A medicine monitoring and tracing method and system based on a tracing label

Abstract

This invention relates to the field of drug monitoring, specifically a drug monitoring and traceability method and system based on traceability labels. During the drug production stage, traceability labels are generated for each level of packaging, and traceability codes, anti-counterfeiting information, and basic drug information are linked and written into master data. At the production line, fixed scanning equipment completes the initial reading and consistency verification, enabling label activation or cancellation management. In the circulation and usage stages, various nodes collect public codes and business event information through terminals and upload them to the monitoring and traceability platform. The platform maintains the label's lifecycle from pending activation to dispensing, returned, and cancelled states. At high-risk nodes, concealed anti-counterfeiting information and electronic tag identification are collected simultaneously. A first monitoring indicator is constructed based on activation behavior, geographical distribution of scanning, and anti-counterfeiting consistency. A second monitoring indicator is constructed by combining state migration trajectories, cross-regional circulation, and return / cancellation behaviors. Finally, a comprehensive monitoring score is calculated, triggering early warnings from enterprises and regulators.

Description

Technical Field

[0001] This invention relates to the field of drug monitoring, specifically a drug monitoring and traceability method and system based on traceability tags. Background Technology

[0002] Current drug traceability systems generally adopt a technical approach of one item, one code, end-to-end scanning and reporting, and centralized platform storage. The common practice is to print one-dimensional barcodes, QR codes, or RFID tags on the surface of each level of drug packaging unit as the carrier of the drug traceability code. At the end of the production line, manufacturers assign a unique traceability code to each packaging unit and establish a parent-child relationship between packaging levels. Wholesalers, distribution centers, retail pharmacies, and medical institutions read the traceability code through scanning devices during receiving, dispatching, allocation, and returns, uploading events such as time, location, and business actions to the enterprise traceability system and industry collaboration platform. This enables the recording of drug flow and post-event traceability after quality problems occur. Furthermore, some systems also provide consumers or medical personnel with a scanning query function to verify basic drug information and approval qualifications, and support batch-by-batch recalls in the event of quality risks.

[0003] However, from the perspectives of labeling, anti-counterfeiting capabilities, and data reliability, existing technologies primarily use single, visible traceability codes as label carriers. The encoding rules and visual forms of these codes are public and fixed, making them easy to copy and transfer. Anti-counterfeiting technologies often rely on graphic anti-counterfeiting measures or adhesive sealing, which are relatively independent of the traceability system, lacking a close logical connection between the two. Existing systems typically only concern themselves with whether the traceability code itself can be read, lacking a systematic verification mechanism for the consistency between the traceability code and hidden anti-counterfeiting information and electronic label identification. They also fail to incorporate physical anti-counterfeiting features such as image features and digital signatures into the unified data model of the traceability platform. This leads to situations where, even if the surface traceability code originates from legitimate channels, the platform still struggles to promptly identify the risks of label authenticity and reuse in scenarios such as repackaging genuine packaging or copying genuine traceability codes and affixing them to counterfeit drugs.

[0004] From the perspective of traceability data collection and circulation process monitoring, existing technologies typically only achieve online detection of traceability code printing quality at the production line end, without introducing a strict first activation concept at the system level. The behavior of the traceability code from generation to its first use in circulation lacks verifiable records. Circulation and usage rely heavily on manual scanning, which can easily lead to missed scans, incorrect scans, or scanning only the outer box while neglecting the inner box during busy warehousing operations or complex business processes. This results in breaks or false continuity in the platform's recorded flow path. For high-risk business nodes (such as primary wholesale warehousing, high-value drug dispensing, and hospital dispensing windows), data collection often still relies solely on reading a single traceability code, without additionally collecting hidden anti-counterfeiting zone information or electronic tag information according to risk level requirements. Furthermore, a unified and configurable high-risk node verification strategy has not been established between the terminal and the platform. Therefore, existing traceability systems cannot guarantee at the collection level that each event record corresponds to a real, uncopyable, and untampered label entity.

[0005] From the perspective of platform-side data processing and risk monitoring capabilities, the mainstream implementation of existing technologies is to simply accumulate scanning events by label or batch, mainly used to query the historical flow of a certain label or to trace the chain of responsibility when problems occur. A few systems may make blacklists or threshold judgments based on simple rules such as the number of scans and regional distribution, but overall, they lack the ability to quantitatively model label behavior in multiple dimensions. For example, existing systems usually do not conduct systematic statistics and anomaly detection on the time interval from the first activation to reaching the key node, and rarely model information such as the spatial dispersion of the location where the same label is read, whether the state transition path violates the preset business logic, etc. Furthermore, they lack a mechanism to combine these multiple characteristics such as time anomaly, spatial dispersion, state behavior anomaly, and physical anti-counterfeiting consistency deviation into continuous risk indicators through clear mathematical relationships. As a result, regulators and enterprise quality management often have to rely on post-event spot checks and manual experience analysis to determine whether there are abnormal behaviors such as reverse circulation, bulk reselling, and frequent returns and resales, lacking real-time and quantitative comprehensive judgment means for hidden risks such as counterfeit drugs with genuine codes, label duplication, and packaging recycling and reuse. Summary of the Invention

[0006] The purpose of this invention is to provide a drug monitoring and traceability method and system based on traceability labels to solve the technical problems mentioned in the background.

[0007] Based on the above ideas, the present invention provides the following technical solution:

[0008] A drug monitoring and traceability method and system based on traceability tags, comprising:

[0009] S1. During the drug production stage, a traceability label containing a publicly visible area and a hidden anti-counterfeiting area is generated for each level of drug packaging unit. The publicly visible area carries a drug traceability code that conforms to national and industry standards, and the hidden anti-counterfeiting area carries anti-counterfeiting information that corresponds one-to-one with the drug traceability code. Electronic tags can be superimposed on some packaging. The drug traceability code, anti-counterfeiting information and electronic tags are associated with the basic drug information and written into the enterprise's traceability master data.

[0010] S2. After the production line is coded, each traceability label is read for the first time using a fixed scanning device. At the same time, the drug traceability code in the publicly visible area, the anti-counterfeiting information in the hidden anti-counterfeiting area, and the electronic tag identifier are collected. The consistency between the three is verified. When the verification is successful, the activation time and production line identifier are recorded and the tag status corresponding to the traceability label is set to activated. When the verification fails, the corresponding tag status is set to invalid.

[0011] S3. During the circulation and use of medicines, circulation nodes such as wholesale enterprises, distribution centers, retail pharmacies and medical institutions read the publicly visible area information of the traceability label through terminal equipment, collect information such as node type, business operation type, time and geographical location, and upload it to the monitoring traceability platform. The monitoring traceability platform updates the traceability label status based on the information. The label status includes at least one of the following: pending activation, activated, in transit, warehoused, dispensed, returned and invalid.

[0012] S4. At the preset high-risk business nodes, while the corresponding terminal completes the reading of the publicly visible area information, it collects the anti-counterfeiting information of the hidden anti-counterfeiting area and / or the electronic tag identification according to the strategy prompts of the platform side. The monitoring and traceability platform constructs a first monitoring indicator to characterize the authenticity and reuse risk of the traceability tag based on the activation behavior of the traceability tag, the geographical distribution of the code scanning and the consistency of physical anti-counterfeiting.

[0013] S5. The monitoring and traceability platform constructs a second monitoring indicator to characterize the risk of abnormal drug circulation behavior based on the status migration trajectory of the traceability label at each business node, cross-regional circulation, and return and cancellation behavior, and dynamically corrects the second monitoring indicator when the label status is updated.

[0014] S6. The monitoring and traceability platform calculates a comprehensive monitoring score based on the first monitoring indicator and the second monitoring indicator. When the comprehensive monitoring score exceeds a preset threshold, the corresponding traceability label is marked as a suspected anomaly, and a warning message is pushed to the enterprise quality management end, the regulatory end, and the terminal query interface to indicate that the drug has a risk of authenticity or abnormal circulation.

[0015] By introducing a composite traceability label with a publicly visible area and a concealed anti-counterfeiting area on the label side, and completing the consistency verification and activation status registration of the internal information of the label during the first coding on the production line, abnormal packaging with inconsistent label information and untraceable characteristics can be eliminated before the medicine leaves the production process, significantly reducing front-end risks such as genuine codes being mislabeled or codes not matching the goods. During circulation and use, information such as node type, business operation type, time, and geographical location is continuously collected by the terminal to monitor the traceability platform for lifecycle management of the label status, ensuring that every migration of the label status from pending activation, activated, in transit, in storage, dispensed, returned, to invalid is traceable. Furthermore, the joint reading of concealed anti-counterfeiting information and electronic tag identification is enforced at high-risk business nodes, and two types of monitoring indicators and a comprehensive monitoring score are established on the platform side to address authenticity risk and circulation risk, realizing the transformation from only post-event traceability to process monitoring and real-time early warning. Overall, this method improves the reliability of data on the label and the observability of the circulation process without changing the existing one-item-one-code architecture. This enables regulators and enterprise quality management to identify problematic batches and abnormal flow earlier, reducing the spread of quality incidents and the cost of handling them.

[0016] Specifically, after completing the consistency verification of the publicly visible area information, the hidden anti-counterfeiting area information, and the electronic tag identification of the traceability tag, the monitoring and traceability platform in S4 constructs the first monitoring indicator based on at least the following steps:

[0017] S4.1 The time anomaly parameter is used to characterize the deviation of the average time interval between the first activation of the traceability tag and the first reading at the preset high-risk node from the reference time interval.

[0018] S4.2. The spatial dispersion parameter is used to characterize the degree of dispersion of the spatial distance distribution between each reading location and the drug production location of the traceability tag in the most recent reading events, relative to the reference spatial distribution range.

[0019] S4.3. The physical anti-counterfeiting consistency deviation parameter is used to characterize the consistency score between the publicly visible area elements and the hidden anti-counterfeiting area elements obtained based on image recognition or digital signature comparison. The result is the degree of deviation from the full score and the result is normalized.

[0020] S4.4 The monitoring and traceability platform calculates the first monitoring index based on the time anomaly parameter, the spatial dispersion parameter, and the physical anti-counterfeiting consistency deviation parameter. This is achieved by normalizing the time and spatial deviations and non-linearly superimposing them, while simultaneously amplifying the physical anti-counterfeiting consistency deviation parameter. As the time anomaly, spatial anomaly, and physical anti-counterfeiting consistency deviation increase, the first monitoring index increases accordingly, thus characterizing the increased risk of authenticity and reuse of the corresponding traceability label.

[0021] By explicitly introducing time anomaly, spatial dispersion, and physical anti-counterfeiting consistency deviation parameters when constructing the first monitoring indicator, the previously difficult-to-quantify risks of label authenticity and reuse are decomposed into three dimensions with clear physical meaning: first, the time anomaly from activation to the first reading at a high-risk node; second, the dispersion and abrupt changes in the geographical location where the label is read; and third, the degree of consistency deviation between the publicly visible area and the hidden anti-counterfeiting area in image features or digital signatures. This parameter system can simultaneously capture multiple risk patterns, such as frequent scanning at abnormal locations within a short period, abnormal cross-regional transfer of labels, and alteration or incomplete copying of anti-counterfeiting areas. Compared to a rough judgment based solely on the number of scans or a single blacklist rule, using these three types of parameters to construct the first monitoring indicator allows the system to more precisely characterize the degree of anomaly of each traceability label in terms of time, space, and anti-counterfeiting features. This provides richer basic information for subsequent risk quantification and threshold determination, and also facilitates adaptive threshold adjustment based on historical data.

[0022] Specifically, the first monitoring indicator is quantitatively calculated in the following manner:

[0023] The time anomaly parameter and the spatial dispersion parameter are standardized according to their respective reference fluctuation ranges to obtain normalized time deviation and spatial deviation. The square roots of the normalized time deviation and the normalized spatial deviation are then superimposed to obtain a comprehensive deviation reflecting the overall degree of time and spatial deviation. The comprehensive deviation is then exponentially amplified according to the physical anti-counterfeiting consistency deviation parameter, and a nonlinear term of the physical anti-counterfeiting consistency deviation parameter is added. This ensures that when any of the time anomaly, spatial dispersion, or physical anti-counterfeiting consistency deviation increases, the first monitoring indicator increases nonlinearly, thereby enhancing the sensitivity to significant abnormal labels.

[0024] By standardizing the temporal anomaly and spatial dispersion parameters during the calculation of the first monitoring indicator, and then combining them non-linearly, while introducing amplification and non-linear superposition of the physical anti-counterfeiting consistency deviation parameter, this monitoring indicator has the following advantages: First, after standardization, the temporal and spatial characteristics of different drug varieties and delivery routes are unified into a comparable range, avoiding inconsistencies in parameter dimensions caused by differences in business rhythms, making the indicator more suitable for cluster analysis and cross-enterprise horizontal comparison. Second, the use of non-linear superposition and amplification ensures that the risk curve corresponding to the first monitoring indicator remains relatively smooth within a small disturbance range, exhibiting a certain robustness to occasional minor delays or positional shifts. However, when the temporal anomaly, spatial dispersion, or physical anti-counterfeiting deviation increases significantly, the indicator value rises sharply, thus exhibiting higher sensitivity to extreme risks such as malicious label duplication and batch substitution. Third, by comprehensively mapping the three types of deviation into a single first monitoring indicator, it facilitates the use of a unified threshold for authenticity determination and interception decisions within the system, reducing the computational complexity of the terminal and platform.

[0025] Specifically, in step S5, the monitoring and tracing platform updates the status of the tracing tag and constructs the second monitoring indicator based on the status transition event of the tracing tag, wherein:

[0026] S5.1 Obtain the time anomaly parameter mentioned in S4 to reflect the abnormal time intervals that occur during the circulation of the tag;

[0027] S5.2 Obtain the spatial dispersion parameter mentioned in S4 to reflect the degree of path anomaly of the tag during cross-regional circulation;

[0028] S5.3. The state behavior anomaly parameter characterizes the degree to which the number of abnormal transitions that violate the preset state transition constraints is higher than the normal transition frequency in the state transitions that occur in the traceability tag per unit time.

[0029] S5.4 The monitoring and traceability platform constructs a second monitoring indicator based on the time anomaly parameter, the spatial dispersion parameter, and the state behavior anomaly parameter to comprehensively reflect the degree of coupling between time anomalies, spatial anomalies, and state behavior anomalies, and to characterize the risk of abnormal behavior in drug distribution.

[0030] When constructing the second monitoring indicator, the temporal anomaly parameter and spatial dispersion parameter are combined with the state behavior anomaly parameter. This allows the monitoring and traceability platform to simultaneously assess the risk of abnormal circulation from two levels: when and where it occurs, and how the state changes. The state behavior anomaly parameter identifies behaviors that do not conform to the normal pharmaceutical supply chain logic by statistically analyzing the deviation between the number of abnormal migrations that violate preset state migration constraints per unit time and the normal migration frequency. These behaviors include repeated regressions of the drug's status from "delivered" to "in transit" or "in storage," subsequent circulation events occurring after the drug has been discarded, and repeated inbound and outbound transactions in multiple non-adjacent areas within a short period of time. When temporal anomaly and spatial dispersion are superimposed with this state behavior anomaly, the second monitoring indicator can highlight the complex risk scenarios of abnormal state changes occurring at abnormal times and locations. Compared to simple rules based solely on the number of state changes or cross-regional circulation, this is more effective in detecting hidden reverse circulation, frequent returns and resales, and gray distribution behaviors.

[0031] Specifically, the second monitoring indicator is constructed through the following relationship:

[0032] The time anomaly parameter and the state behavior anomaly parameter are multiplied to reflect the coupling effect when the abnormal time interval and the frequency of abnormal state migration are superimposed. At the same time, the spatial dispersion parameter and the state behavior anomaly parameter are combined in a nonlinear manner to reflect the amplification effect of cross-regional circulation anomalies on the overall risk under high-frequency abnormal state migration. Furthermore, a logarithmic or similar nonlinear function that enhances the state behavior anomaly parameter with larger values ​​is introduced, so that when the frequency of state behavior anomalies is high, the second monitoring indicator increases faster than linearly. Thus, in cases where drugs are frequently reverse-circulated, frequently returned, or recirculated after being discarded, higher weight is given to the relevant abnormal behaviors.

[0033] By incorporating the product of the temporal anomaly parameter and the state / behavior anomaly parameter into the calculation process, and nonlinearly combining the spatial dispersion parameter with the state / behavior anomaly parameter, and applying a nonlinear function to the state / behavior anomaly parameter whose increase accelerates with increasing value, this approach highlights high-risk scenarios that exhibit both significant cross-regional or temporal anomalies and frequent illegal state migrations. This results in these scenarios showing significantly amplified numerical characteristics in the second monitoring indicator. Furthermore, applying a stronger amplification effect than linear growth to cases with high state / behavior anomalies prevents high-frequency anomalies from being diluted by other normal behaviors, improving the ability to identify systemic risks such as long-term reverse circulation chains and cyclical return chains. Overall, this calculation method amplifies the expression of the coupling effect among temporal, spatial, and state / behavior anomalies, enabling the second monitoring indicator to characterize both single-dimensional anomalies and to keenly capture high-risk scenarios caused by the superposition of multi-dimensional anomalies.

[0034] Specifically, S6 further includes the following steps:

[0035] S6.1 After obtaining the first monitoring indicator and the second monitoring indicator, the monitoring and tracing platform regards the first monitoring indicator and the second monitoring indicator as components on two independent risk dimensions, and performs joint calculation on the two to obtain a comprehensive monitoring score. The comprehensive monitoring score increases monotonically as the first monitoring indicator and the second monitoring indicator increase.

[0036] S6.2 The monitoring and tracing platform compares the comprehensive monitoring score with a preset threshold.

[0037] S6.3 When the comprehensive monitoring score is lower than the preset threshold, the corresponding traceability label will be kept in a normal state, and its risk evolution trajectory will only be recorded in the background.

[0038] S6.4 When the comprehensive monitoring score is greater than or equal to the preset threshold, the corresponding traceability tag is marked as a suspected abnormal tag, and a multi-level early warning process is triggered on the enterprise quality management terminal, the regulatory terminal, and the terminal query interface.

[0039] By combining the first and second monitoring indicators as two independent risk dimensions, the system can establish a unified quantitative measurement framework between label authenticity and the risks of reuse and abnormal circulation behavior. The comprehensive monitoring score increases monotonically with the increase of both indicators, facilitating the platform's adoption of a unified risk threshold strategy: when the comprehensive monitoring score is below the threshold, risk trajectory recording and long-term trend analysis are only performed in the background to avoid interfering with normal business operations; when the comprehensive monitoring score reaches or exceeds the threshold, the label is automatically marked as suspected abnormal, and a tiered early warning process is simultaneously triggered on the enterprise quality management end, the regulatory end, and the terminal query interface. This design supports both simple binary judgment and can be extended to multi-level risk stratification management, which is conducive to regulatory agencies adopting differentiated inspection frequencies and control measures for drugs with different risk levels, improving the efficiency of regulatory resource utilization. At the same time, the comprehensive monitoring score, as a unified indicator, facilitates risk ranking and horizontal assessment across enterprises, regions, or product categories.

[0040] Specifically, the comprehensive monitoring score is determined through the following relationship:

[0041] The first and second monitoring indicators are used as two components of a two-dimensional risk vector. The length of the two-dimensional risk vector or its equivalent comprehensive value is calculated so that the comprehensive monitoring score can simultaneously reflect the overall level of label authenticity, reuse risk, and abnormal circulation behavior risk. The comprehensive monitoring score increases when any single risk increases, and increases even more when both risks increase simultaneously. It is used to quantify and determine the overall risk of counterfeit drugs with genuine codes, label duplication, packaging recycling and reuse, and abnormal circulation behavior at the label level.

[0042] By treating the first and second monitoring indicators as two components of a two-dimensional risk vector, and obtaining the comprehensive monitoring score by calculating the length of this risk vector or its equivalent comprehensive value, a natural normalization and unified measurement of the two types of risks can be achieved without introducing a large amount of subjective weighting. This approach has the following advantages: First, the geometric meaning is clear; when any single risk increases, the comprehensive monitoring score inevitably increases, and when both risks increase simultaneously, the increase in the comprehensive monitoring score is even more significant, meeting the need for heightened vigilance against dual anomalies in actual supervision. Second, because it does not rely on complex weight configurations, the system can maintain the comparability and interpretability of the scoring results without frequent adjustments to a large number of parameters when promoted across different drug varieties, companies, or regions. Third, compressing two types of risk information into a single comprehensive monitoring score simplifies the implementation of risk ranking, batch screening, and automatic interception rules, facilitating large-scale, automated monitoring and handling processes at both the enterprise and regulatory levels. Through this approach, the system can more accurately identify high-risk scenarios such as counterfeit drugs with genuine codes, label duplication, and packaging recycling at the label level, and issue timely warnings.

[0043] A drug monitoring and traceability system based on traceability tags, using the aforementioned drug monitoring and traceability method based on traceability tags.

[0044] The technical solution of the present invention may include the following beneficial effects:

[0045] This invention introduces a composite traceability label, comprising a publicly visible area, a concealed anti-counterfeiting area, and an optional electronic tag, without altering the overall one-item-one-code architecture. An initial consistency verification and activation mechanism is implemented at the production line, ensuring that each traceability label's anti-counterfeiting information is confirmed one-to-one before entering the distribution chain. This significantly reduces risks at the source, such as mislabeling, discrepancies between the code and the product, and pre-emptive label tampering. Furthermore, it upgrades existing coding, printing, and scanning infrastructure without major modifications, simply by adding an anti-counterfeiting layer or attaching an electronic tag. The upgrade cost is controllable, while the overall label authenticity verification capability is significantly improved, providing more reliable foundational data for subsequent monitoring and traceability.

[0046] On the monitoring and traceability platform, the risks of label authenticity and reuse are abstracted into a first monitoring indicator consisting of temporal anomaly, spatial dispersion, and deviation from physical anti-counterfeiting consistency. Cross-regional anomalies, reverse circulation, frequent returns, and recirculation after invalidation in the drug distribution process are abstracted into a second monitoring indicator consisting of temporal anomaly, spatial dispersion, and state behavior anomaly. A non-linear combination method is used to characterize the coupling effect of these three types of anomalies. Through this parameterized and index-based modeling, the system can amplify the risks when labels are repeatedly read at abnormal times and locations, or frequently appear on unreasonable state migration paths. Compared to methods relying solely on single scans, blacklist rules, or simple counting, this approach can more precisely and robustly identify hidden risk scenarios such as counterfeit drugs with genuine codes, label duplication, and packaging recycling and reuse, achieving a shift from qualitative experience-based judgment to a quantifiable and adjustable risk scoring mechanism.

[0047] By treating the aforementioned first and second monitoring indicators as two independent risk dimensions, a comprehensive monitoring score is calculated that monotonically increases with the simultaneous increase of both types of risks. Based on this comprehensive monitoring score, a unified risk threshold and multi-level early warning process are set, achieving an upgrade from post-event traceability and spot checks to a regulatory model of continuous monitoring and real-time early warning across the entire chain. When the comprehensive monitoring score is low, the system only accumulates risk trajectories in the background for medium- to long-term trend analysis. When the score exceeds the threshold, early warnings and intervention measures are automatically triggered in conjunction with the enterprise's quality management end, the regulatory end, and the terminal query interface. This avoids unnecessary interference with normal business operations and concentrates regulatory and inspection resources on the batches and labels with the highest risks. Thus, while ensuring drug safety and improving the timeliness of problem detection, it significantly improves the refinement of regulatory and enterprise risk control and the efficiency of resource utilization. Attached Figure Description

[0048] Figure 1 This is a flowchart of the drug monitoring and traceability method and system based on traceability tags according to the present invention.

[0049] Figure 2 This describes the specific steps of S4 in the present invention, which describes a drug monitoring and traceability method and system based on traceability tags.

[0050] Figure 3 This describes the specific steps of S5 in the present invention, which describes a drug monitoring and traceability method and system based on traceability tags.

[0051] Figure 4 This describes the specific steps of S6 in the present invention, which describes a drug monitoring and traceability method and system based on traceability tags. Detailed Implementation

[0052] Example 1

[0053] like Figure 1-4A drug monitoring and traceability method and system based on traceability tags, comprising:

[0054] S1. During the drug production stage, a traceability label containing a publicly visible area and a hidden anti-counterfeiting area is generated for each level of drug packaging unit. The publicly visible area carries a drug traceability code that conforms to national and industry standards, and the hidden anti-counterfeiting area carries anti-counterfeiting information that corresponds one-to-one with the drug traceability code. Electronic tags can be superimposed on some packaging. The drug traceability code, anti-counterfeiting information and electronic tags are associated with the basic drug information and written into the enterprise's traceability master data.

[0055] S2. After the production line is coded, each traceability label is read for the first time using a fixed scanning device. At the same time, the drug traceability code in the publicly visible area, the anti-counterfeiting information in the hidden anti-counterfeiting area, and the electronic tag identifier are collected. The consistency between the three is verified. When the verification is successful, the activation time and production line identifier are recorded and the tag status corresponding to the traceability label is set to activated. When the verification fails, the corresponding tag status is set to invalid.

[0056] S3. During the circulation and use of medicines, circulation nodes such as wholesale enterprises, distribution centers, retail pharmacies and medical institutions read the publicly visible area information of the traceability label through terminal equipment, collect information such as node type, business operation type, time and geographical location, and upload it to the monitoring traceability platform. The monitoring traceability platform updates the traceability label status based on the information. The label status includes at least one of the following: pending activation, activated, in transit, warehoused, dispensed, returned and invalid.

[0057] S4. At the preset high-risk business nodes, while the corresponding terminal completes the reading of the publicly visible area information, it collects the anti-counterfeiting information of the hidden anti-counterfeiting area and / or the electronic tag identification according to the strategy prompts of the platform side. The monitoring and traceability platform constructs a first monitoring indicator to characterize the authenticity and reuse risk of the traceability tag based on the activation behavior of the traceability tag, the geographical distribution of the code scanning and the consistency of physical anti-counterfeiting.

[0058] S5. The monitoring and traceability platform constructs a second monitoring indicator to characterize the risk of abnormal drug circulation behavior based on the status migration trajectory of the traceability label at each business node, cross-regional circulation, and return and cancellation behavior, and dynamically corrects the second monitoring indicator when the label status is updated.

[0059] S6. The monitoring and traceability platform calculates a comprehensive monitoring score based on the first monitoring indicator and the second monitoring indicator. When the comprehensive monitoring score exceeds a preset threshold, the corresponding traceability label is marked as a suspected anomaly, and a warning message is pushed to the enterprise quality management end, the regulatory end, and the terminal query interface to indicate that the drug has a risk of authenticity or abnormal circulation.

[0060] Pharmaceutical manufacturers equip their existing production lines with traceability label printing and affixing units, including inkjet printers, label attachers, online visual inspection cameras, and an industrial control computer. The industrial control computer connects to the company's internal traceability master data server via Ethernet. Before each batch of production begins, it retrieves the necessary drug traceability codes, anti-counterfeiting information, and electronic tag identifiers from the server. During printing, it simultaneously outputs both the publicly visible content (1D / 2D traceability code) and the hidden anti-counterfeiting content (encrypted QR code or anti-counterfeiting graphics). For high-value drugs or large packages, RFID or NFC tag affixing positions are reserved on the packaging surface. Electronic tag identifiers are written using a reader / writer, and the correspondence between the traceability code, anti-counterfeiting information, and electronic tag identifiers is transmitted back to the traceability master data server in real time, forming a label master data table with the batch number as the primary key.

[0061] Fixed barcode scanning devices are installed at both the upstream and downstream ends of the production line: the barcode scanning gate at the upstream end is mainly used to verify the label printing quality and readability, while the barcode scanning camera at the downstream end works with the industrial control computer to perform an "initial activation" on each package. The industrial control computer calls the enterprise traceability master data interface to compare the publicly available traceability code, anti-counterfeiting zone decoding result, and electronic tag reading obtained from the barcode scanning with pre-generated records in the server. If all three correspond and are unused, the tag status field of the corresponding record is changed from "pending activation" to "activated," and the activation time, production line number, and operator number are written in; otherwise, the tag status of the record is directly set to "invalid," and an alarm is triggered on the production line to prompt the rejection of the package.

[0062] In the distribution and usage stages, wholesale enterprises, distribution centers, retail pharmacies, and medical institutions are all equipped with handheld scanning terminals running Android or dedicated operating systems. These handheld terminals are pre-installed with traceability data collection applications, supporting camera scanning and NFC / RFID reading. When completing business operations (receiving, outbound, return, allocation, dispensing), the terminals automatically collect information such as the current node type, business operation type, terminal geographic coordinates (obtained via GPS or base station positioning), and timestamps. This information is then reported to the central monitoring and traceability platform via a dedicated network or mobile network. Upon receiving an event, the event processing module of the monitoring and traceability platform queries the corresponding tag record based on the node type and business operation type in the event. It then converts the status field between predefined states such as "activated," "in transit," "in storage," "dispensed," "returned," and "invalid," and records the status transition time and source node ID to form a complete status transition trajectory.

[0063] At pre-defined high-risk business nodes (such as warehousing at primary wholesale enterprises, dispensing of high-value drugs, and dispensing windows in hospital inpatient pharmacies), the terminal application is configured in a mandatory anti-counterfeiting verification mode. This means that after scanning the publicly available traceability code, a prompt will appear requiring the operator to scratch off the hidden anti-counterfeiting area and scan the code again, or to read the electronic tag via NFC / RFID. The terminal application will then report both the anti-counterfeiting information and the electronic tag. The platform's tag verification module will perform consistency checks on the publicly available traceability code, anti-counterfeiting information, and electronic tag, and update the corresponding first monitoring indicator, second monitoring indicator, and comprehensive monitoring score in real time.

[0064] Specifically, after completing the consistency verification of the publicly visible area information, the hidden anti-counterfeiting area information, and the electronic tag identification of the traceability tag, the monitoring and traceability platform in S4 constructs the first monitoring indicator based on at least the following steps:

[0065] S4.1 The time anomaly parameter is used to characterize the deviation of the average time interval between the first activation of the traceability tag and the first reading at the preset high-risk node from the reference time interval.

[0066] S4.2. The spatial dispersion parameter is used to characterize the degree of dispersion of the spatial distance distribution between each reading location and the drug production location of the traceability tag in the most recent reading events, relative to the reference spatial distribution range.

[0067] S4.3. The physical anti-counterfeiting consistency deviation parameter is used to characterize the consistency score between the publicly visible area elements and the hidden anti-counterfeiting area elements obtained based on image recognition or digital signature comparison. The result is the degree of deviation from the full score and the result is normalized.

[0068] S4.4 The monitoring and traceability platform calculates the first monitoring index based on the time anomaly parameter, the spatial dispersion parameter, and the physical anti-counterfeiting consistency deviation parameter. This is achieved by normalizing the time and spatial deviations and non-linearly superimposing them, while simultaneously amplifying the physical anti-counterfeiting consistency deviation parameter. As the time anomaly, spatial anomaly, and physical anti-counterfeiting consistency deviation increase, the first monitoring index increases accordingly, thus characterizing the increased risk of authenticity and reuse of the corresponding traceability label.

[0069] Specifically, the first monitoring indicator X is calculated using the following formula:

[0070] ;

[0071] σ a σ bThe preset positive scaling parameter is used to characterize the reference fluctuation range of the time anomaly parameter a and the spatial dispersion parameter b;

[0072] λ is a positive amplification factor used to adjust the influence of the physical anti-counterfeiting consistency deviation parameter c on the first monitoring indicator X;

[0073] 'a' is the time anomaly parameter;

[0074] b is the spatial dispersion parameter;

[0075] c represents the deviation parameter of the physical anti-counterfeiting consistency.

[0076] When the deviation of a and b from the reference value 1 increases or c increases, the first monitoring indicator X increases accordingly, which is used to indicate that the authenticity risk and reuse risk of the corresponding traceability label increases.

[0077] The monitoring and traceability platform is equipped with a tag risk feature calculation module, which periodically pulls event records for each tag from the event log database for a recent period of time to calculate time anomaly parameters, spatial dispersion parameters, and physical anti-counterfeiting consistency deviation parameters.

[0078] For the time anomaly parameter 'a', the platform first calculates the distribution of time intervals from the first activation of the tag to the first scan at a specified high-risk node (e.g., a provincial wholesale enterprise warehouse) based on product variety, sales channels, and historical normal circulation data, and then calculates the mean of this distribution as a reference time interval Δt. ref Subsequently, for each tag, the actual time interval Δt from activation to the first scan at a high-risk node is taken, and the time anomaly parameter a = Δt / Δt is defined. ref For tags that have not yet reached high-risk nodes, the value of the current time minus the activation time can be substituted into Δt, and the tag can be marked on the risk page as being in transit and not yet having completed the reference path.

[0079] For the spatial dispersion parameter b, the platform maintains a node geographic information table, recording the latitude and longitude information of each production plant, wholesale enterprise, distribution center, retail pharmacy, and medical institution. For each tag, the platform takes its most recent N (e.g., N=10) scanning events and calculates the set of great circle distances {d1, d2, ..., dN} between these N locations and the corresponding production plant location. The standard deviation of this set, std(d), is used as the actual spatial dispersion. Then, a certain empirical reference value d_ref (e.g., the median standard deviation based on normal samples) is used for standardization, defining b = std(d) / d ref If the tag has not yet left the factory or the number of events is insufficient, the parameter b can be set to 0 by default or a lower value, and will be updated after subsequent events accumulate.

[0080] For the physical anti-counterfeiting consistency deviation parameter c, during the anti-counterfeiting label generation stage, the manufacturer stores the standard image feature vector or standard digital signature of each anti-counterfeiting QR code in the anti-counterfeiting feature library. The hidden anti-counterfeiting area image uploaded by high-risk nodes during anti-counterfeiting verification undergoes preprocessing (grayscale conversion, binarization, morphological operations, etc.) and feature extraction (such as local feature descriptors, texture features, or verification values ​​obtained directly from decrypting the anti-counterfeiting QR code) by the platform's image processing submodule. This is compared with the corresponding standard features to obtain a similarity score s in the range of 0 to 1. The physical anti-counterfeiting consistency deviation c can be defined as c = 1s, or the result after linear or nonlinear normalization, making c close to 0 under normal consistency conditions and significantly increasing when the image is altered, incompletely copied, or the anti-counterfeiting QR code is damaged. The calculation results of the above three parameters will serve as input for the subsequent calculation of the first monitoring indicator X.

[0081] Specifically, in step S5, while updating the status of the traceability label, the monitoring and traceability platform constructs the second monitoring indicator based on the status transition event of the traceability label, wherein: S5.1, the time anomaly parameter in step S4 is obtained to reflect the situation where the label has an abnormal time interval during circulation;

[0082] S5.2 Obtain the spatial dispersion parameter mentioned in S4 to reflect the degree of path anomaly of the tag in the cross-regional circulation process; S5.3 Use the state behavior anomaly parameter to characterize the degree of abnormal migration frequency that violates the preset state migration constraint in the state migration of the traceable tag within a unit time, relative to the normal migration frequency.

[0083] S5.4 The monitoring and traceability platform constructs a second monitoring indicator based on the time anomaly parameter, the spatial dispersion parameter, and the state behavior anomaly parameter to comprehensively reflect the degree of coupling between time anomalies, spatial anomalies, and state behavior anomalies, and to characterize the risk of abnormal behavior in drug distribution.

[0084] Specifically, the second monitoring indicator is constructed using the following formula:

[0085] ;

[0086] 'a' is the time anomaly parameter;

[0087] b is the spatial dispersion parameter;

[0088] d is the abnormality parameter of the state behavior;

[0089] When d increases and simultaneously and abnormally superimposed with a and b, the first term This increases significantly, serving to comprehensively reflect the coupling effect of temporal and spatial anomalies during the migration process of anomalous states;

[0090] Second item As d increases monotonically, it is used to nonlinearly amplify the migration of high-frequency abnormal states. Overall, this makes the second monitoring indicator Y larger when the state behavior is more abnormal, which indicates that the risk of abnormal behaviors such as cross-regional circulation, reverse circulation, frequent returns, and recirculation after being discarded is higher.

[0091] In one specific embodiment, the monitoring and traceability platform internally implements a tag state machine management module. This module predefines states such as pending activation, activated, in transit, in storage, dispensed, returned, and invalid, as well as the allowed legal transition relationships between these states. For example, a state can transition from activated to in transit, from in transit to in storage, and from in storage to dispensed or returned. However, a state cannot transition from invalid to any other state, and a state cannot be rolled back from dispensed to in storage. Each time a new business event is received, the module determines the target state based on the current state and the event type, and then checks whether the state transition violates the preset state transition constraints.

[0092] For each tag, the platform counts its state transition events within a given time window (e.g., the last 30 days): recording the number of valid transitions N. normal and the number of abnormal migrations N abnormal Simultaneously, the duration T of this time window is recorded. The state behavior anomaly parameter d can be defined as:

[0093] ;

[0094] Where f ref For reference migration frequency, it can be determined based on the average migration frequency of similar normal tags. If a tag exhibits behaviors such as multiple invalidation events still occurring in transit or upon warehousing within a given time window, returns after dispensing followed by re-dispensing, or repeated warehousing and dispensing in different areas within a short period, then N... abnormal As the time anomaly increases significantly, d also increases. The time anomaly parameter a and the spatial dispersion parameter b reuse the calculation method defined above and are periodically updated by the label risk feature calculation module. During each window recalculation, the label risk scoring module writes a, b, and d together into the risk feature table, providing input for the calculation of the second monitoring indicator Y.

[0095] Specifically, S6 further includes the following steps:

[0096] S6.1. After obtaining the first monitoring metric and the second monitoring metric, the monitoring and tracing platform regards the first monitoring metric and the second monitoring metric as components in two independent risk dimensions, performs a joint operation on them to obtain a comprehensive monitoring score, and the comprehensive monitoring score increases monotonically with the increase of the first monitoring metric and the second monitoring metric;

[0097] S6.2. The monitoring and tracing platform compares the comprehensive monitoring score with a preset threshold;

[0098] S6.3. When the comprehensive monitoring score is lower than the preset threshold, the corresponding tracing label is kept in the normal state, and only its risk evolution track is recorded in the background;

[0099] S6.4. When the comprehensive monitoring score is greater than or equal to the preset threshold, the corresponding tracing label is marked as a suspected abnormal label, and a multi-level early warning process for the enterprise quality management end, the supervision end, and the terminal query interface is triggered.

[0100] Specifically, the comprehensive monitoring score Z is determined through the following relationship:

[0101] ;

[0102] X is the first monitoring metric mentioned above;

[0103] Y is the second monitoring metric;

[0104] Z is the comprehensive monitoring score based on the risks of label authenticity and abnormal circulation, and its value increases monotonically with the increase of X and Y, and is used to quantitatively measure the overall risk and determine the threshold for true-code fake drugs, label replication, packaging recycling and reuse, and abnormal circulation behaviors at the label level.

[0105] Before deploying in the production environment, enterprises usually establish a threshold calibration process using historical data. This process includes two stages: First, based on the historical confirmed risk labels and normal label samples, calculate the X, Y, and the corresponding Z for each sample, and draw the recall rate and false alarm rate curves at different Z thresholds, or directly draw the ROC curve; Second, select a compromise point according to regulatory requirements. For example, it is desired that the recognition rate of known risk labels is not less than 90%, and at the same time, the false alarm rate is controlled within 5%, and the corresponding Z th threshold is obtained and written into the module configuration. During operation, the comprehensive risk scoring and early warning module compares the Z of each label with Z th . If Z < Z_th, only record the current risk score and risk history track of this label in the database, and do not trigger a foreground alarm; if Z ≥ Z thThen, the event push submodule is invoked to send a suspected anomaly label notification to the enterprise quality management system, the regulatory platform interface, and the terminal-oriented query interface. This notification includes the label identifier, the current X, Y, and Z values, the main anomaly characteristics, and suggested handling measures.

[0106] In addition, this module also supports multi-level threshold strategies, such as configuring Z... th1 Z th2 Two thresholds are used: labels with Z between the two thresholds are marked as moderate risk, and labels with Z above Z are marked as... th2 The label is marked as high risk so that businesses and regulators can allocate inspection resources and process priorities according to risk level.

[0107] A drug monitoring and traceability system based on traceability tags, using the aforementioned drug monitoring and traceability method based on traceability tags.

Claims

1. A drug monitoring and traceability method based on traceability labels, characterized in that, Includes the following steps: S1. During the drug production stage, a traceability label containing a publicly visible area and a hidden anti-counterfeiting area is generated for each level of drug packaging unit. The publicly visible area carries a drug traceability code that conforms to national and industry standards, and the hidden anti-counterfeiting area carries anti-counterfeiting information that corresponds one-to-one with the drug traceability code. Electronic tags can be superimposed on some packaging. The drug traceability code, anti-counterfeiting information and electronic tags are associated with the basic drug information and written into the enterprise's traceability master data. S2. After the production line is coded, each traceability label is read for the first time using a fixed scanning device. At the same time, the drug traceability code in the publicly visible area, the anti-counterfeiting information in the hidden anti-counterfeiting area, and the electronic tag identifier are collected. The consistency between the three is verified. When the verification is successful, the activation time and production line identifier are recorded and the tag status corresponding to the traceability label is set to activated. When the verification fails, the corresponding tag status is set to invalid. S3. During the circulation and use of medicines, wholesale enterprises, distribution centers, retail pharmacies and medical institutions read the publicly visible area information of the traceability label through terminal equipment, collect node type, business operation type, time and geographical location information and upload it to the monitoring traceability platform. The monitoring traceability platform updates the traceability label status based on the information. The label status includes at least one of the following: pending activation, activated, in transit, warehoused, dispensed, returned and invalid. S4. At the preset high-risk business nodes, while the corresponding terminal completes the reading of the publicly visible area information, it collects the anti-counterfeiting information of the hidden anti-counterfeiting area and / or the electronic tag identification according to the strategy prompts of the platform side. The monitoring and traceability platform constructs a first monitoring indicator to characterize the authenticity and reuse risk of the traceability tag based on the activation behavior of the traceability tag, the geographical distribution of the code scanning and the consistency of physical anti-counterfeiting. S5. The monitoring and traceability platform constructs a second monitoring indicator to characterize the risk of abnormal drug circulation behavior based on the status migration trajectory of the traceability label at each business node, cross-regional circulation, and return and cancellation behavior, and dynamically corrects the second monitoring indicator when the label status is updated. S6. The monitoring and traceability platform calculates a comprehensive monitoring score based on the first monitoring indicator and the second monitoring indicator. When the comprehensive monitoring score exceeds a preset threshold, the corresponding traceability label is marked as suspected abnormal, and a warning message is pushed to the enterprise quality management end, the regulatory end and the terminal query interface to indicate that the drug has the risk of authenticity or abnormal circulation. in: After completing the consistency verification of the publicly visible area information, the hidden anti-counterfeiting area information, and the electronic tag identification of the traceability label, the monitoring and traceability platform in S4 constructs the first monitoring indicator based on at least the following steps: S4.1 The time anomaly parameter is used to characterize the deviation of the average time interval between the first activation of the traceability tag and the first reading at the preset high-risk node from the reference time interval. S4.

2. The spatial dispersion parameter is used to characterize the degree of dispersion of the spatial distance distribution between each reading location and the drug production location of the traceability tag in the most recent reading events, relative to the reference spatial distribution range. S4.

3. The physical anti-counterfeiting consistency deviation parameter is used to characterize the consistency score between the publicly visible area elements and the hidden anti-counterfeiting area elements obtained based on image recognition or digital signature comparison. The result is the degree of deviation from the full score and the result is normalized. S4.4 The monitoring and tracing platform calculates the first monitoring index based on the time anomaly parameter, the spatial dispersion parameter, and the physical anti-counterfeiting consistency deviation parameter by normalizing the time and spatial deviations and non-linearly superimposing them, while amplifying the physical anti-counterfeiting consistency deviation parameter. In S5, while updating the status of the traceability tag, the monitoring and tracing platform constructs the second monitoring indicator based on the status transition event of the traceability tag, wherein: S5.1 Obtain the time anomaly parameter mentioned in S4 to reflect the abnormal time intervals that occur during the circulation of the tag; S5.2 Obtain the spatial dispersion parameter mentioned in S4 to reflect the degree of path anomaly of the tag during cross-regional circulation; S5.

3. The state behavior anomaly parameter characterizes the degree to which the number of abnormal transitions that violate the preset state transition constraints is higher than the normal transition frequency in the state transitions that occur in the traceability tag per unit time. S5.4 The monitoring and traceability platform constructs a second monitoring indicator based on the time anomaly parameter, the spatial dispersion parameter, and the state behavior anomaly parameter to comprehensively reflect the degree of coupling between time anomalies, spatial anomalies, and state behavior anomalies, and to characterize the risk of abnormal behavior in drug distribution.

2. The drug monitoring and traceability method based on traceability tags according to claim 1, characterized in that, The first monitoring indicator is quantitatively calculated in the following manner: The time anomaly parameter and the spatial dispersion parameter are standardized according to their respective reference fluctuation ranges to obtain normalized time deviation and spatial deviation. The squares of the normalized time deviation and the normalized spatial deviation are then superimposed and the square root is calculated to obtain a comprehensive deviation that reflects the overall degree of time and spatial deviation. The comprehensive deviation is then exponentially amplified according to the physical anti-counterfeiting consistency deviation parameter, and the nonlinear term of the physical anti-counterfeiting consistency deviation parameter is added.

3. The drug monitoring and traceability method based on traceability labels according to claim 1, characterized in that, The second monitoring metric is constructed through the following relationship: The time anomaly parameter and the state behavior anomaly parameter are multiplied together, and the spatial discretization parameter and the state behavior anomaly parameter are combined in a nonlinear manner; and a logarithmic nonlinear function is introduced into the state behavior anomaly parameter.

4. The drug monitoring and traceability method based on traceability tags according to claim 3, characterized in that, S6 further includes the following steps: S6.1 After obtaining the first monitoring indicator and the second monitoring indicator, the monitoring and tracing platform regards the first monitoring indicator and the second monitoring indicator as components on two independent risk dimensions, and performs joint calculation on the two to obtain a comprehensive monitoring score. The comprehensive monitoring score increases monotonically as the first monitoring indicator and the second monitoring indicator increase. S6.2 The monitoring and tracing platform compares the comprehensive monitoring score with a preset threshold. S6.3 When the comprehensive monitoring score is lower than the preset threshold, the corresponding traceability label will be kept in a normal state, and its risk evolution trajectory will only be recorded in the background. S6.4 When the comprehensive monitoring score is greater than or equal to the preset threshold, the corresponding traceability tag is marked as a suspected abnormal tag, and a multi-level early warning process is triggered on the enterprise quality management terminal, the regulatory terminal, and the terminal query interface.

5. A drug monitoring and traceability method based on traceability tags according to claim 4, characterized in that, The comprehensive monitoring score is determined through the following relationship: ； X is the first monitoring indicator mentioned above; Y represents the second monitoring indicator; Z represents a comprehensive monitoring score based on the risks of label authenticity and abnormal circulation.

6. A drug monitoring and traceability system based on traceability tags, characterized in that, Use the drug monitoring and traceability method based on traceability tags as described in any one of claims 1-5.

Images