Infusion bag management system and method and infusion cabinet

Abstract

The invention relates to an infusion bag management system and method and an infusion cabinet, and belongs to the technical field of medical information. Comprising a three-dimensional identification module used for obtaining position information of each infusion bag in an infusion cabinet; according to the information, obtaining imaging data of particulate suspended matters in the infusion bag; the infusion bag monitoring module is used for monitoring liquid in the infusion bag and identifying the internal state of the current infusion bag according to the monitoring result and the imaging data; marking state data of the current infusion bag according to the internal state; the verification module is used for performing encryption uplink and multi-party verification on the state data; the sorting module is used for obtaining the verification result and the state data and carrying out sorting operation; the data management and tracing module is used for storing related data, verification results and sorting data; encryption tracing of the data is completed by constructing a private chain. The stability of the medicine storage environment, real-time monitoring of the state of the infusion bag, patient identity verification, instant updating of inventory data and medicine standardized management are achieved.

Description

Technical Field

[0001] This invention relates to the field of medical information technology, and in particular to an infusion bag management system, method, and infusion cabinet. Background Technology

[0002] In modern healthcare, the storage, management, and use of medications are crucial for ensuring patient treatment outcomes and safety. However, existing traditional equipment and technologies have many shortcomings in practical applications, making it difficult to meet the ever-increasing healthcare demands.

[0003] Traditional RFID smart medicine cabinets only provide basic medication storage and retrieval records, making their functionality limited and unable to meet the higher demands of modern medical settings for monitoring the storage environment and status of medications. Specifically, traditional cabinets lack temperature and humidity control, failing to monitor and automatically adjust the internal temperature and humidity in real time, thus hindering the stability of the storage environment. This poses a safety hazard for medications with stringent environmental requirements, such as those requiring cold chain logistics. Furthermore, traditional cabinets lack IV bag deformation monitoring, making it impossible to detect leaks promptly, which could negatively impact patient treatment outcomes.

[0004] Barcode-based infusion management systems also present several problems in practical applications. First, the system requires manual barcode scanning, a time-consuming process that adds an average of 30% to the overall operation time, severely impacting the efficiency of medical staff. Second, barcode scanning is prone to errors or omissions, making it difficult to guarantee the accuracy of medication information verification and posing certain safety risks. Traditional infusion cabinets use paper labels combined with manual verification for patient identification, relying solely on single information such as patient name or bed number. In cases of duplicate names or similar cases, the risk of misoperation increases significantly, with an error rate exceeding 0.3%. This method of identity verification cannot effectively guarantee the safety of patient medication and can easily lead to medical disputes.

[0005] Furthermore, traditional infusion cabinets lack environmental sensor components, making it impossible to monitor the temperature and humidity inside the cabinet in real time. For medications with strict storage temperature requirements, such as cold chain drugs, the cabinet cannot issue timely warnings when the storage temperature deviation exceeds ±3℃, potentially leading to drug deterioration and ineffectiveness. Simultaneously, traditional infusion cabinets cannot monitor the deformation of infusion bags, making it difficult to detect expired or deteriorated medications in a timely manner, posing certain safety hazards. Medication storage and retrieval records for traditional infusion cabinets require manual registration, resulting in significant delays in inventory data updates, typically requiring 4-6 hours. This lagging inventory management model leads to an increase in emergency medication shortage rates of over 12%, particularly in departments with high requirements for timely drug supply, such as emergency departments, where this problem is more pronounced and seriously affects patient treatment outcomes.

[0006] Traditional infusion cabinets lack biometric technology, and medication dispensing permissions are managed solely through mechanical locks. This management model has significant loopholes, making it prone to unauthorized medication dispensing and drug loss. Statistics show that the average annual loss rate in tertiary hospitals can reach 0.8%, causing unnecessary economic losses and posing a challenge to the standardization of drug management. Summary of the Invention

[0007] Therefore, the technical problem to be solved by the present invention is to overcome the shortcomings of the prior art, such as single function, lack of temperature and humidity linkage control and infusion bag deformation monitoring function, cumbersome operation, low verification accuracy, easy error in patient identification method, lack of environmental monitoring and drug status monitoring, lagging inventory management and non-standard drug requisition authority management.

[0008] In a first aspect, to solve the above-mentioned technical problems, the present invention provides an infusion bag management system, comprising:

[0009] The three-dimensional recognition module includes a positioning submodule and an imaging submodule; the positioning submodule is used to acquire the position information of each infusion bag in the infusion cabinet; the imaging submodule is used to acquire imaging data of particulate matter suspended inside each infusion bag based on the position information.

[0010] The infusion bag monitoring module includes a monitoring submodule and a marking submodule; the monitoring submodule is used to monitor the liquid in the infusion bag and identify the current internal state of the infusion bag based on the monitoring results and the imaging data; the marking submodule is used to mark the current state data of the infusion bag based on the internal state.

[0011] The verification module is used to encrypt and upload the status data to the blockchain and perform multi-party verification to obtain the verification result;

[0012] The sorting module is used to acquire the verification results and status data in real time and sort the infusion bags in the infusion cabinet.

[0013] The data management and traceability module is used to store the imaging data, the status data, the verification results, and the sorting data in the sorting module; and to complete the encrypted traceability of the sorting data by constructing a private chain.

[0014] In one embodiment of the present invention, the sorting module includes:

[0015] The medical order driver encapsulation submodule is used to parse electronic medical orders, automatically generate drug mixing schemes, and drive servo motors to complete quantitative dispensing to obtain packaged drugs;

[0016] The grabbing submodule is used for grabbing medicines during sorting and for grabbing the packaged medicines.

[0017] In one embodiment of the present invention, the medical order-driven encapsulation submodule includes a packaging unit, which is used to construct a medical knowledge graph-driven model and automatically generate drug formulation schemes.

[0018] In one embodiment of the present invention, the grasping submodule includes a quantum dot spectral recognition unit and an execution unit. The quantum dot spectral recognition unit is used to acquire drug information and identify the drug to be grasped in the infusion cabinet based on the drug information, and drive the execution unit to grasp the drug.

[0019] In one embodiment of the present invention, the positioning submodule includes a phase extraction unit, the phase extraction unit comprising:

[0020] A ranging setting component is used to set a ranging range and determine the phase change range of the subcarrier at this time based on the ranging range.

[0021] A conversion component is used to perform bandpass sampling and conversion on the transmitted signal and the received signal according to the phase change range, and then send them into the digital domain;

[0022] A distance calculation component is used to estimate the phase of the subcarrier component in the transmit / receive signal sent by the conversion component, and calculate the distance between the tag on the infusion bag and the tag reading device based on the phase.

[0023] A positioning component is used to determine the position information of the label on the infusion bag based on the distance.

[0024] In one embodiment of the present invention, the monitoring submodule includes:

[0025] A density detection unit is used to obtain the density and viscosity of a liquid by measuring the deviation of the frequency.

[0026] A contamination detection unit is used to acquire spectral information of liquids in the terahertz band; extract feature information related to microbial contamination based on the spectral information; and calculate the microbial contamination index of the liquid based on the feature information.

[0027] The identification unit is used to extract key parameters of the density, viscosity, and microbial contamination index; and to establish an identification model based on the key parameters.

[0028] In one embodiment of the present invention, the monitoring submodule further includes an environmental sensing unit, the environmental sensing unit comprising:

[0029] Temperature and humidity detection components are used to monitor the temperature and humidity inside the medicine cabinet in real time;

[0030] A pressure detection component is used to monitor the air pressure inside the medicine cabinet in real time.

[0031] A light detection component is used to monitor the light intensity inside the medicine cabinet in real time.

[0032] An alarm component is used to determine whether there are any abnormalities in the temperature and humidity, air pressure, light intensity, and liquid in the infusion bag. When an abnormality is detected, an emergency alarm signal is issued.

[0033] In one embodiment of the present invention, the control platform interacts bidirectionally with the hospital HIS / intravenous compounding center system, supporting hierarchical management of access permissions for special drugs.

[0034] Secondly, to solve the above-mentioned technical problems, the present invention provides a method for managing infusion bags, comprising:

[0035] Obtain the position information of each infusion bag in the infusion cabinet; based on the position information, obtain imaging data of particulate matter suspended inside each infusion bag;

[0036] The liquid in the infusion bag is monitored, and the current internal state of the infusion bag is identified based on the monitoring results and the imaging data.

[0037] Based on the internal state, mark the current status data of the infusion bag; encrypt and upload the status data to the blockchain and perform multi-party verification to obtain the verification result;

[0038] Obtain the verification results and status data, sort the infusion bags in the infusion cabinet, and record the sorting data;

[0039] The imaging data, status data, verification results, and sorting data are stored, and the sorting data is encrypted and traceable by constructing a private chain.

[0040] Thirdly, in order to solve the above-mentioned technical problems, the present invention provides an infusion cabinet, including the infusion bag management system described above.

[0041] Compared with the prior art, the above-described technical solution of the present invention has the following advantages:

[0042] (1) The infusion bag management system, method, and infusion cabinet described in this invention, through the efficient cooperation of the positioning submodule and the imaging submodule, can accurately acquire the location and internal status information of the infusion bag and update this data in real time, enabling medical staff to keep track of the infusion bag status at any time, thereby significantly improving work efficiency and safety. The monitoring submodule monitors the physical and chemical properties of the infusion bag liquid in real time, and, combined with imaging data, comprehensively assesses the internal status of the infusion bag, providing data support for precise management. The marking submodule automatically marks the status of the infusion bag based on the monitoring results, which not only facilitates subsequent management and traceability but also effectively reduces human operation errors. The verification module uses blockchain technology to encrypt and upload the status data to the chain, ensuring the security and immutability of the data, while supporting a multi-party verification mechanism to further ensure the authenticity and consistency of the data. The sorting module quickly and accurately classifies and processes the infusion bags according to the verification results and status data, reducing manual intervention, lowering operational errors, and improving overall work efficiency. The data management and traceability module centrally stores various types of data, ensuring the integrity and traceability of the data, facilitating subsequent analysis and management. By using private blockchain technology to encrypt and trace sorting data, the overall reliability of the system is further improved. Through the collaborative work of the above modules, high-precision identification, comprehensive monitoring, intelligent tagging, security verification, and efficient sorting of infusion bags are achieved, while traceability capabilities are enhanced and resources and costs are saved.

[0043] (2) This invention effectively solves the problems existing in the prior art by using functions such as temperature and humidity linkage control, infusion bag deformation monitoring, biometric technology, real-time environmental monitoring, accurate inventory management and secure drug requisition authority management. This enables the stability and safety of the drug storage environment, real-time monitoring of the infusion bag status, accurate verification of patient identity, instant updates of inventory data and standardization of drug management, thereby significantly improving medical efficiency and patient medication safety. Attached Figure Description

[0044] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings, wherein:

[0045] Figure 1 This is a structural diagram of an infusion bag management system according to a preferred embodiment of the present invention;

[0046] Figure 2 This is a flowchart of an infusion bag management method according to a preferred embodiment of the present invention. Detailed Implementation

[0047] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0048] Example 1

[0049] Reference Figure 1 As shown, an embodiment of the present invention provides an infusion bag management system, comprising:

[0050] The 3D recognition module includes a positioning submodule and an imaging submodule; the positioning submodule is used to obtain the position information of each infusion bag in the infusion cabinet; the imaging submodule is used to obtain imaging data of particulate matter suspended inside each infusion bag based on the position information.

[0051] The infusion bag monitoring module includes a monitoring submodule and a marking submodule. The monitoring submodule is used to monitor the liquid in the infusion bag and identify the current internal state of the infusion bag based on the monitoring results and imaging data. The marking submodule is used to mark the current state data of the infusion bag based on the internal state.

[0052] The verification module is used to encrypt and upload status data to the blockchain and perform multi-party verification to obtain the verification result.

[0053] The sorting module is used to acquire verification results and status data in real time and sort the infusion bags in the infusion cabinet.

[0054] The data management and traceability module is used to store imaging data, status data, verification results, and sorting data from the sorting module; it achieves encrypted traceability of sorting data by building a private chain.

[0055] This invention provides an infusion bag management system that enables real-time monitoring of infusion bag status, accurate verification of patient identity, instant updates of inventory data, and standardized management of medications. Through the collaborative work of the positioning and imaging submodules, the system can accurately acquire the location and internal status of infusion bags, updating this information in real time. This allows medical staff to monitor the infusion bags at any time, improving work efficiency and safety. The monitoring submodule can monitor the physical and chemical properties of the infusion bag fluid in real time, and, combined with imaging data, comprehensively assess the internal status of the infusion bag. The marking submodule automatically marks the status of the infusion bags based on the monitoring results, facilitating subsequent management and traceability, and reducing human error. The verification module uses blockchain technology to encrypt and upload status data to the blockchain, ensuring data security and immutability, enhancing system credibility, and supporting multi-party verification mechanisms to ensure data authenticity and consistency. Based on the verification results and status data, the sorting module can quickly and accurately classify and process the infusion bags, improving sorting efficiency, reducing manual intervention, minimizing operational errors, and thus improving overall work efficiency. The data management and traceability module centrally stores various types of data, ensuring data integrity and traceability, facilitating subsequent analysis and management. Private blockchain technology is used to encrypt and trace sorting data, ensuring data security and traceability, and improving the overall reliability of the system. This embodiment of the invention, through the collaborative work of the above modules, achieves high-precision identification, comprehensive monitoring, intelligent tagging, security verification, efficient sorting, and data traceability of infusion bags, while saving resources and costs.

[0056] Specifically, the 3D recognition module adopts a multimodal perception fusion architecture, and realizes UHF RFID phase positioning and terahertz imaging through the positioning submodule and imaging submodule, thereby obtaining the position information of each infusion bag in the infusion cabinet and the imaging data of the internal particulate matter of each infusion bag.

[0057] Specifically, the positioning submodule includes multiple readers and a phase extraction unit. The readers communicate with the tags on the infusion bags and extract the EPC information from the tags; the supported protocol is the EPC Gen2v2 protocol. The phase extraction unit processes the transmit and receive subcarrier signals, extracting the phase of these two signals for ranging and positioning.

[0058] Furthermore, communication between the reader and the tag on the infusion bag is based on the ITF (Interrogator Talk First) mechanism, i.e., a half-duplex mechanism. The process of selecting the tag's returned PC+EPC+CRC16 information is used as a reference for signal phase extraction and tag positioning. During the positioning process, multiple readers will measure the distance to tags on the same infusion bag. Moreover, both signal phase extraction and tag positioning are implemented in the phase extraction unit.

[0059] Furthermore, the phase extraction unit includes a ranging setting component, a conversion component, a distance calculation component, and a positioning component. The specific functions and principles of each component are as follows:

[0060] In the ranging setting component, select the reader's transmitted signal carrier frequency fc and set the ranging range. To resolve the phase ambiguity issue within the ranging range, a single-frequency subcarrier amplitude modulation method is used. Based on the ranging range, select the subcarrier frequency and ensure that the phase change range of the subcarrier is within a suitable interval.

[0061] For example, the reader transmits a signal carrier frequency fc of 915MHz, and the ranging range is set to 0.3–5m. Based on the ranging range, the subcarrier frequency is selected as 2MHz, at which point the phase change range of the subcarrier is 1.44°–30°, which is within a suitable range.

[0062] In the conversion component, a low-frequency subcarrier signal is modulated onto the CW signal using AM modulation. The transmitted signal s(t) and the received signal r(t) are then bandpass sampled and A / D converted before being sent into the digital domain.

[0063] The distance calculation component is used to estimate the phase of the subcarrier component in the transmit / receive signal sent by the conversion component, and calculate the distance between the tag on the infusion bag and the tag reading device based on the phase.

[0064] For example, in the distance calculation component, a discrete spectrum correction method is used to estimate the phase of the subcarrier component in the transmit and receive signals. and The phase difference between the transmitted and received subcarrier signals was calculated. Let the subcarrier frequency be f0, then the distance d between the reader and the tag is expressed as:

[0065]

[0066] Where c is the speed of light and π is the symbol for pi.

[0067] A positioning component is used to determine the position information of the tag on the infusion bag based on a distance d. Within the positioning component, data from multiple readers measuring the distance to the same tag is acquired. Each reader obtains its distance information to the tag using the phase extraction and ranging methods described above. Combining this with the least squares method of PDoA (Phase Difference of Arrival), the tag's position information is determined based on the distances measured by each reader.

[0068] To verify the positioning accuracy of the positioning submodule, 18,000 grasping tests were conducted. The positioning submodule performed excellently in positioning irregularly shaped medicines (such as ampoules), with the positioning error controlled within ±2 to 8 mm. Further analysis revealed a significant negative correlation between this error and the object size; that is, the larger the object, the smaller the positioning error.

[0069] Specifically, the imaging submodule includes a transmission component, a detector, an imaging component, and a data processing component. The terahertz source in the transmission component can be either a laser-based terahertz pulse source or a continuous-wave terahertz source. Laser pulse sources are suitable for time-resolved terahertz imaging, capable of capturing rapidly changing signals; while continuous-wave sources are more suitable for high-resolution spectral imaging, providing finer spectral information. For the detector, thermoelectric detectors, photoconductive detectors, or quantum cascade detectors can be used, which can accurately detect the intensity and phase information of terahertz waves. The imaging component is responsible for focusing the terahertz waves and forming an image. The data processing component uses advanced signal processing and image reconstruction algorithms to convert the terahertz signal acquired by the detector into a high-resolution image, thereby obtaining imaging data of particulate matter suspended inside the infusion bag.

[0070] Furthermore, the 3D recognition module also includes millimeter-wave radar for detecting leaks in the infusion bag. During detection, the millimeter-wave radar monitors minute changes on the surface of the infusion bag and its surrounding environment by emitting high-frequency millimeter-wave signals and receiving their reflected signals. When leakage occurs, the dielectric properties of the liquid alter the amplitude and phase of the reflected signal, which is then accurately detected by the radar. The 3D recognition module analyzes these changes in real time and immediately triggers an alarm when an abnormal signal is detected.

[0071] The 3D recognition module integrates heterogeneous sensor technologies, including UHF RFID phase matrix (0.3-5m dynamic tuning), terahertz band imaging (0.1-0.3THz), and 60GHz millimeter-wave radar for leakage detection, enabling full-dimensional perception of the internal and external conditions of the infusion bag. Through error compensation algorithms (such as least squares method, Kalman filter, and neural network), the liquid volume detection error is controlled to less than 0.8%, ensuring high accuracy and reliability of the detection results.

[0072] Specifically, the monitoring submodule includes a density detection unit, a contamination detection unit, and an identification unit. These units work collaboratively to comprehensively monitor the liquid's state. The density detection unit obtains the liquid's density and viscosity information by measuring frequency deviations. The contamination detection unit acquires the liquid's spectral information in the terahertz band and extracts characteristic information related to microbial contamination based on this; subsequently, it calculates the liquid's microbial contamination index. The identification unit, based on the extracted key parameters of density, viscosity, and microbial contamination index, establishes an identification model based on these parameters to achieve accurate identification of the liquid's state. When establishing the identification model, various methods such as Bayesian methods, deep learning, and machine learning can be employed to ensure the model's accuracy and reliability.

[0073] Furthermore, the monitoring submodule also includes an environmental sensing unit to comprehensively monitor the environmental conditions inside the medicine cabinet. The environmental sensing unit includes a temperature and humidity detection component, an air pressure detection component, a light intensity detection component, and an alarm component. Specifically: the temperature and humidity detection component monitors the temperature and humidity inside the medicine cabinet in real time; the air pressure detection component monitors the air pressure inside the medicine cabinet in real time; the light intensity detection component monitors the light intensity inside the medicine cabinet in real time; and the alarm component determines whether there are any abnormalities in the temperature, humidity, air pressure, light intensity, and the fluid in the infusion bag, and issues an emergency alarm signal when any abnormality is detected.

[0074] For example, in the temperature and humidity detection component, a MEMS temperature and humidity array is used, with a temperature monitoring range of -20 to 50°C and an accuracy of ±0.2 to 0.5°C; the humidity monitoring range is 5% to 98% RH, and the response time is less than 8 seconds. In the density detection unit, a surface acoustic wave (SAW) sensor array is used, with a liquid density detection accuracy of ±0.005 to 0.02 g / cm³. 3 The frequency deviation detection accuracy is ±0.1ppm. By combining a MEMS temperature and humidity array with a surface acoustic wave (SAW) sensor, an early warning function for infusion deterioration can be realized. When the pH value deviation of the liquid exceeds 0.3, the system will automatically trigger an alarm to ensure infusion safety.

[0075] Furthermore, the integration of a surface acoustic wave (SAW) sensor array (frequency deviation detection accuracy ±0.1ppm) with terahertz time-domain spectroscopy (THz-TDS) allows for the simultaneous acquisition of liquid density (accuracy ±0.002 g / cm³), viscosity (accuracy ±0.05 cP), and microbial contamination index. This combined technology not only improves the accuracy and reliability of detection but also comprehensively monitors the physical and chemical properties of liquids, providing strong support for pharmaceutical quality control.

[0076] Furthermore, the labeling submodule employs quantum dot spectral labeling technology. Through nano-fluorescent encoding, this technology can not only label the current state of the infusion bag but also accurately label batch information, particularly for high-risk drugs. This labeling method provides efficient and reliable technical support for drug management and traceability. In addition, the status data covers important information such as the deterioration status of the infusion bag and batch data (including production, transportation, preparation, and use), ensuring traceability throughout the entire drug lifecycle and refined management.

[0077] Specifically, the verification module is used to encrypt and upload the status data to the blockchain and perform multi-party verification to obtain the verification result. The specific operation steps of the verification module are as follows:

[0078] Step 1: Encrypt the state data using methods including AES-256 and RSA encryption algorithms. Generate a hash value from the encrypted state data; in this embodiment, the SHA-256 hash algorithm is used. The generation of the hash value ensures the integrity and uniqueness of the data. Any minor modification to the data will result in a significant change in the hash value, thus allowing for rapid detection of data tampering.

[0079] Step 2: Package the hash value and encrypted data into a block and write the block to the private blockchain via a smart contract. The private blockchain is built on the Hyperledger Fabric framework, supporting efficient transaction processing and data storage. Before the block is written to the private blockchain, a consensus mechanism (such as the improved HotStuff consensus algorithm) is used to verify the block, ensuring the correctness and consistency of the block data. The consensus mechanism ensures that all participating nodes reach a consensus on the block content, thereby guaranteeing the stability and reliability of the blockchain.

[0080] Step 3: Conduct multi-party verification. Invite multiple authorized participants (such as the hospital information center, pharmacy department, regulatory authorities, etc.) to join the verification process. Share the encrypted status data and its hash value with each participant, and provide the necessary decryption keys so that the participants can verify the authenticity of the data.

[0081] Step 4: Each participant uses the same hash algorithm to hash the state data and compares it with the hash value on the blockchain. If the hash values ​​match, the verification passes; otherwise, the verification fails. This process ensures that data undergoes rigorous verification at every stage from generation to storage, thereby guaranteeing the authenticity and integrity of the data.

[0082] Specifically, the sorting module includes a prescription-driven packaging submodule and a gripping submodule, which work together to achieve efficient sorting and management of medicines. The prescription-driven packaging submodule is used to parse electronic prescriptions (including accurate verification of patient identity), automatically generate drug dispensing plans, and drive a servo motor to complete quantitative dispensing, thereby obtaining the dispensed medicines. The gripping submodule is used to grip the medicines (i.e., the original medicines) and the dispensed medicines during sorting, ensuring that the medicines can be accurately transported to the designated location.

[0083] Furthermore, the medical order-driven encapsulation submodule includes a packaging unit, which is used to construct a medical knowledge graph-driven model and automatically generate drug formulation schemes.

[0084] For example, the prescription-driven encapsulation submodule parses prescriptions based on the HL7 FHIR protocol. In the dispensing stage, a piezoelectric ceramic injection valve is used for precise dispensing of the medication. This valve achieves droplet volume control accuracy of ±0.1–0.3 μL, ensuring high dispensing precision. Furthermore, to achieve efficient management and traceability of drug information, this module also uses laser-induced graphene (LIG) technology to print NFC smart tags. These tags can store the drug's temperature profile and batch information, with a read / write speed of less than 0.5 seconds, improving the level of intelligence in drug management.

[0085] For example, the dispensing unit preferably uses the Transformer-XL model as the core of the medical knowledge graph-driven process. This model has 12-24 dynamically adjustable attention heads, enabling it to efficiently process complex medical orders and generate accurate drug dispensing plans. In the dispensing stage, a piezoelectric ceramic injection valve is used for precise dispensing of the drug solution. This injection valve achieves a droplet volume control accuracy of ±0.05μL. Combined with the Transformer-XL model, it enables precise dosage distribution for multiple drug combinations, with errors controlled within ±0.3%. Through advanced piezoelectric ceramic technology, high-precision control of droplet volume is ensured, thereby guaranteeing high precision and consistency in dispensing.

[0086] Furthermore, the prescription-driven packaging submodule can employ dynamic radiation shielding packaging technology during the packaging process. This technology utilizes a lead-rubber-tungsten powder composite material (lead equivalent 0.25-0.35 mmPb) to construct an intelligent variable stiffness structure, and leverages shape memory alloy (SMA) to achieve on-demand shielding of chemotherapy drug packaging, ensuring a gamma ray attenuation rate exceeding 99.9%.

[0087] Furthermore, the grasping submodule includes a quantum dot spectral recognition unit and an execution unit. The quantum dot spectral recognition unit is used to acquire drug information and identify the drug to be grasped in the infusion cabinet based on the drug information, thereby driving the execution unit to grasp the drug.

[0088] For example, the execution unit preferably uses a six-axis collaborative robotic arm equipped with a quantum dot spectral recognition unit with a spectral resolution of 1–5 nm. With this advanced configuration, the six-axis collaborative robotic arm can achieve efficient sorting of infusion bags or bottles at a speed of 50–80 pieces / minute, with a load range of 0.2–8 kg. After 23,000 rigorous tests, the breakage rate is extremely low, below 0.005%, ensuring high precision and reliability of the operation.

[0089] For example, the execution unit employs a six-axis collaborative robotic arm equipped with force-position hybrid control technology, achieving a contact force closed-loop accuracy of ±0.02N. Simultaneously, it incorporates a quantum dot spectral recognition unit based on InP / ZnS quantum dot fluorescent labeling, with a wavelength resolution of 1–5 nm, constructing a spectral feature library. Through these technologies, the robotic arm can grasp glass ampoules with a breakage rate of less than 0.003%, achieving a high standard of zero breakage and further enhancing operational safety and reliability.

[0090] Furthermore, to improve the positioning accuracy and operational reliability of the six-axis collaborative robotic arm, a combined opto-mechanical-electronic calibration method can be employed. Specifically, by combining a femtosecond laser interferometer (wavelength stability ±0.01nm) with a MEMS micromirror array, sub-micron-level online calibration of the robotic arm's end effector can be achieved, ensuring a repeatability accuracy of ±5μm. This high-precision calibration method can effectively improve the stability and accuracy of the robotic arm in complex operating environments, meeting the needs of high-precision industrial applications.

[0091] Furthermore, for the grasping path planning of the grasping submodule, a spiking neural network (SNN) algorithm, which mimics the neural network of the human cerebellum, is adopted. This algorithm can realize the real-time generation of grasping paths for irregularly shaped medicine bottles in 3D point cloud space (resolution range of 0.1-0.5mm), with the planning time controlled to less than 50 milliseconds, significantly improving the efficiency and accuracy of grasping.

[0092] Through the above design, the sorting module can not only efficiently complete the sorting task of medicines, but also quickly identify and handle abnormal medicines and promptly clear the infusion cabinet inventory. At the same time, this module ensures accurate recording and full traceability of medicine information, providing solid and powerful support for the drug management of medical institutions.

[0093] Specifically, the data management and traceability module performs several functions. First, it stores imaging data, status data, verification results, and sorting data generated by the sorting module, ensuring the proper preservation and readily accessible nature of all information. Second, it achieves encrypted traceability of sorting data by constructing a private blockchain. Third, this module employs a blockchain architecture and dynamic security strategies to comprehensively guarantee data security and traceability.

[0094] Furthermore, during the encrypted traceability process of sorted data, the data management and traceability module ensures data security and immutability by constructing a private blockchain. Specifically, an improved HotStuff consensus algorithm is used to build the private blockchain. This algorithm has a Byzantine fault tolerance rate of over 33%, effectively resisting node failures and malicious attacks, and ensuring the stability and reliability of on-chain data. Simultaneously, zero-knowledge proof (ZK-SNARKs) technology is used to achieve encrypted traceability of drug distribution data. In this process, the on-chain latency is strictly controlled to less than 0.8 seconds, while the system throughput (TPS) exceeds 3000 transactions per second, significantly improving data processing efficiency and providing efficient and secure technical support for the full lifecycle management of pharmaceuticals.

[0095] Furthermore, regarding the blockchain architecture, a consortium blockchain was built using the Hyperledger Fabric 2.3 framework, ensuring that all data related to the entire lifecycle of pharmaceuticals, from production and configuration to use and recycling, is fully recorded on the blockchain. After multiple rounds of rigorous stress testing, transaction processing latency was precisely controlled within 0.2–1.5 seconds, demonstrating a significant performance advantage compared to traditional centralized methods (whose processing latency typically exceeds 3 seconds), and greatly improving data processing efficiency.

[0096] Furthermore, if the infusion volume deviates from the prescribed value by 5-20%, the system will quickly and automatically initiate a biometric verification process. Iris recognition technology, with its extremely low false recognition rate (only 1e-6 to 1e-5), and finger vein recognition technology, with its relatively low false recognition rate (1e-5 to 1e-4), together ensure high accuracy and reliability in the verification process. For special drugs such as chemotherapy drugs, double-layer radiation-proof packaging is used, with a lead equivalent of 0.25-0.35 mmPb, effectively blocking radiation leakage. It is also equipped with gamma-ray sterilization traceability tags, enabling full traceability of the sterilization process for these special drugs, ensuring drug safety and traceability from multiple dimensions.

[0097] Through the above design, the data management and traceability module not only ensures the immutability and high transparency of drug data, but also enhances the security and reliability of drug management through dynamic security strategies. Furthermore, the module is responsible for storing imaging data, status data, verification results, and sorting data from the sorting module, and achieves encrypted traceability of sorting data by building a private blockchain, providing medical institutions with a comprehensive data management and traceability solution.

[0098] Furthermore, the infusion bag management system provided in this embodiment of the invention also includes an intelligent temperature control module. This module works closely with the various components in the environmental sensing unit to precisely regulate the temperature and humidity of the drug storage environment, significantly reducing energy consumption for drug storage and effectively extending the shelf life of photosensitive drugs. This intelligent temperature control management provides strong support for the long-term stable storage of drugs, further enhancing the system's practicality and economic benefits.

[0099] Furthermore, the infusion bag management system provided in this embodiment of the invention also includes a control platform that interacts bidirectionally with the hospital's HIS (Hospital Information System) / intravenous compounding center system, supporting hierarchical management of access permissions for special drugs. In addition, the 3D recognition module also transmits the leakage location and related information to the control platform, enabling medical staff to take timely measures to ensure the safety and reliability of the infusion process.

[0100] Furthermore, the infusion bag management system provided in this embodiment of the invention also employs a self-healing safety control protocol. This protocol is based on a formally verified deep reinforcement learning (DRL) framework and uses runtime verification technology to detect control command anomalies in real time, achieving a detection coverage rate exceeding 99.999%. Once an anomaly is detected, the system can quickly activate redundant actuators to achieve fault self-healing, with a switching time controlled to less than 10 milliseconds. This highly reliable control mechanism effectively ensures the stable operation of the system, reduces downtime caused by faults, and improves the overall operational safety and reliability.

[0101] Furthermore, the infusion bag management system provided in this embodiment of the invention employs an edge-cloud collaborative control circuit. This system is based on an NVIDIA Jetson Orin FPGA heterogeneous computing architecture with a reconfigurable computing power range between 275 and 750 TOPS. Simultaneously, the system incorporates a quantum tunneling effect sensing circuit, possessing extremely high current sensitivity, reaching 10... -12 Level A. Thanks to this, the system can achieve millisecond-level anomaly response, with latency controlled to less than 5 milliseconds, thereby ensuring real-time and accurate monitoring of the infusion bag status.

[0102] Furthermore, compared to traditional management systems, the embodiments of this invention exhibit several advantages. Specifically, traditional infusion cabinets have significant blind spots in multimodal recognition; for example, the detection error for the volume of liquid in transparent soft bags is as high as 5% or more. In terms of special drug management, the compliance rate for storing light-proof and temperature-controlled drugs is less than 85%. In addition, the problem of medical data silos is prominent, with a serious disconnect between medical order execution and drug traceability. To address these issues, this invention systematically integrates a series of cutting-edge technologies. First, it introduces terahertz imaging technology to accurately detect particulate matter suspended inside the infusion bag, with a resolution of up to 0.1 mm.3 Secondly, quantum dot spectral labeling technology is employed to rapidly identify high-risk drug batches through nano-fluorescent coding, with a response time of less than 50ms. Thirdly, an array of surface acoustic wave (SAW) sensors is included, enabling simultaneous detection of liquid density, viscosity, and temperature changes with an accuracy of ±0.01g / cm³. Building upon this foundation, a six-axis force feedback robotic arm is integrated to achieve non-destructive handling of irregularly shaped drug packaging (such as ampoules and pre-filled syringes), thereby comprehensively enhancing the performance and safety of the infusion cabinet.

[0103] Furthermore, based on a medical knowledge graph, this system accurately analyzes electronic medical orders, automatically generates drug dispensing plans, and achieves quantitative dispensing of drugs through a high-precision servo motor (accuracy up to ±1mL). For drug labeling, laser-induced graphene (LIG) technology is used to print tamper-proof labels. These labels integrate NFC chips and thermochromic ink, effectively preventing information tampering and monitoring drug status in real time. Regarding full-cycle drug traceability, the system builds a traceability network based on a consortium blockchain architecture, utilizing the Hyperledger Fabric framework to encrypt and upload infusion batch data (covering production, transportation, preparation, and use) to the blockchain for multi-party verification, ensuring data immutability. Its Byzantine fault tolerance exceeds 33%. Compared with existing technologies, this system improves the security and efficiency of drug management. The mis-dispensing rate of high-risk drugs is reduced from 0.07% to 0.002%, and the timeliness of cold chain disruption risk warnings is improved by 12 times, greatly enhancing the reliability and traceability of drug management.

[0104] Furthermore, compared to traditional systems, this system exhibits higher accuracy and reliability. Based on multimodal sensing fusion technology (UHF RFID positioning error ±3-8mm, terahertz foreign object detection resolution 0.05-0.3mm), it achieves this. 3 The system achieves an accuracy rate of ≥99.98% for identifying infusion drugs (compared to ≤96.5% for traditional systems), and the dosage error for medication orders is stabilized within ±0.2-0.5% (compared to ±3-8% for manual operation). Through quantum dot spectral sorting (breakage rate <0.005%) and piezoelectric ceramic injection valves (dispensing accuracy ±0.1-0.3μL), the qualified rate of drug preparation is increased to 99.7-99.9% (compared to 92-95% for traditional methods).

[0105] Furthermore, this system has been optimized in terms of efficiency and energy consumption compared to traditional systems. The six-axis collaborative robotic arm achieves a sorting speed of 45-80 pieces / minute (compared to 15-25 pieces / minute for manual operation), the edge computing node (NVIDIA Jetson Orin) has a real-time response latency of <5-20ms, and power consumption is reduced by 60-75% compared to traditional industrial PCs (peak power consumption 15-35W). The prescription parsing engine is based on the Transformer model (processing speed 0.5-1.2 seconds / prescription), which improves configuration efficiency by 2.6-4.1 times, and the average daily processing capacity of a single cabinet reaches 300-500 doses (compared to 120-180 doses for traditional cabinets).

[0106] Furthermore, this system is enhanced in terms of security and traceability compared to traditional systems. This system (TPS2000-3000, latency 0.3-1.5 seconds) enables the on-chain storage of drug lifecycle data, reducing the traceability time for abnormal operations from the traditional 48-72 hours to 5-15 seconds. Dynamic security strategies (iris recognition false recognition rate 1e-6-1e-5) combined with radiation shielding encapsulation (lead equivalent 0.25-0.35 mmPb) reduce the chemotherapy drug mis-mixing rate to 0.003-0.018‰ (compared to 0.12-0.35‰ in traditional systems), and reduce nurses' radiation exposure by 90-95%.

[0107] Furthermore, compared to traditional systems, this system is more economical in terms of resources and costs. The intelligent temperature control module (semiconductor cooling COP value 3.2-4.5) reduces drug storage energy consumption by 40-60% and extends the shelf life of photosensitive drugs by 2-3 months. The robotic arm path planning algorithm (SNN response time <50ms) reduces equipment wear and tear, increasing the lifespan of key components by 3-5 times (to 80,000-120,000 hours). The user interface integrates voice and touch interaction, reducing the time for a single nurse operation to 8-15 seconds (compared to 45-90 seconds in traditional processes), and reducing labor intensity by 70-85%.

[0108] It should be noted that the above data were obtained through rigorous experimental verification and practical application testing, covering system performance evaluation and comparative analysis in multiple scenarios.

[0109] This invention relates to an infusion bag management system designed based on the Internet of Things (IoT) and multimodal sensing technologies. This system can be used for the full-process digital management and control of intravenous infusion drugs in medical institutions. The system highly integrates UHF RFID electronic tags, multispectral environmental sensors, edge computing nodes, and biometric modules. Utilizing intelligent shelves, it achieves precise positioning of infusion preparations, real-time dynamic inventory updates, automatic early warning of near-expiration dates (error range controlled within ±24 hours, warning time error not exceeding 5 minutes, data obtained through multiple experiments), and accurate traceability of operator biometrics. The built-in control platform achieves bidirectional interaction with the hospital's HIS system and intravenous compounding center system, supporting hierarchical access control for special drugs such as chemotherapy drugs and antibiotics. Furthermore, the system is equipped with a 3D weight sensor array, achieving gram-level accuracy verification of drug dispensing (error range controlled within ±1g), effectively avoiding errors that may occur with traditional manual verification.

[0110] Example 2

[0111] Based on the same inventive concept, this embodiment provides an infusion therapy management method. The principle of solving the problem is similar to that of the infusion bag management system provided in Embodiment 1, and the repeated parts will not be described again.

[0112] Reference Figure 2 As shown, this embodiment provides a method for managing intravenous infusion therapy, including but not limited to the following steps:

[0113] Acquire the position information of each infusion bag in the infusion cabinet; based on the position information, acquire imaging data of particulate matter suspended inside each infusion bag;

[0114] The fluid in the infusion bag is monitored, and the current internal state of the infusion bag is identified based on the monitoring results and imaging data.

[0115] Based on the internal state, mark the current status data of the infusion bag; encrypt the status data, upload it to the blockchain, and verify it from multiple parties to obtain the verification result;

[0116] Obtain verification results and status data, sort the infusion bags in the infusion cabinet, and record the sorting data;

[0117] It stores imaging data, status data, verification results, and sorting data, and completes encrypted traceability of sorting data by building a private chain.

[0118] Example 3

[0119] This embodiment provides an infusion cabinet, including an infusion bag management system provided in Embodiment 1.

[0120] The infusion cabinet provided in this embodiment effectively solves the problems existing in the prior art through functions such as temperature and humidity linkage control, infusion bag deformation monitoring, biometric technology, real-time environmental monitoring, accurate inventory management, and secure drug requisition permission management. This enables the stability and security of the drug storage environment, real-time monitoring of the infusion bag status, accurate verification of patient identity, instant updates of inventory data, and standardization of drug management, significantly improving medical efficiency and patient medication safety.

[0121] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0122] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0123] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0124] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0125] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. An infusion bag management system, characterized by, The method comprises the following steps: A three-dimensional identification module comprises a positioning sub-module and an imaging sub-module; The positioning sub-module is used to obtain the position information of each infusion bag in the infusion cabinet; The imaging sub-module is used to obtain the imaging data of the internal particulate suspension of each infusion bag according to the position information; An infusion bag monitoring module comprises a monitoring sub-module and a marking sub-module; the monitoring sub-module is used to monitor the liquid in the infusion bag and identify the internal state of the current infusion bag according to the monitoring result and the imaging data; The marking sub-module is used to mark the state data of the current infusion bag according to the internal state; A verification module is used to encrypt and chain the state data and perform multi-party verification to obtain a verification result; A sorting module is used to obtain the verification result and the state data in real time and sort the infusion bags in the infusion cabinet; A data management and traceability module is used to store the imaging data, the state data, the verification result and the sorting data in the sorting module; The sorting data is encrypted and traced through a private chain.

2. The infusion bag management system of claim 1, wherein, The sorting module comprises: A medical order driving packaging sub-module is used to analyze electronic medical orders, automatically generate a drug ratio scheme and drive a servo motor to complete quantitative dispensing to obtain dispensed drugs; A grabbing sub-module is used to grab the drugs during sorting and the dispensed drugs.

3. The infusion bag management system of claim 2, wherein, The medical order driving packaging sub-module comprises a dispensing unit, which is used to construct a medical knowledge graph driving model to automatically generate a drug ratio scheme.

4. The infusion bag management system of claim 2, wherein, The grabbing sub-module comprises a quantum dot spectrum recognition unit and an execution unit, the quantum dot spectrum recognition unit is used to obtain drug information and identify the drugs to be grabbed in the infusion cabinet according to the drug information to drive the execution unit to grab.

5. The infusion bag management system of claim 1, wherein, The positioning sub-module comprises a phase extraction unit, which comprises: A distance measurement setting component is used to set a distance measurement range and determine the phase change range of the subcarrier at this time according to the distance measurement range; A conversion component is used to convert the transmission signal and the reception signal into the digital domain after bandpass sampling and conversion according to the phase change range; A distance calculation component is used to estimate the phase of the subcarrier component in the transceiver signal sent by the conversion component and calculate the distance between the label on the infusion bag and the label reading device according to the phase; A positioning component is used to determine the position information of the label on the infusion bag according to the distance.

6. The infusion bag management system of claim 1, wherein, The monitoring sub-module comprises: A density detection unit is used to obtain the density and viscosity of the liquid by measuring the deviation of the frequency; A pollution detection unit is used to obtain the spectrum information of the liquid in the terahertz wave band; to extract feature information related to microbial pollution according to the spectrum information; and to calculate the microbial pollution index of the liquid according to the feature information; An identification unit is used to extract the key parameters of the density, the viscosity and the microbial pollution index; and to establish an identification model according to the key parameters.

7. The infusion bag management system of claim 1 or 6, wherein, The monitoring sub-module further comprises an environment perception unit, which comprises: A temperature and humidity detection component is used to monitor the temperature and humidity in the medicine cabinet in real time; An air pressure detection component is used to monitor the air pressure in the medicine cabinet in real time; An illumination detection component is configured to monitor the illumination intensity in the medicine cabinet; An alarm component is configured to determine whether the temperature and humidity, the air pressure, the illumination intensity, and the liquid in the infusion bag are abnormal, and to send an emergency alarm signal when an abnormality exists.

8. The infusion bag management system of claim 1, wherein, The system further comprises a control platform that interacts with a hospital HIS / compounding center system in both directions and supports hierarchical management of special medicines.

9. An infusion bag management method characterized by, The system comprises: acquiring position information of each infusion bag in the infusion cabinet; acquiring imaging data of internal particulate suspensions of each infusion bag according to the position information; monitoring the liquid in the infusion bag and identifying the internal state of the current infusion bag according to the monitoring result and the imaging data; marking state data of the current infusion bag according to the internal state; encrypting and chaining the state data and performing multi-party verification to obtain a verification result; acquiring the verification result and the state data, sorting the infusion bags in the infusion cabinet, and recording sorting data; storing the imaging data, the state data, the verification result, and the sorting data, and completing encrypted traceability of the sorting data by constructing a private chain.

10. An infusion cabinet characterized in that, The system comprises an infusion bag management system according to any one of claims 1-8.

Images