution performance reflects the activity level of the target drug's distribution between different regions, reflecting the drug's cross-regional spatial flow. Integrating these three indicators allows for a comprehensive and integrated description of the complex characteristics of the target drug's distribution at each node and along each distribution path, encompassing multiple dimensions such as reflux, inter-hierarchical propagation, and interregional distribution, thus yielding the drug distribution complexity.

[0084] Preferably, in one embodiment of the present invention, the method for obtaining drug circulation complexity includes:

[0085] The normalized value of the cross-regional circulation performance of the target drug at each node and on each circulation path is used as the cross-regional circulation weight of the target drug at each node and on each circulation path. When the cross-regional circulation weight is larger, it indicates that the cross-regional circulation of the target drug at that node is more significant, which also indicates that it has more complex and more outward circulation characteristics. Therefore, more attention should be paid to the drug outflow performance characteristics on that circulation path. Conversely, when the cross-regional circulation weight is smaller, it indicates that the drug outflow performance is not obvious enough. Then, more attention should be paid to the drug return performance characteristics on that circulation path.

[0086] Therefore, in the flow diagram, for each flow path of the target drug at each node, the hierarchical flow dominance is weighted using the weight of cross-regional flow. Then, the drug return performance is weighted using the difference between the constant 1 and the cross-regional flow weight. This achieves the aforementioned logical relationship: when the cross-regional flow weight is large, the hierarchical flow dominance plays a more important role in the calculation; conversely, when the cross-regional flow weight is small, the drug return performance plays a more important role. Finally, the weighted sum is normalized and used as the drug flow complexity of the target drug at each node and on each flow path. The greater the drug flow complexity, the more complex the drug flow situation is considered on that flow path, and the more likely abnormal drug flow situations will occur, which is more conducive to the infiltration and spread of counterfeit drugs. Therefore, it needs to be given more attention. Normalization is a technique well-known to those skilled in the art. The choice of normalization function can be linear normalization or standard normalization, etc., and the specific normalization method is not limited here.

[0087] Step S4: In the flow diagram, abnormal flow paths are screened and warnings are issued based on the drug flow complexity corresponding to the flow path of each node.

[0088] In the aforementioned steps, the drug circulation complexity corresponding to each circulation path of each node in the circulation graph can be calculated. In this step, abnormal circulation paths can be screened out in the circulation graph based on the drug circulation complexity, thereby providing early warning.

[0089] Preferably, in one embodiment of the present invention, in the flow graph, abnormal flow paths are screened and early warnings are issued based on the drug flow complexity corresponding to the flow path of each node, including:

[0090] Given that the greater the complexity of drug distribution paths, the more likely it is to encounter abnormal drug distribution situations and the more likely it is to be mixed in and spread by counterfeit drugs, the drug distribution complexity of each distribution path at each node in the distribution diagram is compared with a preset complexity threshold. If it is greater than or equal to the preset complexity threshold, the distribution path is marked as an abnormal distribution path.

[0091] Then, in the flow diagram, the sum of the drug flow complexity of all abnormal flow paths is normalized and used as the warning level value. If the warning level value is greater than the preset warning threshold, an warning needs to be issued and the abnormal flow path is highlighted (red).

[0092] It should be noted that in this embodiment of the present invention, the preset complexity threshold is 0.6 and the preset warning threshold is 0.65. The specific values ​​can be adjusted according to the implementation scenario and are not limited here.

[0093] To facilitate calculations, all index data involved in the calculations in this embodiment of the invention have undergone data preprocessing to eliminate the influence of dimensions. The specific methods for eliminating the influence of dimensions are well-known to those skilled in the art and are not limited here.

[0094] In summary, obtaining a distribution map based on the target drug's supply chain and recording the inflow and outflow of each batch of the target drug between various nodes clearly presents the complete path and distribution records of the drug from production to use. Based on the distribution map, the distribution path corresponding to each node can be determined. Then, by analyzing the changing and fluctuating characteristics of the target drug's distribution records between nodes on each distribution path, the cross-regional distribution performance can be obtained. This allows for the keen identification of anomalies in the cross-regional distribution of drugs, enabling early detection of these potential safety risks and facilitating timely intervention by regulatory authorities to prevent problematic drugs from entering the market. Furthermore, analyzing the differences in the distribution records of the target drug between adjacent nodes on each distribution path at each node yields the drug return performance and hierarchical distribution dominance, accurately identifying drug return phenomena and significant characteristics of the target drug's distribution between hierarchical nodes. Finally, by comprehensively considering the cross-regional distribution performance, drug return performance, and hierarchical distribution dominance, the drug distribution complexity can be obtained, enabling a more comprehensive and accurate assessment of the complex changes in drugs along each distribution path. Finally, abnormal distribution paths can be screened and early warnings issued based on the complexity of drug distribution, allowing alerts to be issued before problematic drugs cause serious consequences. Through in-depth analysis of drug distribution data, this invention enables regulatory authorities to understand the characteristics and risk levels of drug distribution at different nodes and stages, thereby allocating regulatory resources in a targeted manner and achieving precise regulation.

[0095] This invention also provides an AI-driven smart medical product circulation data traceability system. Please refer to [link / reference]. Figure 3 The diagram shows a system block diagram, including a data acquisition module 301 for implementing step S1 in the above method embodiment; a circulation analysis module 302 for implementing steps S2 and S3 in the above method embodiment; and an early warning and traceability module 303 for implementing step S4 in the above method embodiment.

[0096] It should be noted that the system provided in the above embodiments is only an example of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above. In addition, the AI-driven smart medical product circulation data traceability system and the AI-driven smart medical product circulation data traceability method embodiment provided in the above embodiments belong to the same concept, and their specific implementation process can be found in the method embodiment, which will not be repeated here.

[0097] Please see Figure 4 This diagram illustrates a system architecture of an AI-driven smart medical product circulation data traceability system according to an embodiment of the present invention. The system includes a processor 400, a memory 401, a bus 402, and a communication interface 403. The processor 400, communication interface 403, and memory 401 are connected via the bus 402. The memory 401 may contain a high-speed random access memory, and the bus 402 may be an ISA bus, PCI bus, or EISA bus, etc. The processor 400 may be an integrated circuit chip with signal processing capabilities. The memory 401 stores at least one instruction, at least one program, code set, or instruction set. When the processor loads and executes the at least one instruction, at least one program, code set, or instruction set, it implements the steps in an AI-driven smart medical product circulation data traceability method.

[0098] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0099] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. An AI-driven method for tracing the circulation data of smart medical products, characterized in that, The method includes: Based on the supply chain of the target drug, a distribution map is obtained, and the distribution records of the target drug in each batch are obtained between various nodes on the distribution map; wherein, the distribution records include inflow and outflow. Obtain all circulation paths corresponding to each node in the circulation diagram; on each circulation path of each node, analyze the change characteristics and fluctuation characteristics of the target drug's circulation records between nodes to obtain the cross-regional circulation performance of the target drug on each circulation path of each node. On each flow path at each node, the differences in the flow records of the target drug between adjacent nodes are analyzed to obtain the drug reflux performance and hierarchical flow dominance of the target drug on each flow path at each node; the drug reflux performance and hierarchical flow dominance are fused based on the off-site flow performance to obtain the drug flow complexity on each flow path at each node. In the flow diagram, abnormal flow paths are screened and warnings are issued based on the drug flow complexity corresponding to the flow path of each node. Methods for obtaining the performance of cross-regional circulation include: For any node in the flow diagram, on each flow path of that node, the local flowability of the target drug in each batch at that node is determined based on the quantitative relationship between the total outflow and total inflow of the target drug at that node. On each flow path of this node, the off-site flow rate of the target drug at this node is determined based on the quantitative relationship between the total inflow and total outflow of the target drug under each batch in other nodes. On each distribution path of this node, the standard deviation of the local circulation of the target drug in all batches at this node is negatively correlated and normalized, and the result is used as the local circulation stability of the target drug at this node. On each flow path of this node, analyze the changing characteristics of the flow records of different batches of target drugs between nodes to determine the outflow trend of the target drugs at this node; Based on the local circulation stability and outflow trend of the target drug on each circulation path at this node, the off-site circulation performance of the target drug on each circulation path at this node is determined, wherein the off-site circulation performance is positively correlated with the outflow trend and negatively correlated with the local circulation stability. Methods for obtaining the circulation path include: In the flow diagram, for any node, each node directly connected to that node is considered as a directly affected node, and each node directly connected to each directly affected node is considered as an indirectly affected node, thus obtaining a flow path for that node. The method involves analyzing the differences in the flow records of the target drug between adjacent nodes along each flow path at each node to obtain the drug reflux performance and hierarchical flow dominance of the target drug along each flow path at each node, including: In the flow diagram, any node is selected as the node to be tested. The ratio of the inflow of the target drug from all batches to all directly affected nodes in each flow path of the node to the total outflow of the target drug in each flow path of the node to be tested is used as the direct outflow performance of the target drug in each flow path of the node to be tested. The ratio of the inflow of the target drug from all batches to the indirectly affected nodes along each flow path of the tested node to the total outflow of the target drug to the directly affected nodes along each path of the tested node is used as the indirect outflow performance of the target drug along each flow path of the tested node. For each flow path of the node under test, the ratio of the inflow of the target drug from the directly affected node to the node under test to the inflow from the node under test to the directly affected node is used as the drug backflow performance of the target drug in each flow path of the node under test. The normalized value of the difference between the indirect outflow performance and the direct outflow performance of the target drug in each circulation path of the node under test is used as the hierarchical circulation dominance of the target drug in each circulation path of the node under test.

2. The AI-driven smart medical product circulation data traceability method according to claim 1, characterized in that, The method for obtaining local circulation includes: The ratio of the total outflow to the total inflow of the target drug in each batch at this node is used as the local circulation of the target drug in each batch at this node. If there is no inflow, the local circulation is a preset value.

3. The AI-driven smart medical product circulation data traceability method according to claim 1, characterized in that, The method for obtaining the cross-regional flow rate includes: Choose any node in the flow diagram as the core node. On each flow path of the core node, take the other nodes besides the core node as hierarchical nodes. Take any batch of the target drug as the target batch. The sum of the local circulation of the target drug in all hierarchical nodes under the target batch is used as the off-site circulation of the target drug in the core node under the target batch.

4. The AI-driven smart medical product circulation data traceability method according to claim 1, characterized in that, The method for obtaining the outflow trend includes: On each flow path of this node, the out-of-town flow of the target drug in each batch at this node is arranged in the order of the batches to obtain a sorting sequence. In the sorting sequence, the first-order difference sequence of the out-of-town flow is calculated. The sum of all values ​​in the first-order difference sequence is normalized and used as the outflow trend of the target drug at this node.

5. The AI-driven smart medical product circulation data traceability method according to claim 1, characterized in that, The methods for obtaining the complexity of drug distribution include: The normalized value of the cross-regional circulation performance of the target drug at each node and each circulation path is used as the cross-regional circulation weight of the target drug at each node and each circulation path. In the flow diagram, for each flow path of the target drug at each node, the hierarchical flow dominance is weighted using the weight of cross-regional flow, and the drug return performance is weighted using the difference between the constant 1 and the weight of cross-regional flow. The normalized value of the weighted sum is used as the drug flow complexity of the target drug at each flow path of each node.

6. The AI-driven smart medical product circulation data traceability method according to claim 1, characterized in that, In the flow diagram, abnormal flow paths are screened and early warnings are issued based on the drug flow complexity corresponding to the flow path of each node, including: In the flow diagram, the drug flow complexity of each flow path at each node is compared with a preset complexity threshold. If it is greater than or equal to the preset complexity threshold, the flow path is marked as an abnormal flow path. In the flow diagram, the sum of the drug flow complexity of all abnormal flow paths is normalized and used as the warning level value. If the warning level value is greater than the preset warning threshold, a warning needs to be issued.

7. An AI-driven smart medical product circulation data traceability system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the AI-driven smart medical product circulation data traceability method as described in any one of claims 1 to 6.

Images