A method and system for clinical pharmacy wisdom management and a storage medium

Abstract

The present application relates to the technical field of pharmaceutical management, and in particular to a method and system for intelligent management of clinical pharmacy and a storage medium. The method comprises the following steps: acquiring real-time drug treatment chain full-factor data, including prescription instruction flow, drug preparation parameters, drug administration execution records and patient physiological feedback signals; constructing a drug treatment chain dynamic evaluation engine; using the drug treatment chain dynamic evaluation engine to establish a digital twin mapping relationship between drug administration operations and clinical terminals for the drug treatment chain full-factor data, and generating a drug digital relationship chain; identifying drug treatment chain fracture risk points based on a multidimensional deviation detection algorithm for the drug digital relationship chain, and obtaining drug treatment abnormal risk points. The present application improves the intelligent management and precise control of the drug treatment chain, and improves the intelligence and precision of the drug treatment chain management through real-time data acquisition, digital twin mapping, intelligent risk identification and intervention.

Description

Technical Field

[0001] This invention relates to the field of pharmaceutical management technology, and in particular to a method, system and storage medium for intelligent management of clinical pharmacy. Background Technology

[0002] Initially, pharmaceutical management was limited to manually recording drug procurement, inventory, and dispensing, relying on human experience to judge drug compatibility and dosage. In the 1990s, with the gradual development of Hospital Information Systems (HIS), clinical pharmacy began to integrate with computer technology, and drug management gradually became digital, especially in drug inventory management and prescription review, where Pharmaceutical Information Systems (PIS) began to play a role. With intelligent pharmaceutical management systems utilizing machine learning and natural language processing technologies, they can automatically identify and warn of potential risks such as adverse drug reactions and drug interactions, improving the accuracy and safety of drug management. However, current technologies for analyzing the drug therapy chain are often static and singular, lacking data-driven dynamic assessment. Furthermore, the risks in the drug therapy chain often rely on human judgment and experience, which can easily lead to delays or omissions, resulting in lower levels of intelligence and accuracy in drug therapy chain management. Summary of the Invention

[0003] Therefore, it is necessary to provide a method, system, and storage medium for intelligent management of clinical pharmacy to solve at least one of the above-mentioned technical problems.

[0004] To achieve the above objectives, a method for intelligent management of clinical pharmacy is provided, the method comprising the following steps:

[0005] Step S1: Real-time acquisition of all elements of the drug therapy chain data, including prescription instruction flow, drug preparation parameters, drug administration records and patient physiological feedback signals;

[0006] Step S2: Construct a dynamic evaluation engine for the drug therapy chain; use the dynamic evaluation engine to match the clinical drug use mapping relationship of all elements of the drug therapy chain data to generate a personalized drug use path diagram; use digital twin technology to extract the topology chain from the personalized drug use path diagram to generate a digital relationship chain of drugs;

[0007] Step S3: Identify drug treatment chain breakage risk points in the drug digital relationship chain based on the multi-dimensional deviation detection algorithm to obtain abnormal drug treatment risk points; set drug intervention nodes in the drug digital relationship chain according to the abnormal drug treatment risk points to obtain drug intervention nodes;

[0008] Step S4: Determine the coverage of drug intervention nodes; trigger the intelligent compensation mechanism based on the coverage of drug intervention nodes, and use the intelligent compensation mechanism to intelligently regulate drug dosage for all elements of the drug treatment chain, so as to execute intelligent management of clinical pharmacy.

[0009] This invention achieves full-process monitoring of drug use by acquiring real-time data on drug preparation, usage records, and physiological feedback. This method ensures that every step of the drug management process can be tracked and recorded in real time, thereby improving management accuracy and response speed. By constructing a digital twin mapping of the drug treatment chain, a real-time digital model can be provided for each drug management link. This dynamic mapping relationship can provide real-time feedback, helping relevant personnel to optimize and adjust the drug management process at any time, thereby reducing human error and improving system operational efficiency. Using a multi-dimensional deviation detection algorithm to analyze the drug management process can identify potential risk points in advance. This algorithm can efficiently and accurately identify deviations in the process, helping relevant personnel to adjust work plans in a timely manner, thereby reducing any defects or hidden dangers in the system. Once an abnormal risk is detected, the system will automatically adjust according to the set intervention nodes. This automated intervention ensures that the drug management process can respond to individual needs, thereby improving the system's adaptability and ensuring optimal management results. When the system detects a need to supplement a link in the drug management process, an intelligent compensation mechanism is automatically triggered to adjust drug management strategies, ensuring the continuity and efficiency of the process. This intelligent mechanism not only improves drug management efficiency but also ensures the accuracy and timeliness of each link. Through the construction of a digital drug relationship chain, data at each link can be clearly tracked. This provides managers with complete operation records and facilitates post-event analysis and evaluation, ensuring the transparency and operability of each link. With the help of a dynamic evaluation engine and data analysis tools, the system can provide managers with real-time and accurate data support, thereby optimizing the decision-making process. This enables managers to more efficiently identify problems, formulate response strategies, and improve the overall operational efficiency of the management system. Therefore, this invention, through real-time data acquisition, digital twin mapping, and intelligent risk identification and intervention, enhances the intelligent management and precise control of the drug therapy chain, improving the intelligence and accuracy of drug therapy chain management.

[0010] Preferably, the construction of the drug therapy chain dynamic evaluation engine in step S2 includes:

[0011] A dynamic pharmaceutical ontology framework is constructed, which includes decision rules in three dimensions: drug-gene interaction, metabolic pathway coupling, and drug delivery device adaptation.

[0012] By employing path discretization modeling technology, the full-element data of the drug treatment chain is decomposed into medication decision units with time dependencies;

[0013] The medication decision-making units are connected to the medication path; the medication path is simulated in real time based on fuzzy Petri nets, and the activation probability of each medication decision-making unit is dynamically updated based on the results of the real-time simulation.

[0014] Establish a path credibility verification mechanism. When the newly generated medication path matches the preset historical successful case database by less than 65%, a manual review process is triggered.

[0015] This invention constructs a dynamic pharmaceutical ontology framework, combining multi-dimensional decision rules based on drug-gene interactions, metabolic pathway coupling, and drug delivery device adaptation. This framework enables comprehensive and accurate dynamic evaluation of each link in the drug therapy chain. It better integrates multiple influencing factors of drugs, ensuring decisions are based on comprehensive information and improving the accuracy and scientific rigor of drug use management. The path discretization modeling technique transforms all elements of the drug therapy chain into time-dependent drug decision units, allowing each unit to make optimal decisions at specific points in time. This time-series management approach captures any temporal changes during treatment, avoiding the omission of key factors affecting treatment efficacy. Real-time extrapolation of the drug delivery path using fuzzy Petri nets dynamically updates the activation probabilities of each drug decision unit, enabling the system to adapt to changes in the patient's physiological state or the external environment. This adaptive capability ensures flexible adjustments to the drug delivery plan throughout the treatment process, improving the adaptability and precision of treatment. Through a path credibility verification mechanism, the system can compare newly generated drug delivery paths with historical successful cases. When the matching degree falls below 65%, a manual review process is triggered, ensuring the reliability and safety of the medication pathway. This mechanism avoids the limitations of automated decision-making relying excessively on algorithms, and ensures the accuracy of decisions through manual review, further enhancing the system's credibility and security. Through real-time simulation and pathway verification mechanisms, potential risk points in the drug treatment chain can be quickly identified, and corresponding adjustment measures can be taken. The system's real-time simulation of the medication pathway allows for timely correction of each step of the medication decision, thereby avoiding the execution of unsuitable drug treatment plans for patients and significantly reducing the risk of adverse reactions during drug treatment.

[0016] Preferably, the method for constructing the dynamic pharmaceutical ontology framework includes:

[0017] A drug action quadruple model was established, in which the quadruple includes a chemical entity, a target protein, a metabolic enzyme system, and a transporter, with each element associated with at least 3,000 validated biomedical entities; the drug action quadruple model was adapted to generate drug-gene interaction decision rules.

[0018] Define a clinical pathway constraint rule base, which includes three types of constraints: contraindication conflict detection rules, dosing timing optimization rules, and device compatibility verification rules; adapt the decision rules to the clinical pathway constraint rule base to generate metabolic pathway coupled decision rules.

[0019] A drug property decay model is constructed, and decision rules are adapted to the drug property decay model to generate drug delivery device adaptation decision rules. A dynamic pharmaceutical ontology framework is constructed based on drug-gene interaction decision rules, metabolic pathway coupling decision rules, and drug delivery device adaptation decision rules.

[0020] This invention establishes a four-tuple model of drug action (including chemical entities, target proteins, metabolic enzyme systems, and transport carriers) to comprehensively capture the interaction between drugs and organisms. This four-tuple model not only accurately describes the drug's mechanism of action but also ensures the comprehensiveness and accuracy of drug action by associating a large number of validated biomedical entities (each element is associated with at least 3000 entities), contributing to more in-depth drug efficacy prediction and optimization. Adapting the decision rules of the four-tuple model generates drug-gene interaction decision rules. These rules help managers better understand the relationship between drugs and genes, thereby developing more precise drug administration plans and improving the personalization and effectiveness of drug therapy. The defined clinical pathway constraint rule base (including contraindication conflict detection rules, dosing sequence optimization rules, and device compatibility verification rules) provides an important optimization framework for the drug therapy process. By adapting the decision rules of the clinical pathway constraint rule base to generate metabolic pathway coupling decision rules, contraindication conflicts can be effectively detected and avoided, dosing sequence optimized, and device compatibility ensured. This pathway optimization helps improve the safety and success of drug therapy. By adapting drug property decay models and their decision rules, a drug delivery device adaptation decision rule is generated. This makes the matching of drug properties with the drug delivery device more accurate, ensuring that the drug's effectiveness is maximized throughout the treatment process. The dynamic pharmaceutical ontology framework can automatically adjust decisions to adapt to changes in different drug properties, ensuring that the treatment plan always remains consistent with the patient's actual needs. By combining decision rules for drug-gene interactions, metabolic pathway coupling, and drug delivery device adaptation to construct a dynamic pharmaceutical ontology framework, the system can provide multi-dimensional decision support for drug management. This integrated decision framework can optimize drug use from multiple perspectives, avoiding the limitations of single rules, thereby improving the overall efficiency of drug therapy management.

[0021] Preferably, the medication pathway is simulated in real time based on fuzzy Petri nets, and the activation probability of each medication decision unit is dynamically updated based on the results of the real-time simulation, including:

[0022] Set the path node activation threshold function, where the formula for the path node activation threshold function is as follows:

[0023]

[0024] In the formula, P activateHere, k is the activation threshold for the path node, and c is the drug sensitivity coefficient. actual For real-time blood drug concentration monitoring values, c standard This is a standard reference value;

[0025] The activation probability of the unit node in the medication path is calculated by using the path node activation threshold function.

[0026] Based on the activation probability of the path unit node, the medication path is analyzed for the difference in activation probability between adjacent nodes. When the difference in activation probability between adjacent nodes exceeds 40%, a pharmaceutical monitoring marker is automatically inserted.

[0027] By using pharmaceutical monitoring markers, differential paths are selected for medication routes, and path conflict is inferred from differential paths using fuzzy Petri nets to obtain real-time inference results.

[0028] The activation probability of the medication decision unit is dynamically adjusted based on the results of real-time simulation.

[0029] This invention, by setting an activation threshold function for path nodes, can calculate the activation probability of path nodes based on the difference between real-time blood drug concentration and standard reference values. This mechanism ensures that the activation state of each medication decision unit during drug treatment is closely linked to the patient's actual drug response, thereby achieving precise medication pathway management. Based on real-time simulation results, the activation probability of each medication decision unit can be dynamically adjusted. When the drug's response in the body changes, the system can automatically update the activation probability to ensure that the treatment plan matches the patient's current physiological state. This dynamic adjustment keeps the drug treatment process in an optimal state, thereby improving treatment efficacy and safety. Through path node activation probability difference analysis, when the difference in activation probability between adjacent nodes exceeds 40%, the system automatically inserts a pharmaceutical monitoring marker. This mechanism can promptly detect abnormal fluctuations or potential problems in the drug treatment pathway, helping medical personnel identify risk points in the treatment process and take timely intervention measures. Using pharmaceutical monitoring markers, the system can differentiate between different pathways. By using fuzzy Petri nets to perform path conflict simulation on the differentiated pathways, the system can assess the conflicts or inconsistencies caused by different pathways, providing a safer and more stable treatment pathway and avoiding risks caused by inconsistent pathways. The results of real-time simulations provide data-driven decision support for adjusting treatment pathways. By dynamically adjusting the activation probability of medication decision units, the treatment process can not only respond to changes in the patient but also be continuously optimized based on real-time data, ensuring the scientific and rational use of drugs during treatment.

[0030] Preferably, step S3 includes the following steps:

[0031] Step S31: Calculate the path morphology difference degree in the drug digital relationship chain based on the multi-dimensional deviation detection algorithm to obtain the path morphology difference degree of the relationship chain;

[0032] Step S32: Perform path deviation interaction analysis on the digital relationship chain of drugs based on the difference in relationship chain path form, and generate path deviation interaction data;

[0033] Step S33: Use path deviation interaction data to identify path intersections in the drug digital relationship chain, and mark the medication decision-making units corresponding to the identified path intersections as abnormal risk points for drug treatment;

[0034] Step S34: Set up drug intervention nodes in the drug digital relationship chain based on the abnormal risk points of drug treatment to obtain drug intervention nodes.

[0035] This invention utilizes a multi-dimensional deviation detection algorithm to calculate the path morphology difference in the digital relationship chain of drugs, effectively identifying abnormal deviations and inconsistencies in the drug treatment path. This method ensures that any deviation from the preset path during drug treatment can be promptly identified and addressed, preventing potential treatment problems. By performing path deviation interaction analysis on the digital relationship chain of drugs, the system obtains detailed path deviation interaction data. This data provides more accurate path analysis, helping to identify and analyze key nodes or factors leading to inconsistencies in drug treatment plans, enhancing the controllability of the drug treatment process. Through path intersection point identification, abnormal risk points in drug treatment caused by path deviations can be effectively marked. This identification mechanism ensures that the drug treatment path always conforms to standards, promptly identifying and marking potential risk points, effectively improving the safety of the treatment process. The setting of drug intervention nodes allows for timely insertion of intervention measures into the drug treatment path, ensuring that the drug treatment chain does not produce adverse reactions due to path deviations or abnormalities. By identifying abnormal risk points in drug treatment, the system can proactively adjust the treatment plan and perform necessary interventions, thereby avoiding instability during the treatment process. By calculating path morphology differences and identifying path intersections, treatment pathways in the drug digital relationship chain can be optimized, making them more stable and reliable. This method reduces the probability of inconsistencies or erroneous paths during treatment, improving the safety of treatment plans. Through real-time analysis of path deviations and dynamic setting of intervention nodes, the system can implement intelligent pharmaceutical monitoring, automatically identifying and intervening in potential problems in the treatment pathway. This intelligent decision support not only improves treatment efficiency but also ensures the personalization and precision of the drug treatment process.

[0036] Preferably, the multi-dimensional deviation detection algorithm in step S31 is as follows:

[0037]

[0038] In the formula, D path The relationship chain path pattern difference is represented by N, where N is the total number of path nodes, and w is the number of nodes in the path. geo For geometric deviation weights, d geo (i) represents the degree of deviation of the i-th node in geometric space, w topo For topological deviation weights, d topo (i) To calculate the topological deviation of the i-th node, w time For time deviation weights, d time (i) represents the deviation of the i-th node in the time dimension, w attr For attribute deviation from weight, d attr (i) represents the deviation of the attribute information of the i-th node.

[0039] This invention analyzes and integrates a multi-dimensional deviation detection algorithm formula. The formula works by comprehensively considering the deviation degrees of each node in the path across four dimensions: geometry, topology, time, and attributes, to calculate the difference in the path's morphology. Used for multi-dimensional deviation detection, it helps identify the deviations of each node on the path from the desired state in different dimensions, thereby evaluating the overall change or anomaly in the path's morphology. This calculation method can comprehensively reflect potential anomalies and changes on the path when processing complex spatial data. The formula uses D... path This represents the path morphology difference, measuring the overall deviation of all nodes in the path across various dimensions. (D) path The larger the value, the more pronounced the changes or anomalies in the path's shape. N represents the total number of path nodes, reflecting the path's length or complexity. The more path nodes there are, the higher the value of D. path The more complex the dimensions involved in the calculation, the better. geo w topo w time and w attr These four parameters are weighting coefficients, representing the importance of the geometric, topological, temporal, and attribute dimensions in the overall variance. Adjusting these weights controls the contribution of each dimension to the overall variance. Different application scenarios require different weight settings, which should be adjusted according to the specific task requirements. geo (i) represents the difference between the node's current position in geometric space and its desired position. Geometric deviation reflects changes in the node's spatial position caused by path curvature or movement. d topo (i) Describe the topological changes of this node relative to other nodes. Topology deviation is often used to detect changes in the connectivity between nodes, such as network topology interruptions. d time(i) represents the difference between the timestamp of this node and the expected time point. Time deviation is used to analyze temporal changes in a path or data stream, detecting whether data points are captured or processed within the predetermined time. attr (i) represents the difference between the attribute data of the node and the expected value. Attribute deviation is often used to analyze whether the feature information of a node (such as attribute values, classification labels, etc.) has changed. When using conventional multi-dimensional deviation detection algorithms in the field, the relationship chain path morphology difference degree can be obtained. By applying the multi-dimensional deviation detection algorithm provided by this invention, the relationship chain path morphology difference degree can be calculated more accurately. By combining geometric, topological, temporal, and attribute deviations, the anomaly of the path can be evaluated more comprehensively. For example, a single geometric deviation is insufficient to fully describe the path problem, but by comprehensively considering topological, temporal, and attribute deviations, more detailed potential problems can be discovered. By weighted summarizing the deviations of each node in different dimensions, it is possible to locate which specific dimensions or nodes caused the overall path anomaly. This not only detects anomalies on the path but also reveals the root cause of the problem. This algorithm can automatically extract deviation information from the data, is highly adaptable, and can be used to process different types of spatial, temporal, topological, and attribute data, making it highly versatile in practical applications.

[0040] Preferably, step S4 includes the following steps:

[0041] Step S41: Determine the coverage of drug intervention nodes;

[0042] Step S42: Analyze the scope of intervention of drug digital relationship chain based on the coverage of drug intervention nodes, and generate the scope of intervention of nodes;

[0043] Step S43: Analyze the path propagation impact of the drug digital relationship chain through the scope of node intervention, and generate path propagation impact data; use the path propagation impact data to intelligently regulate drug usage in the drug digital relationship chain, so as to execute intelligent clinical pharmacy management operations.

[0044] This invention ensures precise intervention by assessing the coverage of drug intervention nodes, guaranteeing that interventions reach key points and minimizing omissions or over-intervention. By combining digital twin technology, it analyzes the scope and path propagation impact of drug intervention nodes, enabling real-time monitoring of dynamic changes in drug use and intelligent regulation. This regulation mechanism dynamically adjusts drug dosage based on individual patient needs, treatment progress, and drug response, avoiding overdosing or underdosing and maximizing treatment effectiveness. The generation and analysis of path propagation impact data provide scientific decision support for clinical pharmacy, optimizing drug use strategies, improving medical efficiency, reducing unnecessary waste of medical resources, and lowering the incidence of drug-related adverse reactions. Path propagation impact analysis provides a basis for drug use decisions, enhancing the intelligence and scientific rigor of clinical management.

[0045] Preferably, step S41 includes:

[0046] The formula for calculating the coverage radius is defined as follows:

[0047] R cover =λ·D path ;

[0048] Among them, R cover Let λ be the coverage radius of the drug intervention node, and D be the coverage radius adjustment factor. path The degree of difference in the relationship chain path form;

[0049] The coverage area of ​​the drug intervention node is calculated using the coverage radius calculation formula to obtain the coverage area data;

[0050] Adequacy assessment of drug intervention nodes is conducted, and adequacy assessment data for drug intervention nodes is generated;

[0051] If the coverage of a drug intervention node is determined, no action is taken if the coverage adequacy assessment data is greater than or equal to the coverage data; otherwise, an intelligent compensation mechanism is triggered based on the coverage of the drug intervention node.

[0052] This invention uses a set formula for calculating the coverage radius (R) cover =λ·D path This method can accurately calculate the coverage of drug intervention nodes, thereby ensuring that drug intervention can cover the required key nodes. It incorporates path morphology difference (D) pathThe addition of a regulation factor (λ) makes the calculation of the coverage radius more accurate, thereby improving the precision and efficiency of interventions. Through coverage adequacy assessment (based on the coverage radius and the coverage range of intervention nodes), the actual intervention effect of drug intervention nodes can be comprehensively evaluated. When the coverage adequacy assessment data is greater than or equal to the coverage range data, no further intervention is needed, reducing unnecessary intervention operations. If the coverage adequacy assessment data is less than the coverage range data, the system can automatically trigger an intelligent compensation mechanism. The intelligent compensation mechanism supplements the intervention based on the coverage of the drug intervention node, ensuring that no drug intervention is missed and improving the effectiveness of intervention measures. Through accurate coverage judgment and the intelligent compensation mechanism, the system can optimize the use of drug resources according to the actual situation, avoiding ineffective or excessive interventions. This approach improves the efficiency of drug use while reducing potential resource waste.

[0053] This specification provides a smart management system for clinical pharmacy, used to execute the above-described smart management method for clinical pharmacy. The smart management system for clinical pharmacy includes:

[0054] The data acquisition module is used to acquire real-time data on all elements of the drug therapy chain, including prescription instruction flow, drug preparation parameters, drug administration records, and patient physiological feedback signals.

[0055] The relationship chain construction module is used to build a dynamic evaluation engine for drug treatment chains; the dynamic evaluation engine for drug treatment chains is used to match clinical drug use mapping relationships on all elements of drug treatment chain data to generate individualized drug use path diagrams; and the topology chain is extracted from the individualized drug use path diagrams through digital twin technology to generate digital relationship chains for drugs.

[0056] The risk point identification module is used to identify drug treatment chain breakage risk points in the drug digital relationship chain based on a multi-dimensional deviation detection algorithm, thereby obtaining abnormal drug treatment risk points; and to set drug intervention nodes in the drug digital relationship chain based on the abnormal drug treatment risk points, thereby obtaining drug intervention nodes.

[0057] The drug dosage compensation module is used to determine the coverage of drug intervention nodes; based on the coverage of drug intervention nodes, an intelligent compensation mechanism is triggered, and the drug dosage is intelligently adjusted through the intelligent compensation mechanism to perform intelligent management of clinical pharmacy data.

[0058] The present invention also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method for intelligent management of clinical pharmacy.

[0059] The beneficial effects of this invention lie in its ability to acquire real-time data across the drug treatment chain through the data acquisition module, including prescription instructions, drug preparation parameters, dosing records, and patient physiological feedback signals. This function ensures accurate real-time data support for each stage of the drug treatment process, providing a sufficient data foundation for subsequent analysis and decision-making. The relationship chain construction module establishes a dynamic evaluation engine for the drug use chain and utilizes digital twin technology to achieve precise mapping between all elements of the drug use chain data and dosing operations and clinical terminals. The digital twin relationship chain makes the drug use process more transparent and traceable, improving the level of intelligence in pharmaceutical management. The risk point identification module uses a multi-dimensional deviation detection algorithm to identify risk points in the drug digital relationship chain, promptly detecting potential abnormal risks. Early identification of risk points allows for timely adjustments to related operations, thereby reducing unnecessary risks. Through the setting of drug intervention nodes, the system can intelligently intervene in the drug use chain and trigger an intelligent compensation mechanism based on the coverage of drug intervention nodes. The intelligent compensation mechanism can dynamically adjust drug dosage, making personalized adjustments according to actual conditions to optimize operational effects. The drug dosage compensation module supports intelligent regulation, enabling real-time adjustments to drug usage in dynamic environments to ensure accuracy and suitability. The system provides intelligent decision support, helping managers make more precise medication adjustments. Through this system, drug management achieves more precise, personalized, and intelligent control, reducing manual intervention and improving work efficiency. Intelligent drug regulation and timely identification of risk points help reduce operational risks and optimize the safety of the entire process. Therefore, this invention, through real-time data acquisition, digital twin mapping, and intelligent risk identification and intervention, enhances the intelligent management and precise regulation of the drug therapy chain, improving the intelligence and accuracy of drug therapy chain management. Attached Figure Description

[0060] Figure 1 This is a flowchart illustrating the steps involved in a smart management method for clinical pharmacy.

[0061] Figure 2 for Figure 1 A detailed flowchart illustrating the implementation steps of step S3.

[0062] Figure 3 for Figure 1 A detailed flowchart illustrating the implementation steps of step S4.

[0063] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0064] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0065] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0066] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0067] To achieve the above objectives, please refer to Figures 1 to 3 A method for intelligent management of clinical pharmacy, the method comprising the following steps:

[0068] Step S1: Real-time acquisition of all elements of the drug therapy chain data, including prescription instruction flow, drug preparation parameters, drug administration records and patient physiological feedback signals;

[0069] Step S2: Construct a dynamic evaluation engine for the drug therapy chain; use the dynamic evaluation engine to match the clinical drug use mapping relationship of all elements of the drug therapy chain data to generate a personalized drug use path diagram; use digital twin technology to extract the topology chain from the personalized drug use path diagram to generate a digital relationship chain of drugs;

[0070] Step S3: Identify drug treatment chain breakage risk points in the drug digital relationship chain based on the multi-dimensional deviation detection algorithm to obtain abnormal drug treatment risk points; set drug intervention nodes in the drug digital relationship chain according to the abnormal drug treatment risk points to obtain drug intervention nodes;

[0071] Step S4: Determine the coverage of drug intervention nodes; trigger the intelligent compensation mechanism based on the coverage of drug intervention nodes, and use the intelligent compensation mechanism to intelligently regulate drug dosage for all elements of the drug treatment chain, so as to execute intelligent management of clinical pharmacy.

[0072] This invention achieves full-process monitoring of drug use by acquiring real-time data on drug preparation, usage records, and physiological feedback. This method ensures that every step of the drug management process can be tracked and recorded in real time, thereby improving management accuracy and response speed. By constructing a digital twin mapping of the drug treatment chain, a real-time digital model can be provided for each drug management link. This dynamic mapping relationship can provide real-time feedback, helping relevant personnel to optimize and adjust the drug management process at any time, thereby reducing human error and improving system operational efficiency. Using a multi-dimensional deviation detection algorithm to analyze the drug management process can identify potential risk points in advance. This algorithm can efficiently and accurately identify deviations in the process, helping relevant personnel to adjust work plans in a timely manner, thereby reducing any defects or hidden dangers in the system. Once an abnormal risk is detected, the system will automatically adjust according to the set intervention nodes. This automated intervention ensures that the drug management process can respond to individual needs, thereby improving the system's adaptability and ensuring optimal management results. When the system detects a need to supplement a link in the drug management process, an intelligent compensation mechanism is automatically triggered to adjust drug management strategies, ensuring the continuity and efficiency of the process. This intelligent mechanism not only improves drug management efficiency but also ensures the accuracy and timeliness of each link. Through the construction of a digital drug relationship chain, data at each link can be clearly tracked. This provides managers with complete operation records and facilitates post-event analysis and evaluation, ensuring the transparency and operability of each link. With the help of a dynamic evaluation engine and data analysis tools, the system can provide managers with real-time and accurate data support, thereby optimizing the decision-making process. This enables managers to more efficiently identify problems, formulate response strategies, and improve the overall operational efficiency of the management system. Therefore, this invention, through real-time data acquisition, digital twin mapping, and intelligent risk identification and intervention, enhances the intelligent management and precise control of the drug therapy chain, improving the intelligence and accuracy of drug therapy chain management.

[0073] In this embodiment of the invention, reference is made to Figure 1 The diagram shown is a flowchart illustrating the steps of a smart management method for clinical pharmacy according to the present invention. In this example, the smart management method for clinical pharmacy includes the following steps:

[0074] Step S1: Real-time acquisition of all elements of the drug therapy chain data, including prescription instruction flow, drug preparation parameters, drug administration records and patient physiological feedback signals;

[0075] In this embodiment of the invention, 100 prescription data entries are acquired per second through the Medical Information System (HIS) interface, and the prescription content, including drug name, dosage, route of administration, frequency, etc., is parsed. A high-efficiency data stream processing engine is employed to control data latency to ≤200ms, ensuring the real-time nature of prescription instructions. Through an automated pharmacy dispensing system, 50 sets of drug preparation parameters (such as concentration, dissolution rate, mixing ratio, etc.) are collected per minute and standardized, with errors controlled within ±0.5%. An intelligent dispensing algorithm is used, combined with historical data to optimize the dispensing scheme, improving accuracy. Through intelligent infusion pumps, automated drug delivery devices, etc., 20 drug administration execution data entries are recorded per second, including actual infusion volume, infusion rate, remaining dose, etc., and data integrity is verified (missing data rate ≤0.1%). A data synchronization mechanism is employed to ensure a high degree of consistency between drug administration records and prescription instructions, with a data synchronization rate ≥99.9%. Through wearable devices and a physiological monitoring system, 10 sets of key physiological parameters (such as blood pressure, heart rate, blood oxygen, blood glucose, etc.) are collected per second and analyzed in real time. An intelligent anomaly detection model is employed to ensure a false alarm rate of ≤1% for physiological data and an anomaly identification accuracy of ≥98%. A distributed data storage architecture is used to store over 1TB of drug treatment data per hour, and formatting is performed to improve query efficiency. Data tagging technology enables efficient fusion of data from different sources, providing high-quality input for subsequent analysis, ultimately generating comprehensive data on the entire drug treatment chain.

[0076] Step S2: Construct a dynamic evaluation engine for the drug therapy chain; use the dynamic evaluation engine to match the clinical drug use mapping relationship of all elements of the drug therapy chain data to generate a personalized drug use path diagram; use digital twin technology to extract the topology chain from the personalized drug use path diagram to generate a digital relationship chain of drugs;

[0077] In this embodiment of the invention, a dynamic evaluation model of the drug treatment chain is constructed using a deep reinforcement learning (DRL) framework. The model inputs include prescription instruction flow, drug preparation parameters, drug administration records, and patient physiological feedback signals. The reward function for reinforcement learning is defined as: R = α·Ceff - β·Crisk - γ·Cdev; where: Ceff is the treatment effectiveness score (based on physiological feedback improvement, ranging from 0 to 100), Crisk is the drug administration risk coefficient (based on allergic reactions, drug interactions, etc., ranging from 0 to 1), Cdev is the operational deviation value (e.g., deviation between actual and planned drug administration, ranging from 0 to 1), and α, β, and γ are regulatory weights, with initial values ​​set to 1, 0.8, and 0.5, respectively. Multimodal data fusion technology is used to construct a mapping model between drug administration operations and clinical terminals, including: structural mapping: the mapping relationship between drug dosage, route, frequency, and patient physiological state; and behavioral mapping: the temporal synchronization association between nurse / doctor operations, patient feedback, and the drug administration process. Feedback Mapping: Matching short-term trends of patient physiological signals with drug effects to calculate the drug efficacy feedback coefficient (range 0-1, above 0.8 is considered significant efficacy). A high-efficiency computing engine is employed, keeping the mapping calculation time ≤500ms. Knowledge graph construction technology is used to generate a relationship graph of drugs, patients, prescriptions, administration processes, and physiological feedback: Nodes: Drug (name, dosage), Patient (ID, physiological parameters), Prescription (doctor, time), Administration execution (device, time), Physiological feedback (data point). Edges: Temporal associations, causal relationships, feedback paths. Weights: Influence weights between different factors (e.g., the influence coefficient of drug dosage on physiological parameters), generating a digital twin relationship chain database storing ≥1 million related records, supporting real-time queries, with a response time ≤1s.

[0078] Step S3: Identify drug treatment chain breakage risk points in the drug digital relationship chain based on the multi-dimensional deviation detection algorithm to obtain abnormal drug treatment risk points; set drug intervention nodes in the drug digital relationship chain according to the abnormal drug treatment risk points to obtain drug intervention nodes;

[0079] In this embodiment of the invention, a multi-dimensional deviation detection algorithm combining Temporal Anomaly Detection (TAD), High-Dimensional Data Deviation Analysis (HDA), and Bayesian Risk Assessment (BRA) is employed to identify anomalies in the digital relationship chain of pharmaceutical products. A deviation detection formula is defined to obtain the relationship chain path morphology difference degree D. path If D pathA deviation of ≥0.7 (range 0-1) is considered a risk point for a break in the drug treatment chain. Combining historical data backtracking and real-time data stream analysis, the following strategies are used to detect anomalies: three consecutive deviations within the last 10 minutes, an anomaly probability >5% within the past 24 hours, and the use of the DBSCAN clustering algorithm to identify high-density anomaly regions. Drug treatment anomaly risk points are identified, and a drug treatment anomaly risk point dataset is generated. An adaptive dynamic intervention strategy is adopted, setting drug intervention nodes near the risk points: if the deviation of the drug treatment anomaly risk point is ≥0.5, the dosing time is adjusted; if the deviation is ≥0.4, the dosage is adjusted; if the deviation is ≥0.6, it is recommended to change the drug or adjust the combination. A drug intervention node dataset is generated and updated in the drug digital relationship chain.

[0080] Step S4: Determine the coverage of drug intervention nodes; trigger the intelligent compensation mechanism based on the coverage of drug intervention nodes, and use the intelligent compensation mechanism to intelligently regulate drug dosage for all elements of the drug treatment chain, so as to execute intelligent management of clinical pharmacy.

[0081] In this embodiment of the invention, the coverage of drug intervention nodes in the drug treatment chain is calculated using the Area Coverage Rate (ACR) model, resulting in a drug intervention node coverage value Ccover (range 0-1). A threshold is set: if Ccover ≥ 0.85, the drug intervention node coverage is considered sufficient, requiring no additional compensation. If Ccover < 0.85, an intelligent compensation mechanism is triggered. If Ccover < 0.7, a new drug intervention node is added, and the timing and regimen of drug administration are optimized. If 0.7 ≤ Ccover < 0.85, the dosage parameters of existing drug intervention nodes are adjusted. If the intervention nodes are sufficient but patient feedback is abnormal, the feedback processing flow is optimized to improve the response accuracy of physiological signals. An Intelligent Dosage Adjustment (IDA) algorithm is used, combined with individualized patient data, to perform dose optimization calculations. An intelligent adjustment range is set, with a maximum dose adjustment range ≤ ±20% to avoid overdose risks. Real-time updates to the digital twin relationship chain ensure that the drug dosage after intelligent compensation meets individualized treatment needs. A 5-minute feedback interval is set to monitor patient physiological signals in real time and make secondary adjustments. A multi-objective optimization (MOO) approach is employed to balance efficacy, risk, and dose stability. An intelligent compensation mechanism is integrated into the hospital pharmacy system (HPS) to achieve automated and intelligent pharmacy management.

[0082] Preferably, the construction of the drug therapy chain dynamic evaluation engine in step S2 includes:

[0083] A dynamic pharmaceutical ontology framework is constructed, which includes decision rules in three dimensions: drug-gene interaction, metabolic pathway coupling, and drug delivery device adaptation.

[0084] By employing path discretization modeling technology, the full-element data of the drug treatment chain is decomposed into medication decision units with time dependencies;

[0085] The medication decision-making units are connected to the medication path; the medication path is simulated in real time based on fuzzy Petri nets, and the activation probability of each medication decision-making unit is dynamically updated based on the results of the real-time simulation.

[0086] Establish a path credibility verification mechanism. When the newly generated medication path matches the preset historical successful case database by less than 65%, a manual review process is triggered.

[0087] In this embodiment of the invention, known drug-gene interaction (DGI) data, such as FDA drug labels and the PharmGKB database, are collected. An ontology mapping method is used to link concepts such as drugs, genes, and proteins to a unified knowledge graph. DGI-based medication decision rules are constructed, for example, considering the impact of gene mutations on drug metabolism (e.g., the effect of CYP2C19 on the efficacy of clopidogrel). Metabolic databases such as KEGG and Reactome are used to analyze drug interactions in different metabolic pathways. Metabolic pathway topology analysis is employed to assess the degree of drug influence at different pathway nodes, generating dynamic metabolic pathway coupling rules to guide personalized medication. Adaptability rules are established for different drug dosage forms (oral, injection, implantation) and delivery devices (micropumps, sustained-release systems, etc.). Physicochemical modeling methods are used to evaluate parameters such as drug solubility, release rate, and tissue permeability. The optimal dosing method is recommended based on individual patient characteristics (age, liver and kidney function, etc.). The drug therapy chain is broken down into multiple Medication Decision Units (MDUs), including: initial medication regimen (based on patient characteristics), medication adjustment (adjusting dosage or medication based on treatment feedback), cross-medication management (avoiding drug interactions), and decisions to discontinue or switch medications. A Markov Decision Process (MDP) is used to describe the temporal dependencies of the MDUs. State transition probabilities are set and dynamically updated based on real-time patient feedback data. A knowledge graph of the drug therapy chain is constructed using a graph database (such as Neo4j), with nodes representing MDUs and edges representing decision transition relationships. Optimal medication pathways are dynamically generated by combining individual patient data. Fuzzy Petri Nets (FPNs) are used to extrapolate medication pathways and calculate the probability of different medication regimens. Activation probabilities of each MDU are calculated using fuzzy rules, path weights are updated, and personalized medication recommendations are generated. A historical success case database is used to calculate the matching degree between the new pathway and historical cases. A 65% matching degree threshold is set; if the degree falls below this value, manual review is triggered. An expert system is invoked for in-depth review to evaluate the rationality of the new pathway. If a path is confirmed to be effective, it will be included in the success case library to improve the system's decision-making capabilities.

[0088] Preferably, the method for constructing the dynamic pharmaceutical ontology framework includes:

[0089] A drug action quadruple model was established, in which the quadruple includes a chemical entity, a target protein, a metabolic enzyme system, and a transporter, with each element associated with at least 3,000 validated biomedical entities; the drug action quadruple model was adapted to generate drug-gene interaction decision rules.

[0090] Define a clinical pathway constraint rule base, which includes three types of constraints: contraindication conflict detection rules, dosing timing optimization rules, and device compatibility verification rules; adapt the decision rules to the clinical pathway constraint rule base to generate metabolic pathway coupled decision rules.

[0091] A drug property decay model is constructed, and decision rules are adapted to the drug property decay model to generate drug delivery device adaptation decision rules. A dynamic pharmaceutical ontology framework is constructed based on drug-gene interaction decision rules, metabolic pathway coupling decision rules, and drug delivery device adaptation decision rules.

[0092] In this embodiment of the invention, a quadruple (C, T, M, Tt) is defined, where: C (Chemical Entity) refers to a drug molecule, such as aspirin; T (Target Protein) refers to the protein to which the drug acts, such as COX-1 / COX-2; M (Metabolic Enzyme System), such as the CYP450 enzyme system, affects the drug's metabolic rate; and Tt (Transporter) is responsible for the drug's transmembrane transport, such as P-gp (P-glycoprotein). Natural Language Processing (NLP) technology is used to analyze databases such as PubMed, DrugBank, and KEGG to extract quadruple information. A graph database (Neo4j) is used to store the quadruples, with each element associated with at least 3000 verified biomedical entities. A random walk algorithm is used to calculate the strength of the drug-protein-metabolism-transport association. The gene-level influence of the drug quadruple is analyzed, such as the influence of the CYP2C19 gene on clopidogrel metabolism. A decision tree approach is used to construct drug-gene interaction rules. Example rules: Rule 1 (metabolic effects): CYP2D6*10 mutation -> avoid drugs dependent on CYP2D6 metabolism. Rule 2 (target mutation): EGFR mutation -> select a suitable tyrosine kinase inhibitor (e.g., erlotinib). A clinical pathway constraint rule base is established: e.g., NSAIDs are contraindicated in patients with peptic ulcers; calcium channel blockers and beta-blockers should not be used concurrently; insulin pumps are not suitable for some oral hypoglycemic agent combination therapies. Reasoning over knowledge graphs is used to match medication regimens with the contraindication database. Metabolic network topology analysis is used to identify metabolic pathway conflict risks. Pharmacokinetic (PK) modeling is used to establish drug concentration-time curves and calculate half-life (T1 / 2) and bioavailability (F). Machine learning regression models (e.g., XGBoost) are used to predict drug decay trends under different dosing methods. Adaptation to different delivery devices (oral, injection, implantable devices, etc.). Example rules: Rule 1 (Controlled-release adaptation): For T1 / 2 > 8h, sustained-release formulations can be considered; for T1 / 2 < 2h, frequent dosing or continuous intravenous infusion is required. Rule 2 (Device adaptation): Oral hypoglycemic agents are not suitable for patients relying on insulin pumps with continuous glucose monitoring. An ontology framework is generated by integrating drug-gene interaction decision rules, metabolic pathway coupling decision rules, and drug delivery device adaptation decision rules. Automated reasoning using an ontology language (OWL) and a reasoning engine (such as HermiT) generates personalized medication regimens.

[0093] Preferably, the medication pathway is simulated in real time based on fuzzy Petri nets, and the activation probability of each medication decision unit is dynamically updated based on the results of the real-time simulation, including:

[0094] Set the path node activation threshold function, where the formula for the path node activation threshold function is as follows:

[0095]

[0096] In the formula, P activate Here, k is the activation threshold for the path node, and c is the drug sensitivity coefficient. actual For real-time blood drug concentration monitoring values, c standard This is a standard reference value;

[0097] The activation probability of the unit node in the medication path is calculated by using the path node activation threshold function.

[0098] Based on the activation probability of the path unit node, the medication path is analyzed for the difference in activation probability between adjacent nodes. When the difference in activation probability between adjacent nodes exceeds 40%, a pharmaceutical monitoring marker is automatically inserted.

[0099] By using pharmaceutical monitoring markers, differential paths are selected for medication routes, and path conflict is inferred from differential paths using fuzzy Petri nets to obtain real-time inference results.

[0100] The activation probability of the medication decision unit is dynamically adjusted based on the results of real-time simulation.

[0101] In this embodiment of the invention, a path node activation threshold function is constructed using a logistic function, as shown in the following formula: In the formula, P activate Here, k is the activation threshold for the path node, and c is the drug sensitivity coefficient. actual For real-time blood drug concentration monitoring values, c standardThe standard reference value is used; individualized pharmacokinetic (PK) modeling is used to calculate the patient's k-value, and Bayesian estimation is used to update the parameters. Real-time blood drug concentration monitoring is performed to dynamically adjust the real-time blood drug concentration monitoring value. The activation probability of each path unit node is calculated using a path node activation threshold function, resulting in a path unit node activation probability set: Pnode={P1,P2,...,Pn}; where the P-value corresponding to each node is dynamically calculated based on the real-time blood drug concentration. The difference in activation probabilities between adjacent path unit nodes is calculated: ΔP=|Pi-Pi+1|; a threshold of 40% is set for the difference between adjacent nodes, i.e., ΔP>0.4; if this condition is met, a pharmaceutical monitoring marker is automatically inserted into the path. The pharmacist review process is triggered (if the system determines the risk to be high). The system then enters the difference path selection module (if a potential path optimization scheme exists). For paths with monitoring markers, the system selects the optimal path from the historical successful path library for comparison. The matching degree between the difference paths is calculated using a Dynamic Time Warping (DTW) algorithm. A Fuzzy Petri Net (FPN) model is constructed: The set of states (P) represents different states of the medication pathway (e.g., "drug metabolism stage"). The set of transitions (T) represents pathway switching events (e.g., "drug is metabolized by CYP450"). Fuzzy weights (W) represent uncertainties (e.g., "individual metabolic rate"). Firing probability is determined by the activation probability of pathway unit nodes. The process involves inputting the current medication pathway and candidate pathways, calculating the pathway conflict probability (based on fuzzy weights), identifying conflict points (e.g., abnormal concentrations due to drug metabolic competition), and adjusting pathway switching strategies (e.g., adjusting dosing time or dosage). If the conflict path probability is too high (e.g., >60%), the activation probability Pi of the corresponding node is reduced. If the new path has a high matching degree (e.g., >80%), the activation probability Pi+1 of the new path is increased. Reinforcement learning algorithms (e.g., Q-learning) are used to continuously optimize the decision-making units. Combined with historical medication data, individualized pathway adjustments are performed to improve prediction accuracy.

[0102] As an example of the present invention, reference is made to... Figure 2 As shown, step S3 in this example includes:

[0103] Step S31: Calculate the path morphology difference degree in the drug digital relationship chain based on the multi-dimensional deviation detection algorithm to obtain the path morphology difference degree of the relationship chain;

[0104] Step S32: Perform path deviation interaction analysis on the digital relationship chain of drugs based on the difference in relationship chain path form, and generate path deviation interaction data;

[0105] Step S33: Use path deviation interaction data to identify path intersections in the drug digital relationship chain, and mark the medication decision-making units corresponding to the identified path intersections as abnormal risk points for drug treatment;

[0106] Step S34: Set up drug intervention nodes in the drug digital relationship chain based on the abnormal risk points of drug treatment to obtain drug intervention nodes.

[0107] In this embodiment of the invention, a path morphology deviation detection index is established by selecting three dimensions: temporal features, spatial features, and biocompatibility features. Temporal features include: medication time point, dynamic curve of blood drug concentration, and metabolic half-life. Spatial features include: drug target, receptor distribution, and metabolic pathway. Biocompatibility features include: individual genotype, drug-protein binding capacity, and metabolic enzyme activity. Dynamic Time Warping (DTW) is used to calculate the morphological matching degree between the standard path and the current medication path, obtaining the path morphological difference degree Dshape. If Dshape ≥ 0.3, it indicates a significant path deviation, and the path deviation interactive analysis process begins. Path deviation thresholds are set: mild deviation (0.1 ≤ Dshape < 0.3): allowed to continue execution but monitored; moderate deviation (0.3 ≤ Dshape < 0.6): triggers medication adjustment prompts; severe deviation (Dshape ≥ 0.6): requires manual intervention or path optimization. The local deviation degree of different path nodes is calculated to generate an interactive data matrix. Where D m,n This represents the deviation degree of the nth node on the m-th path in the drug digital relationship chain. In the path deviation interaction data matrix, we find high deviation points where multiple paths intersect: If D m,n A deviation >0.5 is considered a path intersection point. Path intersection points must meet two conditions: multiple paths overlap (at least two paths deviate at this point); and there must be a significant morphological difference (deviation >0.5). The medication decision unit corresponding to the path intersection point is marked as a drug treatment abnormality risk point. If the number of intersecting paths at this point is >3, the intervention node setting process begins. For metabolic conflict risk points: adjust dosing time (avoid metabolic competition), or replace with drugs with different metabolic pathways. For drug interaction risk points: reduce dosage or modify the medication regimen, increase monitoring frequency, and improve safety. A Fuzzy Inference System (FIS) is used to set the drug intervention plan: Input: risk point deviation, metabolic conflict level, individual patient pharmacokinetic parameters. Output: adjusted drug intervention strategy.

[0108] Preferably, the multi-dimensional deviation detection algorithm in step S31 is as follows:

[0109]

[0110] In the formula, D path The relationship chain path pattern difference is represented by N, where N is the total number of path nodes, and w is the number of nodes in the path. geo For geometric deviation weights, d geo (i) represents the degree of deviation of the i-th node in geometric space, w topo For topological deviation weights, d topo (i) To calculate the topological deviation of the i-th node, w time For time deviation weights, d time (i) represents the deviation of the i-th node in the time dimension, w attr For attribute deviation from weight, d attr (i) represents the deviation of the attribute information of the i-th node.

[0111] This invention analyzes and integrates a multi-dimensional deviation detection algorithm formula. The formula works by comprehensively considering the deviation degrees of each node in the path across four dimensions: geometry, topology, time, and attributes, to calculate the difference in the path's morphology. Used for multi-dimensional deviation detection, it helps identify the deviations of each node on the path from the desired state in different dimensions, thereby evaluating the overall change or anomaly in the path's morphology. This calculation method can comprehensively reflect potential anomalies and changes on the path when processing complex spatial data. The formula uses D... path This represents the path morphology difference, measuring the overall deviation of all nodes in the path across various dimensions. (D) path The larger the value, the more pronounced the changes or anomalies in the path's shape. N represents the total number of path nodes, reflecting the path's length or complexity. The more path nodes there are, the higher the value of D. path The more complex the dimensions involved in the calculation, the better. geo w topo w time and w attr These four parameters are weighting coefficients, representing the importance of the geometric, topological, temporal, and attribute dimensions in the overall variance. Adjusting these weights controls the contribution of each dimension to the overall variance. Different application scenarios require different weight settings, which should be adjusted according to the specific task requirements. geo (i) represents the difference between the node's current position in geometric space and its desired position. Geometric deviation reflects changes in the node's spatial position caused by path curvature or movement. d topo (i) Describe the topological changes of this node relative to other nodes. Topology deviation is often used to detect changes in the connectivity between nodes, such as network topology interruptions. d time(i) represents the difference between the timestamp of this node and the expected time point. Time deviation is used to analyze temporal changes in a path or data stream, detecting whether data points are captured or processed within the predetermined time. attr (i) represents the difference between the attribute data of the node and the expected value. Attribute deviation is often used to analyze whether the feature information of a node (such as attribute values, classification labels, etc.) has changed. When using conventional multi-dimensional deviation detection algorithms in the field, the relationship chain path morphology difference degree can be obtained. By applying the multi-dimensional deviation detection algorithm provided by this invention, the relationship chain path morphology difference degree can be calculated more accurately. By combining geometric, topological, temporal, and attribute deviations, the anomaly of the path can be evaluated more comprehensively. For example, a single geometric deviation is insufficient to fully describe the path problem, but by comprehensively considering topological, temporal, and attribute deviations, more detailed potential problems can be discovered. By weighted summarizing the deviations of each node in different dimensions, it is possible to locate which specific dimensions or nodes caused the overall path anomaly. This not only detects anomalies on the path but also reveals the root cause of the problem. This algorithm can automatically extract deviation information from the data, is highly adaptable, and can be used to process different types of spatial, temporal, topological, and attribute data, making it highly versatile in practical applications.

[0112] As an example of the present invention, reference is made to... Figure 3 As shown, step S4 in this example includes:

[0113] Step S41: Determine the coverage of drug intervention nodes;

[0114] Step S42: Analyze the scope of intervention of drug digital relationship chain based on the coverage of drug intervention nodes, and generate the scope of intervention of nodes;

[0115] Step S43: Analyze the path propagation impact of the drug digital relationship chain through the scope of node intervention, and generate path propagation impact data; use the path propagation impact data to intelligently regulate drug usage in the drug digital relationship chain, so as to execute intelligent clinical pharmacy management operations.

[0116] In this embodiment of the invention, the coverage of a drug intervention node refers to the scope of a drug's influence in a digital twin relationship chain, primarily indicating the degree of influence a drug has on different medication decision-making units. The coverage of each intervention node is calculated based on its influence on other nodes in the path. By analyzing the interaction between the drug and adjacent nodes, as well as the biological effects of the drug, its actual impact on each node in the path is determined. Nodes with larger coverage influence more decision-making units, while nodes with smaller coverage only affect local areas. The coverage of each intervention node is calculated using models based on the drug's physiological mechanisms, pharmacodynamics, and pharmacokinetics. The strength of the interaction between the drug and other relevant factors (such as genes, metabolic pathways, etc.) can be used as a metric. Coverage can be represented numerically, typically categorized into low, medium, and high levels to determine the intensity of the intervention node's influence. After obtaining the coverage of each intervention node, its scope of influence needs to be analyzed. Specifically, based on the influence of the intervention node, the scope of the drug intervention's influence on other nodes in the relationship chain is analyzed. This analysis considers the drug's mechanism of action, pharmacokinetics, pharmacodynamics, and clinical pathway constraints (such as dosing sequence, contraindication conflicts, etc.). This multi-dimensional analysis allows for the precise determination of the impact range of drug intervention nodes, including both direct and indirect effects. Based on the analysis results, a map of the impact range of drug intervention nodes is generated. Path analysis algorithms can be used to identify other nodes directly or indirectly affected by the intervention node, thus defining the region or scope of drug intervention. Using the node intervention impact range obtained from the previous analysis, an impact analysis is performed on path propagation in the drug digital relationship chain. During the analysis, the strength of the drug intervention's impact on different nodes along the path is assessed based on the coverage and impact range of the drug intervention node. Through path propagation impact analysis, propagation impact data for each path is generated, reflecting the propagation strength of the drug intervention throughout the path and its impact on different nodes. Based on this data, it is possible to assess whether the drug intervention meets treatment needs and whether adjustments or optimizations are necessary. Intelligent regulation of drug dosage is then implemented based on the path propagation impact data. Specifically, by assessing the impact of drug intervention on various decision-making units, the dosage and timing of drug administration are intelligently adjusted to achieve personalized and precise treatment. This process relies on a clinical pharmacy intelligent management system, which uses algorithms to adjust drug dosage to ensure maximum drug efficacy and minimize the risk of side effects during treatment. For example, when the spread of a certain route is too strong, it is necessary to reduce the drug dosage or adjust the dosing regimen.

[0117] Preferably, step S41 includes:

[0118] The formula for calculating the coverage radius is defined as follows:

[0119] R cover =λ·D path;

[0120] Among them, R cover Let λ be the coverage radius of the drug intervention node, and D be the coverage radius adjustment factor. path The degree of difference in the relationship chain path form;

[0121] The coverage area of ​​the drug intervention node is calculated using the coverage radius calculation formula to obtain the coverage area data;

[0122] Adequacy assessment of drug intervention nodes is conducted, and adequacy assessment data for drug intervention nodes is generated;

[0123] If the coverage of a drug intervention node is determined, no action is taken if the coverage adequacy assessment data is greater than or equal to the coverage data; otherwise, an intelligent compensation mechanism is triggered based on the coverage of the drug intervention node.

[0124] In this embodiment of the invention, a formula for calculating the coverage radius is used to determine the influence range of drug intervention nodes in the drug digital twin relationship chain. The formula is: R cover =λ·D path Among them, R cover Let λ be the coverage radius of the drug intervention node, and D be the coverage radius adjustment factor. pathThis section describes the difference in relationship chain path morphology. Based on the coverage radius calculation formula, the coverage range of each drug intervention node is calculated. The result is the coverage radius (Rcover), representing the range of paths that the drug intervention node can influence. Based on this coverage radius, the number of covered nodes and paths and their degree of influence are further analyzed to form coverage range data. Based on the coverage range data, the coverage sufficiency of the drug intervention node is assessed. Coverage sufficiency assessment determines whether its effect meets treatment needs by comparing the intersection of the intervention node's coverage range and the actual treatment path. The assessment criteria are: if the coverage sufficiency assessment result is greater than or equal to the coverage range data, the impact of the drug intervention is sufficient, and further action can be taken; if the coverage sufficiency assessment result is less than the coverage range data, supplementary measures are required. If the difference between the coverage range data and the actual coverage range of the intervention node is less than a preset threshold, the node is considered to have sufficient coverage and no further processing is needed. If the coverage sufficiency assessment data is lower than the coverage range data, the impact of the intervention node is insufficient, and a compensation mechanism is required. When the coverage of a drug intervention node is insufficient, an intelligent compensation mechanism is triggered. This mechanism enhances the impact of drug intervention on pathways by dynamically adjusting the scope of intervention nodes, ensuring coverage of all critical pathways. The influence of intervention nodes is amplified by adjusting the moderating factor (λ), thereby strengthening the impact on uncovered pathway nodes. The path morphology difference degree of the relationship chain is appropriately modified based on the degree of difference between drug intervention nodes and other pathways, enhancing control over path morphology. New intervention nodes are added in uncovered areas or pathways to ensure the integrity of the drug treatment chain.

[0125] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0126] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for clinical pharmacovigilance intelligence management, characterized in that, The method comprises the following steps: Step S1: Real-time acquisition of drug treatment chain full-factor data, including prescription instruction flow, drug preparation parameters, drug administration execution record and patient physiological feedback signal; Step S2: Constructing a drug treatment chain dynamic evaluation engine; utilizing the drug treatment chain dynamic evaluation engine to perform clinical drug mapping relationship matching on the drug treatment chain full-factor data, to generate an individualized drug use path graph; performing topological chain extraction on the individualized drug use path graph through digital twin technology, to generate a drug digital relationship chain; Step S3: Based on a multi-dimensional deviation detection algorithm, identifying drug treatment chain fracture risk points of the drug digital relationship chain, to obtain drug treatment abnormal risk points; Based on the drug treatment abnormal risk points, setting drug intervention nodes of the drug digital relationship chain, to obtain the drug intervention nodes; wherein, step S3 comprises the following steps: Step S31: Based on a multi-dimensional deviation detection algorithm, calculating relationship chain path form difference degrees of the drug digital relationship chain, to obtain the relationship chain path form difference degrees; Step S32: Based on the relationship chain path form difference degrees, performing path deviation interaction analysis on the drug digital relationship chain, to generate path deviation interaction data; Step S33: Utilizing the path deviation interaction data to identify path intersection points of the drug digital relationship chain, and marking the drug decision unit corresponding to the identified path intersection points as a drug treatment abnormal risk point; Step S34: Based on the drug treatment abnormal risk points, setting drug intervention nodes of the drug digital relationship chain, to obtain the drug intervention nodes; wherein, the multi-dimensional deviation detection algorithm is as follows: In the formula, is the relationship chain path form difference degree, is the total number of path nodes, is the geometric deviation weight, is the deviation degree of the first node in the geometric space, is the topological deviation weight, is the topological deviation degree of the first node, is the time deviation weight, is the deviation degree of the first node in the time dimension, is the attribute deviation weight, is the attribute information deviation degree of the first node. Step S4: Judging the coverage of the drug intervention nodes; based on the coverage of the drug intervention nodes, triggering an intelligent compensation mechanism, and through the intelligent compensation mechanism, performing drug dosage intelligent regulation and control on the drug treatment chain full-factor data, to execute clinical pharmacy wisdom management work.

2. The method for clinical pharmacovigilance management of claim 1, wherein, The construction of the drug treatment chain dynamic evaluation engine in step S2 comprises: Constructing a dynamic pharmacy ontology framework, including three dimensions of decision rules of drug-gene interaction, metabolic pathway coupling and drug administration device adaptation; Using path discretization modeling technology, decomposing the drug treatment chain full-factor data into drug decision units with time-dependent relationship; Connecting the drug decision units into a drug use path; based on fuzzy Petri net, performing real-time deduction on the drug use path, and dynamically updating the activation probability of each drug decision unit through the result of real-time deduction; Establishing a path credibility verification mechanism, when the matching degree of the newly generated drug use path with the preset historical successful case library is lower than 65%, triggering an artificial review process.

3. The method for clinical pharmacovigilance management of claim 2, wherein, The construction method of the dynamic pharmacy ontology framework comprises: Establishing a drug action four-tuple model, wherein the four-tuple includes chemical entities, target proteins, metabolic enzyme systems and transport carriers, and each element is associated with at least 3000 verified biomedical entities; performing decision rule adaptation on the drug action four-tuple model, to generate drug-gene interaction decision rules; Defining a clinical path constraint rule library, including three types of constraint conditions of contraindication conflict detection rules, drug administration timing optimization rules and equipment compatibility verification rules; performing decision rule adaptation on the clinical path constraint rule library, to generate metabolic pathway coupling decision rules; The drug property attenuation model is constructed, the drug property attenuation model is subjected to decision rule adaptation, and a drug delivery device adaptation decision rule is generated; and a dynamic pharmacology ontology framework is constructed according to the drug-gene interaction decision rule, the metabolic pathway coupling decision rule and the drug delivery device adaptation decision rule.

4. The method for clinical pharmacovigilance management of claim 2, wherein, The medication path is deduced in real time based on the fuzzy Petri net, and the activation probability of each medication decision unit is dynamically updated according to the result of real-time deduction, including: The path node activation threshold function is set, and the formula of the path node activation threshold function is as follows: In the formula, is a path node activation threshold, is a drug sensitivity coefficient, is a real-time blood drug concentration monitoring value, is a standard reference value; The path node activation threshold function is used for unit node activation probability calculation of the medication path, and the path unit node activation probability is obtained; According to the path unit node activation probability, the activation probability difference of adjacent nodes of the medication path is analyzed, and when the activation probability difference of adjacent nodes exceeds 40%, a pharmaceutical care marker point is automatically inserted; The pharmaceutical care marker point is used for difference path selection of the medication path, and the fuzzy Petri net is used for path conflict deduction of the difference path, and the result of real-time deduction is obtained; Based on the result of real-time deduction, the activation probability of the medication decision unit is dynamically adjusted.

5. The method for clinical pharmacovigilance management of claim 1, wherein, Step S4 includes the following steps: Step S41: judging the coverage amount of the drug intervention node; Step S42: based on the coverage amount of the drug intervention node, the node intervention range of the drug digital relationship chain is analyzed, and the node intervention range is generated; Step S43: the path propagation influence data is generated by analyzing the path propagation influence of the drug digital relationship chain through the node intervention range; the path propagation influence data is used for intelligent regulation of the drug use of the drug digital relationship chain to execute the clinical pharmacy wisdom management work.

6. The method for clinical pharmacovigilance management of claim 5, wherein, Step S41 includes: The coverage radius calculation formula is set, and the coverage radius calculation formula is as follows: wherein, a coverage radius of the drug intervention node, a coverage radius adjustment factor, a relationship chain path form difference degree; The coverage range of the drug intervention node is calculated through the coverage radius calculation formula, and the coverage range data is obtained; The coverage sufficiency of the drug intervention node is evaluated, and the coverage sufficiency evaluation data of the drug intervention node is generated; The coverage amount of the drug intervention node is judged, if the coverage sufficiency evaluation data is greater than or equal to the coverage range data, no processing is performed; if the coverage sufficiency evaluation data is less than the coverage range data, the intelligent compensation mechanism is triggered based on the coverage amount of the drug intervention node.

7. A system for clinical pharmacovigilance intelligence management, comprising: The system for clinical pharmacy wisdom management is used for executing the method for clinical pharmacy wisdom management, and the system includes: The data acquisition module is used for acquiring the whole element data of the drug treatment chain in real time, including the prescription instruction flow, the drug preparation parameter, the drug administration record and the patient physiological feedback signal; The relationship chain construction module is used for constructing the drug treatment chain dynamic evaluation engine; the drug treatment chain dynamic evaluation engine is used for matching the clinical medication mapping relationship of the whole element data of the drug treatment chain, generating an individualized medication path graph; and the digital twin technology is used for extracting the topological chain of the individualized medication path graph, generating a drug digital relationship chain; The risk point identification module is configured to identify a drug treatment chain breaking risk point of the drug digital relationship chain based on a multidimensional deviation detection algorithm, and obtain a drug treatment abnormality risk point; and set a drug intervention node of the drug digital relationship chain according to the drug treatment abnormality risk point, and obtain the drug intervention node. The drug amount compensation module is configured to judge a coverage amount of the drug intervention node; trigger an intelligent compensation mechanism based on the coverage amount of the drug intervention node, and perform intelligent regulation and control on all-factor data of the drug treatment chain through the intelligent compensation mechanism to execute a clinical pharmacy wisdom management task.

8. A storage medium having stored thereon a computer program, characterized in that The computer program is executed by a processor to implement the method for clinical pharmacy wisdom management according to any one of claims 1-6.

Images